1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
£EPA
[EPA ID for final]
www.epa.gov/iris
TOXICOLOGICAL REVIEW
OF
T richloroethylene
(CAS No. 79-01-6)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
June 2009
NOTICE
This document is an INTER-AGENCY REVIEW draft. This information is distributed solely
for the purpose of pre-dissemination peer review under applicable information quality guidelines.
It has not been formally disseminated by U.S. 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.
U.S. Environmental Protection Agency
Washington, DC
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 DISCLAIMER
2
3
4 This document is a preliminary draft for review purposes only. This information is
5 distributed solely for the purpose of pre-dissemination peer review under applicable information
6 quality guidelines. It has not been formally disseminated by EPA. It does not represent and
7 should not be construed to represent any Agency determination or policy. Mention of trade
8 names or commercial products does not constitute endorsement or recommendation for use.
9
6/22/2009
li
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
SUMMARY
There is substantial potential for human exposure to TCE, as it has a widespread presence
in ambient air, indoor air, soil, and groundwater. At the same time, humans are likely to be
exposed to a variety of compounds that are either metabolites of TCE or which have common
metabolites or targets of toxicity. Once exposed, humans, as well as laboratory animal species,
rapidly absorb TCE, which is then distributed to tissues via systemic circulation, extensively
metabolized, and then excreted primarily in breath as unchanged TCE or C02, or in urine as
metabolites.
Based on the available human epidemiologic data and experimental and mechanistic
studies, it is concluded that TCE poses a potential human health hazard for non-cancer toxicity to
the central nervous system, the kidney, the liver, the immune system, the male reproductive
system, and the developing fetus. The evidence is more limited for TCE toxicity to the
respiratory tract and female reproductive system. Following U.S. EPA (2005a) Guidelines for
Carcinogen Risk Assessment, TCE is characterized as carcinogenic in humans by all routes of
exposure. This conclusion is based on convincing evidence of a causal association between TCE
exposure in humans and kidney cancer. The human evidence of carcinogenicity from
epidemiologic studies of TCE exposure is compelling for Non-Hodgkins Lymphoma (NHL) but
less convincing than for kidney cancer, and more limited for liver and biliary tract cancer.
Further support for the characterization of TCE as carcinogenic in humans by all routes of
exposure is derived from positive results in multiple rodent cancer bioassays in rats and mice of
both sexes, similar toxicokinetics between rodents and humans, mechanistic data supporting a
mutagenic MOA for kidney tumors, and the lack of mechanistic data supporting the conclusion
that any of the MOA(s) for TCE-induced rodent tumors are irrelevant to humans.
As TCE toxicity and carcinogenicity are generally associated with TCE metabolism,
susceptibility to TCE health effects may be modulated by factors affecting toxicokinetics,
including lifestage, gender, genetic polymorphisms, race/ethnicity, pre-existing health status,
lifestyle, and nutrition status. In addition, while these some of these factors are known risk
factors for effects associated with TCE exposure, it is not known how TCE interacts with known
risk factors for human diseases.
For non-cancer effects, the most sensitive types of effects, based either on human
equivalent concentrations/doses or on candidate RfCs/RfDs, appear to be developmental, kidney,
and immunological (adult and developmental) effects. The neurological and reproductive effects
appear to be about an order of magnitude less sensitive, with liver effects another two orders of
magnitude less sensitive. The preferred RfC estimate of 0.001 ppm (1 ppb or 5 (^g/m3) is based
on route-to-route extrapolated results from oral studies for the critical effects of heart
6/22/2009
in
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
malformations (rats), immunotoxicity (mice), and toxic nephropathy (rats, mice), and an
inhalation study for the critical effect of increased kidney weight (rats). Similarly, the preferred
RfD estimate for non-cancer effects of 0.0004 mg/kg/d is based on the critical effects of heart
malformations (rats), adult immunological effects (mice), developmental immunotoxicity (mice),
and toxic nephropathy (rats). There is high confidence in these preferred non-cancer reference
values, as they are supported by moderate- to high-confidence estimates for multiple effects from
multiple studies.
For cancer, the preferred estimate of the inhalation unit risk is 2 x 10~2 per ppm [4 x 10~6
per jug/m31, based on human kidney cancer risks reported by Charbotel et al. (2006) and
adjusted, using human epidemiologic data, for potential risk for tumors at multiple sites. The
preferred estimate of the oral unit risk for cancer is 5 x 10~2 per mg/kg/d, resulting from PBPK
model-based route-to-route extrapolation of the inhalation unit risk estimate based on the human
kidney cancer risks reported in Charbotel et al. (2006) and adjusted, using human epidemiologic
data, for potential risk for tumors at multiple sites. There is high confidence in these unit risks
for cancer, as they are based on good quality human data, as well as being similar to unit risk
estimates based on multiple rodent bioassays. Because there is both sufficient weight of
evidence to conclude that TCE operates through a mutagenic MOA for kidney tumors and a lack
of TCE-specific quantitative data on early-life susceptibility, the default age-dependent
adjustment factors (ADAFs) can be applied for the kidney cancer component of the unit risks for
cancer; however, the application of ADAFs is likely to have a minimal impact on the total cancer
risk except when exposures are primarily during early life.
6/22/2009
iv
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
GUIDE TO READERS OF THIS DOCUMENT
Due to the length of the TCE toxicological review, it is recommended that Chapters 1 and 6
be read prior to Chapters 2-5.
Chapter 1 is the standard introduction to an IRIS Toxicological Review, describing the purpose
of the assessment and the guidelines used in its development.
Chapter 2 is an exposure characterization that summarizes information about TCE sources,
releases, media levels and exposure pathways for the general population (occupational exposure
is also discussed to a lesser extent).
Chapter 3 describes the toxicokinetics and physiologically-based pharmacokinetic (PBPK)
modeling of TCE and metabolites (PBPK modeling details are in Appendix A).
Chapter 4 is the hazard characterization of TCE. Section 4.0 summarizes the evaluation of
epidemiologic studies of cancer and TCE (qualitative details in Appendix B; meta-analyses in
Appendix C). Each of the sections 4.1-4.8 provides self-contained summary and syntheses of
the epidemiologic and laboratory studies on TCE and metabolites, organized by tissue/type of
effects, in the following order: genetic toxicity, central nervous system (CNS), kidney, liver,
immune system, respiratory tract, reproduction and development, and other cancers. Additional
details are provided in Appendix D for CNS effects and Appendix E for liver effects. Section
4.9 summarizes the available data on susceptible lifestages and populations. Section 4.10
describes the overall hazard characterization, including the weight of evidence for non-cancer
effects and for carcinogenicity.
Chapter 5 is the dose-response assessment of TCE. Section 5.1 describes the dose-response
analyses for non-cancer effects, and Section 5.2 describes the dose-response analyses for cancer.
Additional computational details are described in Appendix F for non-cancer dose-response
analyses, Appendix G for cancer dose-response analyses based on rodent bioassays, and
Appendix H for cancer dose-response analyses based on human epidemiologic data.
Chapter 6 is the summary of the major conclusions in the characterization of TCE hazard and
dose response.
6/22/2009
v
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
CONTENTS of TOXICOLOGICAL REVIEW for TRICHLOROETHYLENE
(CAS No. 79-01-6)
DISCLAIMER ii
SUMMARY iii
GUIDE TO READERS OF THIS DOCUMENT v
CONTENTS of TOXICOLOGICAL REVIEW for TRICHLOROETHYLENE vi
FOREWORD xxii
AUTHORS, CONTRIBUTORS, AND REVIEWERS xxiii
CHEMICAL MANAGER xxiii
AUTHORS xxiii
REVIEWERS xxvi
INTERNAL EPA REVIEWERS xxvi
EXTERNAL PEER REVIEWERS xxviii
ACKNOWLEDGMENTS xxviii
1 INTRODUCTION 1
2 EXPO SURE CHARACTERIZATION 5
2.1 ENVIRONMENTAL SOURCES 6
2.2 ENVIRONMENTAL FATE 11
2.3 EXPOSURE CONCENTRATIONS 11
2.4 EXPOSURE PATHWAYS AND LEVELS 23
2.4.1 General Population 23
2.4.1.1 Inhalation 23
2.4.1.2 Ingestion 25
2.4.1.3 Dermal 26
2.4.1.4 Exposure to TCE Related Compounds 27
2.4.2 Potentially Highly Exposed Populations 28
2.4.3 Exposure Standards 31
2.5 Exposure Summary 32
2.6 REFERENCES 32
3 TOXICOKINETICS 39
3.1 ABSORPTION 40
3.1.1 Oral 40
3.1.2 Inhalation 42
3.1.3 Dermal 50
6/22/2009
vi
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.2 DISTRIBUTION AND BODY BURDEN 52
3.3 METABOLISM 60
3.3.1 Introduction 60
3.3.2 Extent of Metabolism 60
3.3.3 Pathways of Metabolism 64
3.3.3.1 Cytochrome P450-Dependent Oxidation 64
3.3.3.1.1 Formation of trichloroethy 1 ene oxide 68
3.3.3.1.2 Formation of CH, TCOH and TCA 69
3.3.3.1.3 Formation of DCA and other products 71
3.3.3.1.4 Tissue distribution of oxidative metabolism and metabolites 73
3.3.3.1.5 Species-, Sex-, and age-dependent differences of oxidative metabolism....77
3.3.3.1.6 CYP isoforms and genetic polymorphisms 79
3.3.3.2 GSH Conjugation Pathway 82
3.3.3.2.1 Formation of DCVG 83
3.3.3.2.2 Formation of DCVC 87
3.3.3.2.3 Formation of NAcDCVC 88
3.3.3.2.4 Beta lyase metabolism of DCVC 88
3.3.3.2.5 Sulfoxidation of DCVC and NAcDCVC 89
3.3.3.2.6 Tissue distribution of GSH metabolism 90
3.3.3.2.7 Sex- and Species-dependent differences in GSH metabolism 93
3.3.3.2.8 Human variability and susceptibility in GSH conjugation 96
3.3.3.3 Relative Roles of the CYP and GSH Pathways 96
3.4 TCE EXCRETION 99
3.4.1 Exhaled Air 100
3.4.2 Urine 102
3.4.3 Feces 104
3.5 PBPK Modeling of TCE and Its Metabolites 130
3.5.1 Introducti on 130
3.5.2 Previous PBPK Modeling of TCE for Risk Assessment Application 130
3.5.3 Development and Evaluation of an Interim "Harmonized" TCE PBPK Model ..132
3.5.4 PBPK Model for TCE and Metabolites Used for this Assessment 135
3.5.4.1 Introduction 135
3.5.4.2 Updated PBPK Model Structure 135
3.5.4.3 Specification of PBPK model parameter prior distributions 139
3.5.4.4 Dose Metric Predictions 140
6/22/2009
vii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.5.5 Bayesian estimation of PBPK model parameters, and their uncertainty and
variability 141
3.5.5.1 Updated Pharmacokinetic Database 141
3.5.5.2 Updated Hierarchical Population Statistical Model 148
3.5.5.3 Use of interspecies scaling to update prior distributions in the absence of other
data 148
3.5.5.4 Implementation 152
3.5.6 Evaluation of Updated PBPK model 152
3.5.6.1 Convergence 152
3.5.6.2 Evaluation of posterior parameter distributions 154
3.5.6.3 Comparison of model predictions with data 161
3.5.6.3.1 Mouse model and data 163
3.5.6.3.2 Rat model and data 166
3.5.6.3.3 Human model 170
3.5.6.4 Summary Evaluation of Updated PBPK Model 174
3.5.7 PBPK Model Dose Metric Predictions 174
3.5.7.1 Characterization of Uncertainty and Variability 174
3.5.7.2 Implications for the Population Pharmacokinetics of TCE 189
3.5.7.2.1 Results 189
3.5.7.2.2 Discussion 190
3.5.7.3 Overall evaluation of PBPK model-based internal dose predictions 194
REFERENCES 197
4 Hazard Characterization 205
4.0 Epidemiologic studies on cancer and TCE—summary evaluation 205
References 235
4.1 Genetic toxicity 242
4.1.1 TCE 242
4.1.1.1 DNA binding Studies 242
4.1.1.2 Bacterial systems - Gene mutations 244
4.1.1.3 Fungal systems - Gene Mutations, conversions and recombination 246
4.1.1.4 Mammalian Systems and Human 248
4.1.1.4.1 Gene Mutations 248
4.1.1.4.2 von Hippel-Lindau (VHL) Gene Mutations 249
4.1.1.4.3 Chromosomal Aberrations 251
4.1.1.4.4 Micronucleus Induction 251
4.1.1.4.5 Sister Chromatid Exchanges (SCEs) 253
6/22/2009
viii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
.254
255
256
.257
265
.266
.266
.267
.267
.268
268
269
.270
270
.271
.271
.272
.272
.273
273
274
.274
274
.275
.276
.276
276
.278
.279
279
.279
279
.287
.287
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.1.1.4.6 Unscheduled DNA Synthesis
4.1.1.4.7 DNA Strand Breaks
4.1.1.4.8 DNA damage related to oxidative stress
4.1.1.4.9 Cell Transformation:
4.1.1.5 Summary
4.1.2 TCA (Trichloroacetic Acid)
4.1.2.1 Bacterial Systems - Gene Mutations
4.1.2.2 Mammalian Systems
4.1.2.2.1 Gene Mutations
4.1.2.2.2 Chromosomal Aberrations
4.1.2.2.3 Micronucleus
4.1.2.2.4 DNA Damage
4.1.2.2.5 Cell Transformation
4.1.2.3 Summary
4.1.3 Di chl oroaceti c Acid (DCA)
4.1.3.1 Bacterial and Fungal Systems - Gene Mutations
4.1.3.2 Mammalian Systems
4.1.3.2.1 Gene Mutations
4.1.3.2.2 Chromosomal Aberrations and Micronucleus
4.1.3.2.3 DNA Damage
4.1.3.3 Summary
4.1.4 Chloral Hydrate
4.1.4.1 DNA binding Studies
4.1.4.2 Bacterial and Fungal Systems - Gene Mutations
4.1.4.3 Mammalian Systems
4.1.4.3.1 Gene Mutations
4.1.4.3.2 Micronucleus Induction
4.1.4.3.3 Chromosomal Aberrations
4.1.4.3.4 Sister Chromatid Exchanges
4.1.4.3.5 DNA Damage
4.1.4.3.6 Cell Transformation
4.1.4.4 Summary
4.1.5 S-(l,2-dichlorovinyl)-L-Cysteine (DCVC) and S-dichlorovinyl glutathione
(DCVG) 280
4.1.6 Trichloroethanol (TCOH)
4.1.7 Synthesis and Overall Summary
6/22/2009
IX
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
293
.311
.311
.311
.313
.318
.320
321
.321
.324
328
330
.330
330
.331
332
332
.335
337
.338
338
.341
.343
.344
.344
344
347
.348
.348
.348
.351
353
.354
354
354
355
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.1.8 References:
4.2 Central Nervous System Toxicity
4.2.1 Alterations in Nerve Conduction
4.2.1.1 Trigeminal Nerve Function: Human Studies
4.2.1.2 Nerve Conduction Velocity - Human Studies
4.2.1.3 Trigeminal Nerve Function: Laboratory Animal Studies
4.2.1.4 Discussion and Conclusions: TCE-induced trigeminal nerve impairment
4.2.2 Auditory Effects
4.2.2.1 Auditory Function: Human Studies
4.2.2.2 Auditory Function: Laboratory Animal Studies
4.2.2.3 Summary and Conclusion of Auditory Effects
4.2.3 Vestibular function
4.2.3.1 Vestibular Function: Human Studies
4.2.3.2 Vestibular function: Laboratory animal data:
4.2.3.3 Summary and Conclusions for the Vestibular Function Studies
4.2.4 Visual Effects
4.2.4.1 Visual Effects: Human Studies
4.2.4.2 Visual Effects: Laboratory animal data
4.2.4.3 Summary and Conclusion of Visual Effects
4.2.5 Cognitive function
4.2.5.1 Cognitive Effects: Human Studies
4.2.5.2 Cognitive Effects: Laboratory animal studies
4.2.5.3 Summary and Conclusions of Cognitive Function Studies
4.2.6 Psychomotor Effects
4.2.6.1 Psychomotor effects: Human Studies
4.2.6.1.1 Reaction Time
4.2.6.1.2 Muscular Dyscoordination
4.2.6.2 Psychomotor effects: Laboratory animal data
4.2.6.2.1 Loss of righting reflex
4.2.6.2.2 Activity, sensory-motor and neuromuscular function
4.2.6.2.3 Locomotor activity
4.2.6.3 Summary and Conclusions for Psychomotor Effects
4.2.7 Mood Effects and Sleep Disorders
4.2.7.1 Effects on Mood: Human Studies
4.2.7.2 Effects on Mood: Laboratory animal findings
4.2.7.3 Sleep Disturbances
6/22/2009
x
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.2.8 Developmental neurotoxicity 355
4.2.8.1 Human Studies 355
4.2.8.2 Animal Studies 356
4.2.8.3 Summary and Conclusions for the Developmental Neurotoxicity Studies 360
4.2.9 Mechanistic studies of TCE neurotoxicity 361
4.2.9.1 Dopamine neuron disruption 361
4.2.9.1.1 Dopamine Neuron Disruption: Human Studies 362
4.2.9.1.2 Dopamine Neuron Disruption: Animal Studies 362
4.2.9.1.3 Summary and Conclusions of Dopamine Neuron Studies 363
4.2.9.2 Neurochemical and Molecular Changes 363
4.2.10 Potential Mechanisms for TCE-mediated Neurotoxicity 367
4.2.11 Overall Summary and Conclusions—Weight of Evidence 369
References 375
4.3 KIDNEY TOXICITY AND CANCER 386
4.3.1 Human studies of kidney 386
4.3.1.1 Nonspecific Markers of Nephrotoxicity 386
4.3.1.2 End-stage Renal Disease 389
4.3.2 Human studies of kidney cancer 389
4.3.2.1 Studies of Job Titles and Occupations with Historical TCE Usage 390
4.3.2.2 Cohort and Case-Controls Studies of TCE Exposure 390
4.3.2.2.1 Discussion of Controversies on Studies in the Arnsberg Region of Germany392
4.3.2.3 Examination of Possible Confounding Factors 395
4.3.2.4 Susceptible Populations - Kidney Cancer and TCE Exposure 397
4.3.2.5 Meta-analysis for Kidney Cancer 398
4.3.3 Human studies of Somatic Mutation of von Hippel-Lindau (VHL) Gene 420
4.3.4 Kidney non-cancer toxicity in laboratory animals 423
4.3.5 Kidney cancer in laboratory animals 433
4.3.5.1 Inhalation Studies of TCE 433
4.3.5.2 Gavage and Drinking Water Studies of TCE 434
4.3.6 Role of metabolism in TCE kidney toxicity 436
4.3.6.1 In vivo studies of the kidney toxicity of TCE metabolites 436
4.3.6.1.1 Role of GSH conjugation metabolites of TCE 436
4.3.6.1.2 Role of oxidative metabolites of TCE 440
4.3.6.2 In vitro studies of kidney toxicity of TCE and metabolites 444
4.3.6.3 Conclusions as to the active agents of TCE-induced nephrotoxicity 445
4.3.7 Mode(s) of Action for Kidney Carcinogenicity 446
6/22/2009
XI
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.3.7.1 Hypothesized Mode of Action: Mutagenicity 446
4.3.7.2 Hypothesized Mode of Action: Cytotoxicity and Regenerative Proliferation.449
4.3.7.3 Additional Hypothesized Modes of Action with Limited Evidence or
Inadequate Experimental Support 451
4.3.7.3.1 Peroxisome proliferation 451
4.3.7.3.2 a2|i-globulin-related nephropathy 452
4.3.7.3.3 Formic acid-related nephrotoxicity 452
4.3.7.4 Conclusions about the Hypothesized Modes of Action 453
4.3.8 Summary: TCE kidney toxicity, carcinogenicity, and mode-of-action 455
References 457
4.4 Liver toxicity and cancer 475
4.4.1 Liver non-cancer toxicity in humans 475
4.4.2 Liver cancer in humans 478
4.4.3 Experimental studies of TCE in rodents - introduction 494
4.4.4 TCE-induced liver non-cancer effects 496
4.4.4.1 Liver weight 497
4.4.4.2 Cytotoxicity 501
4.4.4.3 Measures of DNA synthesis, cellular proliferation, and apoptosis 508
4.4.4.4 Peroxisomal proliferation and related effects 510
4.4.4.5 Oxidative stress 513
4.4.4.6 Bile production 514
4.4.4.7 Summary: TCE-induced non-cancer effects in laboratory animals 515
4.4.5 TCE-induced liver cancer in laboratory animals 516
4.4.5.1 Negative or inconclusive studies of mice and rats 516
4.4.5.2 Positive TCE studies of mice 518
4.4.5.3 Summary: TCE-induced cancer in laboratory animals 521
4.4.6 Role of metabolism in liver toxicity and cancer 521
4.4.6.1 Pharmacokinetics of CH, TCA, and DCA from TCE exposure 521
4.4.6.2 Comparisons between TCE and TCA, DCA, and CH non-cancer effects 522
4.4.6.2.1 Hepatomegaly - qualitative and quantitative comparisons 522
4.4.6.2.2 Cytotoxicity 531
4.4.6.2.3 DNA synthesis and polyploidization 532
4.4.6.2.4 Apoptosis 534
4.4.6.2.5 Glycogen accumulation 535
4.4.6.2.6 Peroxisome proliferation and related effects 537
4.4.6.2.7 Oxidative stress 538
6/22/2009
Xll
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.4.6.3 Comparisons of TCE-induced carcinogenic responses with TCA, DC A, and CH
studies 540
4.4.6.3.1 Studies in Rats 540
4.4.6.3.2 Studies in Mice 542
4.4.6.3.2.1 TCE carcinogenicity dose-response data 542
4.4.6.3.2.2 DCA carcinogenicity dose-response data 545
4.4.6.3.2.3 TCA carcinogenicity dose-response data 547
4.4.6.3.2.4 CH carcinogenic dose-response 551
4.4.6.3.2.5 Degree of concordance among TCE, TCA, DCA, and CH dose-
response relationships 554
4.4.6.3.3 Inferences from liver tumor phenotype and genotype 555
4.4.6.3.3.1 Tumor phenotype - staining and appearance 555
4.4.6.3.3.2 C-Jun staining 558
4.4.6.3.3.3 Tumor genotype: H-ras mutation frequency and spectrum 560
4.4.6.3.4 "Stop" experiments 563
4.4.6.4 Conclusions regarding the role of TCA, DCA, and CH in TCE-induced effects
in the liver 564
4.4.7 MOA for TCE Liver Carcinogenicity 565
4.4.7.1 Mutagenicity 565
4.4.7.2 PPARa receptor activation 567
4.4.7.3 Additional Proposed Hypotheses and Key Events with Limited Evidence or
Inadequate Experimental Support 572
4.4.7.3.1 Increased liver weight 572
4.4.7.3.2 "Negative selection" 573
4.4.7.3.3 Polyploidization 574
4.4.7.3.4 Glycogen storage 574
4.4.7.3.5 Inactivation of GST-Zeta 575
4.4.7.3.6 Oxidative stress 576
4.4.7.3.7 Changes in gene expression (e.g., hypomethylation) 577
4.4.7.3.8 Cytotoxicity 581
4.4.7.4 MOA Conclusions 581
4.4.7.4.1 Qualitative human relevance and susceptibility 582
4.4.7.4.2 Quantitative species differences 583
References 587
4.5 Immunotoxicity and Cancers of the Immune System 610
4.5.1 Human Studies 610
6/22/2009
Xlll
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.5.1.1 Noncancer Immune-Related Effects 610
4.5.1.1.1 Immunosuppression, Asthma, and Allergies 610
4.5.1.1.2 Generalized hypersensitivity skin diseases, with or without hepatitis 611
4.5.1.1.3 Cytokine profiles 614
4.5.1.1.4 Autoimmune disease 619
4.5.1.1.4.1 Disease Clusters and Geographic-Based Studies 619
4.5.1.1.4.2 Case-Control Studies 620
4.5.1.2 Cancers of the Immune System, Including Childhood Leukemia 625
4.5.1.2.1 Description of Studies 625
4.5.1.2.2 Meta-analysis 650
4.5.2 Animal Studies 655
4.5.2.1 Immunosuppression 655
4.5.2.1.1 Inhalation exposures 655
4.5.2.1.2 Oral exposures 657
4.5.2.1.3 Intraperitoneal administration 659
4.5.2.2 Hypersensitivity 665
4.5.2.3 Autoimmunity 669
4.5.2.4 Cancers of the immune system 683
4.5.3 Summary 687
4.5.3.1 Noncancer Effects 687
4.5.3.2 Cancer 688
References 691
4.6 Respiratory tract toxicity and cancer 705
4.6.1 Epidemiologic Evidence 705
4.6.1.1 Chronic Effects: Inhalation 705
4.6.1.2 Cancer 705
4.6.2 Laboratory Animal Studies 718
4.6.2.1 Respiratory Tract Animal Toxicity 718
4.6.2.1.1 Acute and Short-term effects: Inhalation 718
4.6.2.1.1.1 Acute and Short-term effects: Intraperitoneal Injection and Gavage
Exposure 721
4.6.2.1.1.2 Subchronic and Chronic Effects 722
4.6.2.2 Respiratory Tract Cancer 723
4.6.2.2.1 Inhalation 723
4.6.2.2.2 Gavage 723
4.6.3 Role of Metabolism in Pulmonary Toxicity 724
6/22/2009
xiv
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
729
730
730
731
731
733
733
733
735
744
754
754
754
754
754
755
758
761
761
766
776
776
777
779
780
781
782
782
782
787
792
796
798
799
799
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.6.4 Mode of Action for Pulmonary Carcinogenicity
4.6.4.1 Mutagenicity via Oxidative Metabolism
4.6.4.1.1 Experimental Support for the Hypothesized Mode of Action
4.6.4.2 Cytotoxicity leading to increased cell proliferation
4.6.4.2.1 Experimental Support for the Hypothesized Mode of Action
4.6.4.3 Additional Hypothesized Modes of Action with Limited Evidence or
Inadequate Experimental Support
4.6.4.3.1 Role of Formation of DAL Protein Adducts
4.6.4.4 Conclusions about the hypothesized modes of action
4.6.5 Summary and Conclusions
References
4.7 Reproductive and developmental toxicity
4.7.1 Reproductive toxi city
4.7.1.1 Human reproductive outcome data
4.7.1.1.1 Female and male combined human reproductive effects
4.7.1.1.2 Female human reproductive effects
4.7.1.1.3 Male human reproductive effects
4.7.1.1.4 Summary of human reproductive toxicity
4.7.1.2 Animal reproductive toxicity studies
4.7.1.2.1 Inhalation exposures
4.7.1.2.2 Oral exposures
4.7.1.3 Discussion/synthesis of non-cancer reproductive toxicity findings
4.7.1.3.1 Female reproductive toxicity
4.7.1.3.2 Male reproductive toxi city
4.7.1.3.2.1 The role of metabolism in male reproductive toxicity
4.7.1.3.2.2 Mode of acti on for mal e reproductive toxi city
4.7.1.3.3 Summary of non-cancer reproductive toxicity
4.7.2 Cancers of the reproductive system
4.7.2.1 Human data
4.7.2.1.1 Prostate Cancer
4.7.2.1.2 Breast Cancer
4.7.2.1.3 Cervical Cancer
4.7.2.2 Animal studies
4.7.2.3 Mode of action for testicular tumors
4.7.3 Developmental toxicity
4.7.3.1 Human developmental data
6/22/2009
xv
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
800
809
814
827
827
827
830
848
849
849
850
851
851
852
852
854
857
860
861
862
864
866
868
870
891
891
901
908
913
918
918
918
918
920
923
923
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.7.3.1.1 Adverse fetal/birth outcomes
4.7.3.1.2 Postnatal Developmental Outcomes
4.7.3.1.3 Summary of human developmental toxicity
4.7.3.2 Animal developmental toxicology studies
4.7.3.2.1 Mammalian studies
4.7.3.2.1.1 Inhalation exposures
4.7.3.2.1.2 Oral exposures
4.7.3.2.1.3 Intraperitoneal exposures
4.7.3.2.2 Studies in non-mammalian species
4.7.3.2.2.1 Avian
4.7.3.2.2.2 Amphibian
4.7.3.2.2.3 Invertebrate
4.7.3.2.3 In vitro studies
4.7.3.3 Discussion/synthesis of developmental data
4.7.3.3.1 Adverse fetal and early neonatal outcomes
4.7.3.3.2 Cardiac malformations
4.7.3.3.2.1 Mode of action for cardiac malformations
4.7.3.3.2.2 Association of PPAR with developmental outcomes
4.7.3.3.2.3 Summary of the weight of evidence on cardiac malformations.
4.7.3.3.3 Other structural developmental outcomes
4.7.3.3.4 Developmental neurotoxicity
4.7.3.3.5 Developmental immunotoxicity
4.7.3.3.6 Childhood Cancers
References
4.8 Other site-specific cancers
4.8.1 Esophageal cancer
4.8.2 Bladder Cancer
4.8.3 Central Nervous System and Brain Cancers
References
4.9 Susceptible Lifestages and Populations
4.9.1 Lifestages
4.9.1.1 Early Lifestages
4.9.1.1.1 Early Lifestage-Specific Exposures
4.9.1.1.2 Early Lifestage-Specific Toxicokinetics
4.9.1.1.3 Early Lifestage-Specific Effects
4.9.1.1.3.1 Differential effects in early lifestages
6/22/2009 xvi
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
924
925
927
928
928
928
930
930
932
933
933
934
934
935
935
935
936
936
936
936
937
938
938
939
939
941
965
965
965
969
970
972
973
974
975
979
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.9.1.1.3.2 Susceptibility to noncancer outcomes in early lifestages.
4.9.1.1.3.3 Susceptibility to cancer outcomes in early lifestages
4.9.1.2 Later Lifestages
4.9.2 Other Susceptibility Factors
4.9.2.1 Gender
4.9.2.1.1 Gender-Specific Toxicokinetics
4.9.2.1.2 Gender -Specific Effects
4.9.2.1.2.1 Gender susceptibility to non-cancer outcomes
4.9.2.1.2.2 Gender susceptibility to cancer outcomes
4.9.2.2 Genetic Variability
4.9.2.2.1 CYP450 Genotypes
4.9.2.2.2 GST Genotype
4.9.2.2.3 Other Genotypes
4.9.2.3 Race/Ethnicity
4.9.2.4 Pre-Existing Health Status
4.9.2.4.1 Obesity and Metabolic Syndrome
4.9.2.4.2 Diabetes
4.9.2.4.3 Hypertension
4.9.2.5 Lifestyle Factors and Nutrition Status
4.9.2.5.1 Alcohol Intake
4.9.2.5.2 Tobacco Smoking
4.9.2.5.3 Nutritional Status
4.9.2.5.4 Physical Activity
4.9.2.5.5 Socioeconomic Status
4.9.3 Uncertainty of Database for Susceptible Populations
References
4.10 Hazard Characterization
4.10.1 Characterization of Non-Cancer Effects
4.10.1.1 Neurotoxicity
4.10.1.2 Kidney toxicity
4.10.1.3 Liver toxicity
4.10.1.4 Immunotoxi city
4.10.1.5 Respiratory tract toxicity
4.10.1.6 Reproductive toxi city
4.10.1.7 Developmental toxicity
4.10.2 Characterization of Carcinogenicity
6/22/2009
xvii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.10.2.1 Summary evaluation of epidemiologic evidence of TCE and cancer 980
4.10.2.2 Summary of evidence for TCE carcinogenicity in rodents 987
4.10.2.3 Summary of additional evidence on biological plausibility 989
4.10.2.3.1 Toxicokinetics 989
4.10.2.3.2 Toxicity and mode of action 990
4.10.3 Characterization of Factors Impacting Susceptibility 994
References 997
5 Dose-Response Assessment 1029
5.1 Dose-Response Analyses for Non-Cancer Endpoints 1029
5.1.1 Modeling approaches and uncertainty factors for developing candidate reference
values based on applied dose 1031
5.1.2 Candidate Critical Effects by Effect Domain 1035
5.1.2.1 Candidate critical neurological effects on the basis of applied dose 1036
5.1.2.2 Candidate critical kidney effects on the basis of applied dose 1040
5.1.2.3 Candidate critical liver effects on the basis of applied dose 1041
5.1.2.4 Candidate critical body weight effects on the basis of applied dose 1042
5.1.2.5 Candidate critical immunological effects on the basis of applied dose 1044
5.1.2.6 Candidate critical respiratory tract effects on the basis of applied dose 1047
5.1.2.7 Candidate critical reproductive effects on the basis of applied dose 1047
5.1.2.7.1 Male reproductive effects (effects on sperm and reproductive tract) 1047
5.1.2.7.2 Other reproductive effects 1049
5.1.2.8 Candidate critical developmental effects on the basis of applied dose 1053
5.1.2.9 Summary of cRfCs, cRfDs, and Candidate Critical Effects 1058
5.1.3 Application of PBPK model to inter- and intra-species extrapolation for candidate
critical effects 1061
5.1.3.1 Selection of dose metrics for different endpoints 1061
5.1.3.1.1 Kidney toxicity (meganucleuocytosis, increased kidney weight, toxic
nephropathy) 1064
5.1.3.1.2 Liver weight increases (hepatomegaly) 1067
5.1.3.1.3 Developmental toxicity - heart malformations 1069
5.1.3.1.4 Reproductive toxicity - decreased ability of sperm to fertilize oocytes.. 1070
5.1.3.1.5 Other reproductive and developmental effects and neurological effects and
immunologic effects 1071
5.1.3.2 Methods for inter- and intra-species extrapolation using internal doses 1072
5.1.3.3 Results and discussion of p-RfCs and p-RfDs for candidate critical effects.1076
5.1.4 Uncertainties in cRfCs and cRfDs 1088
6/22/2009
xviii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
5.1.4.1 Qualitative uncertainties 1088
5.1.4.2 Quantitative uncertainty analysis of PBPK model-based dose metrics for
LOAEL or NOAEL-based PODs 1090
5.1.5 Summary of non-cancer reference values 1100
5.1.5.1 Preferred candidate reference values (cRfCs, cRfD, p-cRfCs and p-cRfDs) for
candidate critical effects 1100
5.1.5.2 Reference Concentration 1104
5.1.5.3 Reference Dose 1107
5.2 Dose-Response Analysis for Cancer Endpoints 1110
5.2.1 Dose-Response Analyses: Rodent Bioassays 1110
5.2.1.1 Rodent dose-response analyses: Studies and Modeling Approaches 1110
5.2.1.2 Rodent Dose-Response Analyses: Dosimetry 1115
5.2.1.2.1 Selection of dose metrics for different tumor types 1115
5.2.1.2.1.1 Kidney 1118
5.2.1.2.1.2 Liver 1120
5.2.1.2.1.3 Lung 1122
5.2.1.2.1.4 Other sites 1124
5.2.1.2.2 Methods for dose-response analyses using internal dose metrics 1125
5.2.1.3 Rodent dose-response analyses: Results 1131
5.2.1.4 Uncertainties in dose-response analyses of rodent bioassays 1143
5.2.1.4.1 Qualitative discussion of uncertainties 1143
5.2.1.4.2 Quantitative uncertainty analysis of PBPK model-based dose metrics... 1147
5.2.2 Dose-Response Analyses: Human Epidemiologic Data 1153
5.2.2.1 Inhalation Unit Risk Estimate for Renal Cell Carcinoma Derived from
Charbotel et al. (2006) Data 1153
5.2.2.1.1 RCC results from the Charbotel et al. study 1153
5.2.2.1.2 Prediction of lifetime extra risk of RCC incidence from TCE exposure. 1154
5.2.2.1.3 Uncertainties in the RCC unit risk estimate 1157
5.2.2.1.4 Conclusions regarding the RCC unit risk estimate 1160
5.2.2.2 Adjustment of the Inhalation Unit Risk Estimate for Multiple Sites 1160
5.2.2.3 Route-to-route extrapolation using PBPK model 1163
5.2.3 Summary of unit risk estimates 1166
5.2.3.1 Inhalation unit risk estimate 1166
5.2.3.2 Oral unit risk estimate 1167
5.2.3.3 Application of age-dependent adjustment factors 1169
5.2.3.3.1 Example application of ADAFs for inhalation exposures 1170
6/22/2009
xix
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
5.2.3.3.2 Example application of ADAFs for oral exposures 1171
REFERENCES 1175
6 MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE 1185
6.1 Human Hazard Potential 1185
6.1.1 Exposure (Chapter 2) 1185
6.1.2 Toxicokinetics and PBPK modeling (Chapter 3 and Appendix A) 1186
6.1.3 Non-cancer toxicity 1187
6.1.3.1 Neurological effects (Sections 4.2 and 4.10.1.1 and Appendix D) 1188
6.1.3.2 Kidney effects (Sections 4.3.1, 4.3.4, 4.3.6, and 4.10.1.2) 1189
6.1.3.3 Liver effects (Sections 4.4.1, 4.4.3, 4.4.4, 4.4.6,and 4.10.1.3, and Appendix E)1190
6.1.3.4 Immunological effects (Sections 4.5.1.1, 4.5.2, and 4.10.1.4) 1191
6.1.3.5 Respiratory tract effects (Sections 4.6.1.1, 4.6.2.1, 4.6.3, and 4.10.1.5) 1191
6.1.3.6 Reproductive effects (Sections 4.7.1 and 4.10.1.6) 1192
6.1.3.7 Developmental effects (Sections 4.7.3 and 4.10.1.7) 1193
6.1.4 Carcinogenicity (Sections 4.0, 4.1, 4.3.2, 4.3.5, 4.3.7, 4.4.2, 4.4.5, 4.4.6, 4.4.7,
4.5.1.2, 4.5.2.4, 4.6.1.2, 4.6.2.2, 4.6.4, 4.7.2, 4.8, and 4.10.2, and Appendices B and C) .1195
6.1.5 Susceptibility (Sections 4.9 and 4.10.3) 1200
6.2 Dose-Response Assessment 1202
6.2.1 Non-cancer effects (Section 5.1) 1202
6.2.1.1 Background and methods 1202
6.2.1.2 Uncertainties and application of uncertainty factors (UFs) (Section 5.1.1 and
5.1.4) 1203
6.2.1.3 Candidate critical effects and reference values (Sections 5.1.2 and 5.1.3)....1206
6.2.1.3.1 Neurological effects 1207
6.2.1.3.2 Kidney effects 1207
6.2.1.3.3 Liver effects 1208
6.2.1.3.4 Immunological effects 1208
6.2.1.3.5 Reproductive effects 1209
6.2.1.3.6 Developmental effects 1210
6.2.1.3.7 Summary of most sensitive candidate reference values 1211
6.2.1.4 Non-cancer reference values (Section 5.1.5) 1212
6.2.1.4.1 Reference Concentration 1212
6.2.1.4.2 Reference Dose 1213
6.2.2 Cancer (Section 5.2) 1215
6.2.2.1 Background and methods (Rodent: Section 5.2.1.1; Human: 5.2.2.1) 1215
6/22/2009
xx
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 6.2.2.2 Inhalation Unit Risk Estimate (Rodent: Section 5.2.1.3; Human: Section 5.2.2.1
2 and 5.2.2.2) 1217
3 6.2.2.3 Oral Unit Risk Estimate (Rodent: Section 5.2.1.3; Human: Section 5.2.2.3)1219
4 6.2.2.4 Uncertainties in cancer dose-response assessment 1220
5 6.2.2.4.1 Uncertainties in estimates based on human epidemiologic data (Section
6 5.2.2.1.3) 1220
7 6.2.2.4.2 Uncertainties in estimates based on rodent bioassays (Section 5.2.1.4)..1223
8 6.2.2.5 Application of age-dependent adjustment factors (Section 5.2.3.3) 1224
9 6.3 Overall Characterization of TCE Hazard and Dose-response 1226
10
6/22/2009
xxi
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
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
trichloroethylene. It is not intended to be a comprehensive treatise on the chemical or
toxicological nature of trichloroethylene.
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, where applicable, and to characterize the overall
confidence in the quantitative and qualitative aspects of hazard and dose response.
For other general information about this assessment or other questions relating to IRIS,
the reader is referred to EPA's IRIS Hotline at (202) 566-1676 (phone), (202) 566-1749 (fax), or
hotline.iris@epa.gov (email address).
6/22/2009
XXll
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER
Weihsueh A. Chiu
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
AUTHORS
Ambuja Bale
National Center for Environmental Assessment - Immediate Office
U.S. Environmental Protection Agency
Washington, DC
Stanley Bar one
National Center for Environmental Assessment - Immediate Office
U.S. Environmental Protection Agency
Washington, DC
Rebecca Brown
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
Jane C. Caldwell
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
Chao Chen
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
6/22/2009
XXlll
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Weihsueh A. Chiu
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
Glinda Cooper
National Center for Environmental Assessment - Immediate Office
U.S. Environmental Protection Agency
Washington, DC
Ghazi Dannan
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
Marina Evans
National Health and Environmental Effects Research Laboratory
(on detail to National Center for Environmental Assessment - Washington Office)
U.S. Environmental Protection Agency
Research Triangle Park, NC
John Fox
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
Kathryn Z. Guyton
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
Maureen R. Gwinn
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
6/22/2009
xxiv
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Jennifer Jinot
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
Nagalakshmi Keshava
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
John Lipscomb
National Center for Environmental Assessment - Cincinnati Office
U.S. Environmental Protection Agency
Cincinnati, OH
Susan Makris
National Center for Environmental Assessment - Washington Office
U.S. Environmental Protection Agency
Washington, DC
Miles Okino
National Exposure Research Laboratory - Las Vegas Office
U.S. Environmental Protection Agency
Las Vegas, NV
Fred Power
National Exposure Research Laboratory - Las Vegas Office
U.S. Environmental Protection Agency
Las Vegas, NV
John Schaum
National Center for Environmental Assessment - Washington Office
Office of Research and Development
Washington, DC
6/22/2009
XXV
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Cheryl Siegel Scott
National Center for Environmental Assessment - Washington Office
Office of Research and Development
Washington, DC
REVIEWERS
This document has been reviewed by U.S. EPA scientists, reviewers from other Federal
agencies, and the public, and peer reviewed by independent scientists external to U.S. EPA. A
summary and U.S. EPA's disposition of the comments received from the independent external
peer reviewers and from the public is included in Appendix I.
INTERNAL EPA REVIEWERS
Daniel Axelrad
National Center for Environmental Economics
Robert Benson
U.S. EPA Region 8
Ted Birner
National Center for Environmental Assessment - Immediate Office
Nancy Chiu
Office of Water
David Farrar
National Center for Environmental Assessment - Cincinnati Office
Lynn Flowers
National Center for Environmental Assessment - Immediate Office
Stiven Foster
Office of Solid Waste and Emergency Response
Susan Griffin
6/22/2009
xxvi
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
U.S. EPA Region 8
Samantha Jones
National Center for Environmental Assessment - Immediate Office
Leonid Kopylev
National Center for Environmental Assessment - Washington Office
Allan Marcus
National Center for Environmental Assessment - Immediate Office
Gregory Miller
Office of Children's Health Protection and Environmental Education
Deirdre Murphy
Office of Air Quality Planning and Standards
Marian 01 sen
U.S. EPA Region 2
Peter Preuss
National Center for Environmental Assessment - Immediate Office
Kathleen Raffaele
National Center for Environmental Assessment - Washington Office
William Sette
Office of Solid Waste and Emergency Response
Bob Sonawane
National Center for Environmental Assessment - Washington Office
Suryanarayana Vulimiri
National Center for Environmental Assessment - Washington Office
Nina Ching Y. Wang
6/22/2009
xxvii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
National Center for Environmental Assessment - Cincinnati Office
Paul White
National Center for Environmental Assessment - Washington Office
Jean Zodrow
U.S. EPA Region 10
EXTERNAL PEER REVIEWERS
Name
Affiliation
[note: organization affiliation only, spelled out; not complete mailing address]
ACKNOWLEDGMENTS
Drafts of Section 3.3 (TCE metabolism) prepared for the U.S. EPA by Syracuse Research
Corporation under contract. Additional support, including literature searches and retrievals and
drafts of Appendix D were prepared for the U.S. EPA by the Oak Ridge Institute for Science and
Education (ORISE) through interagency agreement number DW-89939822-01-0 with the U.S.
Department of Energy (DOE). ORISE is managed by Oak Ridge Associated Universities under
a contract with DOE. The PBPK modeling sections of this report are dedicated to the memory of
Fred Power (1938-2007). His keen analytical mind will be greatly missed, but his gentle heart
and big smile will be missed even more.
6/22/2009
xxviii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
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
trichloroethylene. IRIS Summaries may include oral reference dose (RfD) and inhalation
reference concentration (RfC) values for chronic and other exposure durations, and a
carcinogenicity assessment.
The RfD and RfC, if derived, provide quantitative information for use in risk assessments
for health effects known or assumed to be produced through a nonlinear (presumed threshold)
mode of action (MO A). 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 ppm or (J,g/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 are generally derived for chronic exposures (up to a lifetime), but may also be derived for
acute (<24 hours), short-term (>24 hours up to 30 days), and subchronic (>30 days up to 10% of
lifetime) exposure durations, all of which are derived based on an assumption of continuous
exposure throughout the duration specified. Unless specified otherwise, the RfD and RfC are
derived for chronic exposure duration.
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 may be derived. 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 may be derived from the application of a
low-dose extrapolation procedure. If derived, the oral slope factor is a plausible upper bound on
the estimate of risk per mg/kg/day of oral exposure. Similarly, an inhalation unit risk is a
plausible upper bound on the estimate of risk per ppm or |ig/m3 in air breathed.
Development of these hazard identification and dose-response assessments for
trichloroethylene 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 the Health Risk Assessment of Chemical Mixtures (U.S. EPA, 1986a),
6/22/2009
1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Guidelines for Mutagenicity Risk Assessment (U.S. EPA, 1986b), Recommendations for and
Documentation of Biological Values for Use in Risk Assessment (U.S. EPA, 1988), Guidelines
for Developmental Toxicity Risk Assessment (U.S. EPA, 1991), 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), Guidelines for Reproductive Toxicity Risk Assessment (U.S. EPA, 1996), Guidelines for
Neurotoxicity Risk Assessment (U.S. EPA, 1998), Science Policy Council Handbook. Risk
Characterization (U.S. EPA, 2000a), Benchmark Dose Technical Guidance Document (U.S.
EPA, 2000b), Supplementary Guidance for Conducting Health Risk Assessment of Chemical
Mixtures (U.S. EPA, 2000c), A Review of the Reference Dose and Reference Concentration
Processes (U.S. EPA, 2002), Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens
(U.S. EPA, 2005b), Science Policy Council Handbook: Peer Review (U.S. EPA, 2006a), and A
Framework for Assessing Health Risks of Environmental Exposures to Children (U. S. EPA,
2006b).
The literature search strategy employed for this compound was based on the Chemical
Abstracts Service Registry Number (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, 2009.
6/22/2009
2
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
REFERENCES
NRC (National Research Council). (1983) Risk assessment in the federal government: managing
the process. Washington, DC: National Academy Press.
U.S. EPA (Environmental Protection Agency). (1986a) Guidelines for the health risk assessment
of chemical mixtures. Federal Register 51(185):34014-34025. Available from:
.
U.S. EPA. (1986b) Guidelines for mutagenicity risk assessment. Federal Register
51(185):34006-34012. Available from: .
U.S. EPA. (1988) Recommendations for and documentation of biological values for use in risk
assessment. Prepared by the Environmental Criteria and Assessment Office, Office of Health and
Environmental Assessment, Cincinnati, OH for the Office of Solid Waste and Emergency
Response, Washington, DC; EPA 600/6-87/008. Available from:
.
U.S. EPA. (1991) Guidelines for developmental toxicity risk assessment. Federal Register
56(234):63798-63826. Available from: .
U.S. EPA. (1994a) Interim policy for particle size and limit concentration issues in inhalation
toxicity studies. Federal Register 59(206):53799. Available from:
.
U.S. EPA. (1994b) Methods for derivation of inhalation reference concentrations and application
of inhalation dosimetry. Office of Research and Development, Washington, DC; EPA/600/8-
90/066F. Available from: .
U.S. EPA. (1995) Use of the benchmark dose approach in health risk assessment. Risk
Assessment Forum, Washington, DC; EPA/630/R-94/007. Available from:
.
U.S. EPA. (1996) Guidelines for reproductive toxicity risk assessment. Federal Register
61(212):56274-56322. Available from: .
6/22/2009
3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
U.S. EPA. (1998) Guidelines for neurotoxicity risk assessment. Federal Register
63(93):26926-26954. Available from: .
U.S. EPA. (2000a) Science policy council handbook: risk characterization. Office of Science
Policy, Office of Research and Development, Washington, DC; EPA 100-B-00-002. Available
from: .
U.S. EPA. (2000b) Benchmark dose technical guidance document [external review draft]. Risk
Assessment Forum, Washington, DC; EPA/63O/R-00/001. Available from:
.
U.S. EPA. (2000c) Supplementary guidance for conducting for health risk assessment of
chemical mixtures. Risk Assessment Forum, Washington, DC; 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. Risk Assessment Forum,
Washington, DC; EPA/630/P-03/001B. Available from: .
U.S. EPA. (2005b) Supplemental guidance for assessing susceptibility from early-life exposure
to carcinogens. Risk Assessment Forum, Washington, DC; EPA/630/R-03/003F. Available from:
.
U.S. EPA. (2006a) Science policy council handbook: peer review. Third edition. Office of
Science Policy, Office of Research and Development, Washington, DC; EPA/100/B-06/002.
Available from: .
U.S. EPA. (2006b) A Framework for Assessing Health Risk of Environmental Exposures to
Children. National Center for Environmental Assessment, Washington, DC, EPA/600/R-
05/093F. Available from: .
6/22/2009
4
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
2 EXPOSURE CHARACTERIZATION
The purpose of this exposure characterization is to summarize information about TCE
sources, releases, media levels and exposure pathways for the general population (occupational
exposure is also discussed to a lesser extent). It is not meant as a substitute for a detailed
exposure assessment for a particular risk assessment application. While this section primarily
addresses TCE, it also includes some information on a number of related compounds. These
related compounds include metabolites of TCE and other parent compounds that produce similar
metabolites as shown in Table 2-1. The first column in this table lists the principal TCE
metabolites in humans (trichloroethanol, trichloroethanol-glucuronide and trichloroacetic acid)
as well as a number of minor ones (ATSDR, 1997a). The subsequent columns list parent
compounds that can produce some of the same metabolites. The metabolic reaction pathways
are much more complicated than implied here and it should be understood that this table is
intended only to provide a general understanding of which parent compounds lead to which TCE
metabolites. Exposure to the TCE-related compounds can alter or enhance TCE's metabolism
and toxicity by generating higher internal metabolite concentrations than would result from TCE
exposure by itself. This characterization is based largely on earlier work by Wu and Schaum
(2000, 2001), but also provides updates in a number of areas.
Table 2-1. TCE Metabolites and Related Parent Compounds*
Parent Compounds
TCE Metabolites Tclrachloro- 1.1-Dichloro- 1.1.1 -Tri- 1.1.1,2-Tctra- 1.2-Dichloro-
clhvlene ethane chloroelhane chloroclhanc ethylene
Oxalic Acid X X
Chloral X
Chloral Hydrate X
Monochloroacetic Acid X X X X X
Dichloroacetic Acid XX X
Trichloroacetic Acid X XX
Trichloroethanol X XX
Trichloroethanol- X XX
glucuronide
* X indicates that the parent compound can produce the corresponding metabolite (Hazardous
Substances Data Bank, http://toxnet.nlm.nih.gov./cgi-bin/sis/htmlgen?HSDB)
6/22/2009
5
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 2.1 ENVIRONMENTAL SOURCES
2 TCE is a stable, colorless liquid with a chloroform-like odor and chemical formula
3 C2C13H (Lewis, 2001). Its chemical properties are listed in Table 2-2.
4
5
6
7
8 Table 2-2. Chemical Properties of TCE
Property
Value
Reference
Molecular Weight
131.39
Lide, 1998
Boiling Point
87.2° C
Lide, 1998
Melting Point
-84.7° C
Lide, 1998
Density
1.4642 at 20° C
Merck Index, 1996
Solubility
1,280 mg/L water at 25° C
Hotvath et al., 1999
Vapor Pressure
69.8 mmHG @ 25°C
Boublik et al., 1984
Vapor Density
4.53 (air = 1)
Merck Index, 1996
Henry's Law Constant
9.85 x 10"3 atm-cu m/mol @
25° C
Leighton, 1981
Octanol/W ater Partition
Coefficient
log Kow = 2.61
Hansch, 1995
Air Concentration
Conversion
1 ppb = 5.38 (J,g/m3
HSDB, 2002
Mo ecu ar Structure of TCE
Figure 2.1
6/22/2009
6
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Trichloroethylene has been produced commercially since the 1920s in many countries by
chlorination of ethylene or acetylene. Its use in vapor degreasing began in the 1920s. In the
1930s, it was introduced for use in dry cleaning. This use was largely discontinued in the 1950s
and was replaced with tetrachloroethylene (ATSDR, 1997a). More recently, 80-90% of
trichloroethylene production worldwide is used for degreasing metals (IARC, 1995). It is also
used in adhesives, paint-stripping formulations, paints, lacquers, and varnishes (SRI, 1992). A
number of past uses in cosmetics, drugs, foods and pesticides have now been discontinued
including use as an extractant for spice oleoresins, natural fats and oils, hops and decaffeination
of coffee (IARC, 1995), and as a carrier solvent for the active ingredients of insecticides and
fungicides, and for spotting fluids (WHO, 1985; ATSDR, 1997a). The production of TCE in the
United States peaked in 1970 at 280 million kg (616 million pounds) and declined to 60 million
kg (132 million pounds) in 1998 (USGS, 2006). In 1996, the United States imported 4.5 million
kg (10 million pounds) and exported 29.5 million kg (65 million pounds) (Chemical Marketing
Reporter, 1997). Table 2-3 summarizes the basic properties and principal uses of the TCE
related compounds.
Releases of TCE from nonanthropogenic activities are negligible (HSDB, 2002). Most of
the TCE used in the United States is released to the atmosphere, primarily from vapor degreasing
operations (ATSDR, 1997a). Releases to air also occur at treatment and disposal facilities, water
treatment facilities, and landfills (ATSDR, 1997a). TCE has also been detected in stack
emissions from municipal and hazardous waste incineration (ATSDR, 1997a). TCE is on the list
for reporting to EPA's Toxics Release Inventory (TRI). Reported releases into air predominate
over other types and have declined over the period 1994 to 2004 (see Table 2-4).
6/22/2009
7
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 2-3. Properties and Uses of TCE Related Compounds
\\;i(er
Solul)ili(\
(niii/l.)
Ysipor Pivssiiiv
(mm ll(>)
Uses
Sourees
Tetrachloroethylene
150
18.5 @25°C
Dry cleaning, degreasing, solvent
1
1,1,1 -T richloroethane
4400
124 @25°C
Solvents, degreasing
1
1,2-Dichloroethylene
3000-6000
273-395 @30°C
Solvents, chemical intermediates
1
1,1,1,2-
Tetrachloroethane
1100
14 @25°C
Solvents, but currently not
produced in United States
1,2
1,1 -Dichloroethane
5500
234 @25°C
Solvents, chemical intermediates
1
Chloral
High
35 @20°C
Herbicide production
1
Chloral Hydrate
High
NA
pharmaceutical production
1
Monochloroacetic Acid
High
1 @43°C
pharmaceutical production
1
Dichloroacetic Acid
High
<1 @20°C
pharmaceuticals, not widely used
1
Trichloroacetic Acid
High
1 @50°C
herbicide production
1
Oxalic Acid
220,000
0.54 @105°C
Scouring/cleaning agent,
degreasing
2
Dichlorovinyl cysteine
Not Available
Not Available
Not Available
T richloroethanol
Low
NA
Anesthetics and chemical
intermediate
3
2 1 - Wu and Schaum, 2001
3 2 - HSDB, 2003
4 3-Lewis, 2001
5
6/22/2009
8
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 2-4. TRI Releases of TCE (pounds/year)
Total On-
Year
On-Sitc
On-Sitc
Total On-
On-Sitc
Total
Total
and Off-
Fugitive
Stack Air
Sitc Air
Su rt'acc
Total On-Sitc
On-Sitc
Off-Site
Site
Air
Emissions
Water
Underground
Releases
Disposal
Disposal
Discharges
Injection
to Land
or Other
or Other
Releases
Releases
1994
15,018,818
15,929,943
30,948,761
1,671
288
4,070
96,312
31,051,102
1995
12,498,086
13,784,853
26,282,939
1,477
550
3,577
74,145
26,362,688
1996
10,891,223
10,995,228
21,886,451
541
1,291
9,740
89,527
21,987,550
1997
9,276,150
8,947,909
18,224,059
568
986
3,975
182,423
18,412,011
1998
6,769,810
6,504,289
13,274,099
882
593
800
136,766
13,413,140
1999
5,861,635
4,784,057
10,645,692
1,034
0
148,867
192,385
10,987,978
2000
5,485,493
4,375,516
9,861,009
593
47,877
9,607
171,952
10,091,038
2001
4,968,282
3,453,451
8,421,733
406
98,220
12,609
133,531
8,666,499
2002
4,761,104
3,436,289
8,197,393
579
140,190
230
139,398
8,477,790
2003
3,963,054
3,121,718
7,084,772
595
90,971
150,642
66,894
7,393,873
2004
3,040,460
3,144,980
6,185,440
216
123,637
2
71,780
6,381,075
2005
2,733,983
2,893,168
5,627,152
533
86,817
4,711
60,074
5,779,287
2006
2,816,241
2,795,184
5,611,425
482
0
77,339
90,758
5,780,004
Source: EPA TRI Explorer, http://www.epa.gov/triexplorer/trends.htm
Under the National-Scale Air Toxics Assessment (NSATA) program, EPA has developed
an emissions inventory for TCE (U.S. EPA, 2007a). The inventory includes sources in the
United States plus the Commonwealth of Puerto Rico and the U.S. Virgin Islands. The types of
emission sources in the inventory include large facilities, such as waste incinerators and factories
and smaller sources, such as dry cleaners and small manufacturers. Figures 2-1 and 2-2 show the
results of the 1999 emissions inventory for TCE. Figure 2-1 shows the percent contribution to
total emissions by source category. A variety of sources have TCE emissions with the largest
ones identified as halogenated solvent cleaners and metal parts and products. Figure 2-2 shows a
national map of the emission density (tons/sq mi-yr) for TCE. This map shows the highest
densities in the far west and northeastern regions of the United States. Emissions range from 0 to
4.12 tons/mi2-yr.
6/22/2009
9
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Trichloroethylene Emissions
1999
2% Municipal Landfills
2% Pulp and Paper Production
2% Aerospace Industries 2% Printing. Coating & Dyeing Of Fabrics
2% Integrated Iron a. aieei lyianuraciuring
2% Consumer and Commercial Products Use
4% Dry Cleaning
6% Miscellaneous Metal Parts & Products (Surface Coating)
19% Other Categories (293 categories)
59% Halogenated Solvent Cleaners
Figure 2-1. Source contribution to TCE emissions
Bi erqarck
Sacrarnsr
Phqsnlx
Distribution of U.S. Emission Densities
Highest In U.S. ¦ , +.12
95 0.063
0.020
Percentile
90
75
50
£5
Lowest I n U.S.
0.004- 0
0.000 73
0.000 21
0.000 000 04-4-
Pollutant Emission Density by County
(tons / yeor / sq. mile )'
1999 NATA National-
Source' U.S. EPA / OAQPS
-Scale AJr Toxics Assessment
1999 County Emission Densities
Trichloroethylene — United States Counties
Figure 2-2. Annual emissions of TCE
6/22/2009
10
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
2.2 ENVIRONMENTAL FATE
Fate in Terrestrial Environments: The dominant fate of trichloroethylene released to
surface soils is volatilization. Because of its moderate water solubility, trichloroethylene
introduced into soil (e.g., landfills) also has the potential to migrate through the soil into
groundwater. The relatively frequent detection of trichloroethylene in groundwater confirms
this. Biodegradation in soil and groundwater may occur at a relatively slow rate (half-lives on
the order of months to years) (Howard et al., 1991).
Fate in the Atmosphere: In the atmosphere, trichloroethylene is expected to be present
primarily in the vapor phase, rather than sorbed to particulate, because of its high vapor pressure.
Some removal by scavenging during wet precipitation is expected because of its moderate water
solubility. The major degradation process affecting vapor phase trichloroethylene is photo-
oxidation by hydroxyl radicals. Photolysis in the atmosphere proceeds very slowly, if at all.
Trichloroethylene does not absorb UV light at wavelengths of less than 290 nm and thus will not
directly photolyze. Based on measured rate data for the vapor phase photo-oxidation reaction
with hydroxyl radicals, the estimated half-life of trichloroethylene in the atmosphere is on the
order of 1 to 11 days with production of phosgene, dichloroacetyl chloride, and formyl chloride.
Under smog conditions, degradation is more rapid (half-life on the order of hours) (HSDB, 2002;
Howard et al., 1991).
Fate in Aquatic Environments: The dominant fate of trichloroethylene released to
surface waters is volatilization (predicted half-life of minutes to hours). Bioconcentration,
biodegradation, and sorption to sediments and suspended solids are not thought to be significant
(HSDB, 2002). Trichloroethylene is not hydrolyzed under normal environmental conditions.
However, slow photo-oxidation in water (half-life of 10.7 months) has been reported (HSDB,
2002; Howard et al., 1991).
2.3 EXPOSURE CONCENTRATIONS
TCE levels in the various environmental media result from the releases and fate processes
discussed in Sections 2.1 and 2.2. No statistically based national sampling programs have been
conducted that would allow estimates of true national means for any environmental medium. A
substantial amount of air and groundwater data, however, has been collected as well as some
data in other media, as described below.
Outdoor Air - Measured Levels: TCE has been detected in the air throughout the
United States. According to ATSDR (1997a), atmospheric levels are highest in areas
6/22/2009
11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
concentrated with industry and population, and lower in remote and rural regions. Table 2-5
shows levels of TCE measured in the ambient air at a variety of locations in the United.
More recent ambient air measurement data for TCE were obtained from EPA's Air
Quality System database at the AirData Web site: http://www.epa.gov/air/data/index.html (U.S.
EPA, 2007b). These data were collected from a variety of sources including state and local
environmental agencies. The data are not from a statistically based survey and cannot be
assumed to provide nationally representative values. The most recent data (2006) come from 258
monitors located in 37 states. The means for these monitors range from 0.03 to 7.73 |ig/m3 and
have an overall average of 0.23 (j,g/m3. Table 2-6 summarizes the data for the years 1999-2006.
The data suggest that levels have remained fairly constant since 1999 at about 0.3 |ig/m3. Table
2-7 shows the monitoring data organized by land setting (rural, suburban, or urban) and land use
(agricultural, commercial, forest, industrial, mobile, and residential). Urban air levels are almost
4 times higher than rural areas. Among the land use categories, TCE levels are highest in
commercial/industrial areas and lowest in forest areas.
6/22/2009
12
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 2-5. Concentrations of Trichloroethylene in Ambient Air
Arosi
Yosir
( oiHTiilriilion (fiii/m )
Mosul
U;in<>o
Rural
Whiteface Mountain, NY (a)
Badger Pass, CA (a)
Reese River, NV (a)
Jetmar, KS (a)
All rural sites
1974
1977
1977
1978
1974-1978
0.5
0.06
0.06
0.07
<0.3-1.9
0.005-0.09
0.005-0.09
0.04-0.11
0.005 - 1.9
Urban and Suburban
New Jersey (a)
1973-79
9.1
ND-97
New York City, NY (a)
1974
3.8
0.6-5.9
Los Angeles, CA (a)
1976
1.7
0.14-9.5
Lake Charles, LA (a)
1976-78
8.6
0.4-11.3
Phoenix, AZ (a)
1979
2.6
0.06-16.7
Denver, CO (a)
1980
1.07
0.15-2.2
St. Louis, MO (a)
1980
0.6
0.1-1.3
Portland, OR (a)
1984
1.5
0.6-3.9
Philadelphia, PA (a)
1983-1984
1.9
1.6-2.1
Southeast Chicago, IL (b)
1986-1990
1.0
East St. Louis, IL (b)
1986-1990
2.1
District of Columbia (c)
1990-1991
1.94
1-16.65
Urban Chicago, IL (d)
pre-1993
0.82-1.16
Suburban Chicago, IL (d)
pre-1993
0.52
300 cities in 42 states (e)
pre-1986
2.65
Several Canadian Cities (f)
1990
0.28
Several US Cities (f)
1990
6.0
Phoenix, AZ (g)
1994-1996
0.29
0-1.53
Tucson, AZ (g)
1994-1996
0.23
0-1.47
All urban/suburban sites
1973-1996
0-97
2 (a) IARC, 1995 (b) Sweet, 1992 (c) Hendler, 1992 (d) Scheff, 1993 (e) Shah, 1988 (f) Bunce, 1994 (g)
3 Zielinska, 1998
6/22/2009
13
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 2-6. TCE Ambient Air Monitoring Data ([j,g/m3)
Year
Number of
Number of
Mean
Standard
Median
Range
Monitors
States
Deviation
1999
162
20
0.30
0.53
0.16
0.01-4.38
2000
187
28
0.34
0.75
0.16
0.01-7.39
2001
204
31
0.25
0.92
0.13
0.01-12.90
2002
259
41
0.37
1.26
0.13
0.01-18.44
2003
248
41
0.35
0.64
0.16
0.02-6.92
2004
256
37
0.32
0.75
0.13
0.00-5.78
2005
313
38
0.43
1.05
0.14
0.00-6.64
2006
258
37
0.23
0.55
0.13
0.03-7.73
2 Source: EPA's Air Quality System database at the AirData Web site: http://www. epa.gov/air/data/index.html
3
4 Table 2-7. Mean TCE Air Levels across Monitors by Land Setting and Use (1985 to 1998)
Ru ral
Subur-
ban
Urban
Agricul-
tural
Com-
mercial
Forest
Indus-
trial
Mobile
Resi-
dential
Mean
Concen-
tration
(lig/m3)
0.42
1.26
1.61
1.08
1.84
0.1
1.54
1.5
0.89
n
93
500
558
31
430
17
186
39
450
5 Source: EPA's Air Quality System database at the AirData Web site: http://www.epa.gov/air/data/index.html
6
7
8 Outdoor Air - Modeled Levels: Under the National-Scale Air Toxics Assessment
9 program, EPA has compiled emissions data and modeled air concentrations/exposures for the
10 Criteria Pollutants and Hazardous Air Pollutants (U.S. EPA, 2007a). The results of the 1999
11 emissions inventory for TCE were discussed earlier and results presented in Figures 2-1 and 2-2.
12 A computer simulation model known as the Assessment System for Population Exposure
13 Nationwide (ASPEN) is used to estimate toxic air pollutant concentrations (U.S. EPA, 2005).
14 This model is based on the EPA's Industrial Source Complex Long Term model (ISCLT) which
15 simulates the behavior of the pollutants after they are emitted into the atmosphere. ASPEN uses
16 estimates of toxic air pollutant emissions and meteorological data from National Weather Service
17 Stations to estimate air toxics concentrations nationwide. The ASPEN model takes into account
18 important determinants of pollutant concentrations, such as:
6/22/2009
14
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
• rate of release;
• location of release;
• the height from which the pollutants are released;
• wind speeds and directions from the meteorological stations nearest to the release;
• breakdown of the pollutants in the atmosphere after being released (i.e., reactive decay);
• settling of pollutants out of the atmosphere (i.e., deposition) and
• transformation of one pollutant into another (i.e., secondary formation).
The model estimates toxic air pollutant concentrations for every census tract in the continental
United States, the Commonwealth of Puerto Rico and the U.S. Virgin Islands. Census tracts are
land areas defined by the U.S. Bureau of the Census and typically contain about 4,000 residents
each. Census tracts are usually smaller than 2 square miles in size in cities but much larger in
rural areas.
Figure 2-3 shows the results of the 1999 ambient air concentration modeling for TCE.
The county median air levels range from 0 to 3.79 |ig/m3 and an overall median of 0.054 |ig/m3.
They have a pattern similar to the emission densities shown in Figure 2-2. These NSATA
modeled levels appear lower than the monitoring results presented above. For example, the 1999
air monitoring data (Table 2-6) indicates a median outdoor air level of 0.16 (j,g/m3 which is about
3 times as high as the modeled 1999 county median (0.054 |ig/m3). However, it should be
understood that the results from these two efforts are not perfectly comparable. The modeled
value is a median of county levels for the entire United States which includes many rural areas.
The monitors cover many fewer areas (n= 162 for 1999) and most are in nonrural locations. A
better analysis is provided by EPA (2007) which presents a comparison of modeling results from
NSATA to measured values at the same locations. For 1999, it was found that formaldehyde
levels were underestimated at 79% of the sites (n = 92). Thus, while the NSATA modeling
results are useful for understanding geographic distributions, they may frequently underestimate
ambient levels.
6/22/2009
15
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
2
3
4 Figure 2-3. Modeled ambient air concentrations of TCE
5
6/22/2009 16
Bi errjarek
¦revl dance
rinapolle
Saeramei
ere on
1990 Estimated County Median Ambient Concentrations
Trichbroethylene — United States Counties
Distribution of U.S.
Hlgh*et In U.S. .
Percentile
95
90
75
50
S5
In U.S.
Ambient
1 3.79
0.1S
0.099
Concentrations
_ County Median Ambient Pollutant Concentration
j}'™ ( micrograms / cubic meter)
a.a+5 Source; as EPA / OAQPS
Q'a19 1999 NATA National—Scale Afr Toxics .Assessment
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Indoor Air: TCE can be released to indoor air from use of consumer products that
contain it (i.e. adhesives and tapes), vapor intrusion (migration of volatile chemicals from the
subsurface into overlying buildings) and volatilization from the water supply. Where such
sources are present, it is likely that indoor levels will be higher than outdoor levels. A number of
studies have measured indoor levels of TCE:
• The 1987 EPA TEAM (Total Exposure Assessment Methodology) study (U.S. EPA,
1987) showed that the ratio of indoor to outdoor TCE concentrations for residences in
Greensboro, NC, was about 5:1.
• In two homes using well water with TCE levels averaging 22 to 128 (J,g/L, the TCE levels
in bathroom air ranged from <500 to 40,000 [j,g /m3 when the shower ran less than 30
minutes (Andelman et al., 1985).
• Shah and Singh (1988) report an average indoor level of 7.2 (J,g/m3 based on over 2000
measurements made in residences and workplaces during 1981-1984 from various
locations across the United States.
• Hers et al. (2001) provides a summary of indoor air TCE measurements at locations in
United States, Canada and Europe with a range of <1 to 165 |ig/m3,
• Sapkota et al. (2005) measured TCE levels inside and outside of the Baltimore Harbor
Tunnel toll booths during the summer of 2001. Mean TCE levels were 3.11 (J,g/m3
indoors and 0.08 (J,g/m3 outdoors based on measurements on 7 days. The authors
speculated that indoor sources, possibly dry cleaning residues on uniforms, were the
primary source of the indoor TCE.
• Sexton et al. (2005) measured TCE levels inside and outside residences in
Minneapolis/St. Paul metropolitan area. Two day samples were collected over three
seasons in 1999. Mean TCE levels were 0.5 (J,g/m3 indoors (// = 292), 0.2 (J,g/m3 outdoors
(n = 132) and 1.0 (J,g/m3 based on personal sampling (n=288).
• Zhu et al. (2005) measured TCE levels inside and outside of residences in Ottawa,
Canada. 75 homes were randomly selected and measurements were made during the
winter of 2002/2003. TCE was above detection limits in the indoor air of 33% of the
residences and in the outdoor air of 19% of the residences. The mean levels were 0.06
[j,g/m3 indoors and 0.08 (J,g/m3 outdoors. Given the high frequency of nondetects, a more
meaningful comparison can be made on basis of the 75th percentiles: 0.08 [ig/m3 indoors
and 0.01 (J,g/m3 outdoors.
TCE levels measured indoors have been directly linked to vapor intrusion at two sites in New
York:
6/22/2009
17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
• TCE vapor intrusion has occurred in buildings/residences near a former Smith Corona
manufacturing facility located in Cortlandville, NY. An extensive sampling program
conducted in 2006 has detected TCE in groundwater (1-13 (J,g/L), soil gas (6-97 (J,g/m3),
subslab gas (2-1600 (J,g/m3) and indoor air (1-17 (J,g/m3) (NYSDEC, 2006a).
• Evidence of vapor intrusion of TCE has also been reported in buildings and residences in
Endicott, NY. Sampling in 2003 showed total VOCs in soil gas exceeding 10,000 (J,g/m3
in some areas. Indoor air sampling detected TCE levels ranging from 1 to 140 (J,g/m3
(NYSDEC, 2006b).
Little et al. (1992) developed attenuation coefficients relating contaminants in soil gas
(assumed to be in chemical equilibrium with the groundwater) to possible indoor levels as a
result of vapor intrusion. On this basis they estimated that TCE groundwater levels of 540 |ig/L,
(a high contamination level) could produce indoor air levels of 5 to 500 |ig/m3. Vapor intrusion
is likely to be a significant source only in situations where residences are located near soils or
groundwater with high contamination levels. USEPA (2002) recommends considering vapor
intrusion when volatiles are suspected to be present in groundwater or soil at a depth of <100
feet. Hers et al. (2001) concluded that the contribution of VOCs from subsurface sources
relative to indoor sources is small for most chemicals and sites.
Water: A number of early (pre-1990) studies measured TCE levels in natural water
bodies (levels in drinking water is discussed later in this section) as summarized in Table 2-8.
According to IARC (1995), the reported median concentrations of TCE in 1983-84 were 0.5
[j,g/L in industrial effluents and 0.1 |ig/L in ambient water. Results from an analysis of the EPA
STORET Data Base (1980-1982) showed that TCE was detected in 28% of 9,295 surface water
reporting stations nationwide (ATSDR, 1997a).
6/22/2009
18
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 2-8. Concentrations of Trichloroethylene in Water Based on Pre-1990 Studies
Water 1\ |)c
I ocation
Year
Moan
Median
Uiingt'
IN urn her of S;im pics
Ref.
Og/L)
Og/L)
(Hg/L)
Industrial
U.S.
83
0.5
NR
I ARC, 1995
Effluent
Surface Waters
U.S.
83
0.1
NR
I ARC, 1995
Rainwater
Portland, OR
84
0.006
0.002-0.02
NR
Ligocki, etal, 1985
Groundwater
MN
83
0.2-144
NR
Sabel etal, 1984
NJ
76
<1530
NR
Burmaster et al., '82
NY
80
<3800
NR
Burmaster et al. ,'82
PA
80
<27300
NR
Burmaster et al. ,'82
MA
76
<900
NR
Burmaster et al. ,'82
AZ
8.9-29
NR
I ARC, 1995
Drinking water
U.S.
76
0.2-49
I ARC, 1995
U.S
77
0-53
I ARC, 1995
U.S.
78
0.5-210
I ARC, 1995
MA
84
max. 267
I ARC, 1995
NJ
84
23.4
max. 67
1130
Cohn et al., 1994
CA
85
8-12
486
U.S. EPA 1987
CA
84
66
486
U.S. EPA, 1987
NC
84
5
48
U.S. EPA, 1987
ND
84
5
48
U.S. EPA, 1987
NR - Not Reported
ATSDR (1997a) has reported that TCE is the most frequently reported organic
contaminant in groundwater and the one present in the highest concentration in a summary of
ground water analyses reported in 1982. It has been estimated that between 9 and 34% of the
drinking water supply sources tested in the United States may have some trichloroethylene
contamination. This estimate is based on available Federal and State surveys (ATSDR, 1997a).
Squillace et al. (2004) reported TCE levels in shallow groundwater based on data from
the National Water Quality Assessment Program managed by USGS. Samples from 518 wells
were collected from 1996 to 2002. All wells were located in residential or commercial areas and
had a median depth of 10 m. 8.3% of the well levels were above the detection limit, 2.3% were
above 0.1 [j,g/L and 1.7% were above 0.2 |ig/L,
The U.S. EPA Office of Ground Water and Drinking Water reported that most water
supplies are in compliance with the Maximum Contaminant Level or MCL (5 |ig/L) and that
only 407 samples out of many thousands taken from community and other water supplies
6/22/2009
19
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
throughout the country over the past 11 years (1987-1997) have exceeded the MCL limit for
TCE (U.S. EPA, 1998).
TCE concentrations in ground water have been measured extensively in California. The
data were derived from a survey of large water utilities (i.e., utilities with more than 200 service
connections). The survey was conducted by the California Department of Health Services (DHS,
1986). From January 1984 through December 1985, wells in 819 water systems were sampled
for organic chemical contamination. The water systems use a total of 5,550 wells, 2,947 of
which were sampled. TCE was found in 187 wells at concentrations up to 440 (J,g/L, with a
median concentration of 3.0 (J,g/L. Generally, the wells with the highest concentrations were
found in the heavily urbanized areas of the state. Los Angeles County registered the greatest
number of contaminated wells (149).
A second California study collected data on TCE levels in public drinking water
(Williams et al., 2002). The data were obtained from the California Department of Health
Services. The data spanned the years 1995 to 2001 and the n's for each year ranged from 3,447
to 4,226. The percent of sources that were above the detection limit ranged from 9.6 to 11.7 per
year (detection limits not specified). The annual average detected concentrations ranged from
14.2 to 21.6 |ig/L. Although not reported, the average over all of the samples (assuming an
average of 20 |ig/L among the samples above the detection limit, 10% detection rate and 0 for
the nondetects) would be about 2 (J,g/L.
The USGS (2006) conducted a national assessment of 55 volatile organic compounds
(VOCs), including trichloroethylene, in ground water. A total of 3,500 water samples were
collected during 1985-2001. Samples were collected at the well head prior to any form of
treatment. The types of wells sampled included 2,400 domestic wells and 1,100 public wells.
Almost 20% of the samples contained one or more of the VOCs above the assessment level of
0.2 (J,g/L. The detection frequency increased to over 50% when a subset of samples was
analyzed with a low level method that had an assessment level of 0.02 (J,g/L. The largest
detection frequencies were observed in California, Nevada, Florida, the New England States and
Mid-Atlantic states. The most frequently detected VOCs (> 1% of samples) include TCE,
tetrachloroethylene, 1,1,1-trichloroethane (methyl chloroform), 1,2 dichloroethylene, and 1,1
dichloroethane. Findings specific to TCE include the following:
• Detection frequency was 2.6% at 0.2 [j.g/L and was 3.8% at 0.02 (J,g/L.
• The median concentration was 0.15 [j,g/L with a range of 0.02 to 100 (J,g/L.
• The number of samples exceeding the MCL (5 (J,g/L) was 6 at domestic wells and 9 at
public wells.
6/22/2009
20
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
USGS (2006) also reported that four solvents (TCE, tetrachloroethylene, 1,1,1-
trichloroethane and methylene chloride) occurred together in 5% of the samples. The most
frequently occurring two-solvent mixture was TCE and tetrachloroethylene. The report stated
that the most likely reason for this co-occurrence is the reductive dechlorination of
tetrachloroethylene to TCE.
Other media: Levels of TCE were found in the sediment and marine animal tissue
collected in 1980-81 near the discharge zone of a Los Angeles County waste treatment plant.
Concentrations were 17 [j,g/L in the effluent, <0.5 (J,g/kg in dry weight in sediment, and 0.3-7
[j,g/kg wet weight in various marine animal tissue (IARC, 1995). TCE has also been found in a
variety of foods. FDA has limits on TCE use as a food additive in decaffeinated coffee and
extract spice oleoresins (see Table 2-15). Table 2-9 summarizes data from two sources:
• IARC (1995) reports average concentrations of TCE in limited food samples collected in
the United States
• Fleming-Jones and Smith (2003) measured VOC levels in over 70 foods collected from
1996 to 2000 as part of the FDA's Total Diet Program. All foods were collected directly
from supermarkets. Analysis was done on foods in a ready-to-eat form. Sample sizes for
most foods were in the 2-5 range.
6/22/2009
21
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 2-9. Levels in Food
(IARC, 1995)
Fleming-Jones and Smith (2003)
Cheese 3.8 (Jg/kg
Butter and Margarine 73.6 (Jg/kg
Cheese 2-3 j_ig/kg
Butter 7-9 (_ig/kg
Margarine 2-21 (Jg/kg
Cheese Pizza 2 (Jg/kg
Peanut Butter 0.5 (Jg/kg
Nuts 2-5 (Jg/kg
Peanut Butter 4-70 (Jg/kg
Ground Beef 3-6 (Jg/kg
Beef Frankfurters 2-105 (J-g/kg
Hamburger 5-9 (Jg/kg
Cheeseburger 7 (J-g/kg
Chicken Nuggets 2-5 (Jg/kg
Bologna 2-20 (Jg/kg
Pepperoni Pizza 2 (Jg/kg
Banana 2 (Jg/kg
Avocado 2-75 (Jg/kg
Orange 2 (Jg/kg
Chocolate Cake 3-57 (Jg/kg
Blueberry Muffin 3-4 (Jg/kg
Sweet Roll 3 (Jg/kg
Chocolate Chip Cookies 2-4 (Jg/kg
Apple Pie 2-4 (Jg/kg
Doughnuts 3 (Jg/kg
Tuna 9-11 (Jg/kg
Cereals 3 (Jg/kg
Grain-based Foods 0.9 j_ig/kg
Cereal 3 (Jg/kg
Popcorn 4-8 (Jg/kg
French Fries 3 (Jg/kg
Potato Chips 4-140 (Jg/kg
Coleslaw 3 (Jg/kg
6/22/2009
22
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
2 Biological Monitoring: Biological monitoring studies have detected TCE in human
3 blood and urine in the United States and other countries such as Croatia, China, Switzerland, and
4 Germany (IARC, 1995). Concentrations of TCE in persons exposed through occupational
5 degreasing operations were most likely to have detectable levels (IARC, 1995). In 1982, 8 of 8
6 human breastmilk samples from 4 U.S. urban areas had detectable levels of TCE. The levels of
7 TCE detected, however, are not specified (HSDB, 2002; ATSDR, 1997a).
8 The Third National Health and Nutrition Examination Survey (NHANES III) examined
9 TCE concentrations in blood in 677 non-occupationally exposed individuals. The individuals
10 were drawn from the general U.S. population and selected on the basis of age, race, gender and
11 region of residence (IARC, 1995; Ashley et al., 1994). The samples were collected during 1988
12 to 1994. TCE levels in whole blood were below the detection limit of 0.01 [j.g/L for about 90%
13 of the people sampled (Table 2-10). Assuming that nondetects equal half of the detection limit,
14 the mean concentration was about 0.017 (J,g/L.
15
16 Table 2-10. TCE Levels in Whole Blood by Population Percentile
Percentiles
ID
2<)
30
40
5<>
(•>0
70
80
Concentration
(l^g/L)
ND
ND
ND
ND
ND
ND
ND
ND
0.012
17 ND = Nondetect, i.e. below detection limit of 0.01 |ig/L.
18 Data from IARC (1995) and Ashley (1994)
19
20 2.4 EXPOSURE PATHWAYS AND LEVELS
21 2.4.1 General Population
22 Because of the pervasiveness of TCE in the environment, most people are likely to have
23 some exposure via one or more of the following pathways: ingestion of drinking water,
24 inhalation of outdoor/indoor air, or ingestion of food (ATSDR, 1997a). As noted earlier, the
25 NHANES survey suggests that about 10% of the population has detectable levels of TCE in
26 blood. Each pathway is discussed below.
27 2.4.1.1 Inhalation
28 As discussed earlier, EPA has estimated emissions and modeled air concentrations for the
29 Criteria Pollutants and Hazardous Air Pollutants under the National-Scale Air Toxics
6/22/2009
23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Assessment program (U.S. EPA, 2007a). This program has also estimated inhalation exposures
on a nationwide basis. The exposure estimates are based on the modeled concentrations from
outdoor sources and human activity patterns (U.S. EPA, 2005). Table 2-11 shows the 1999
results for TCE.
Table 2-11. Modeled 1999 Annual Exposure Concentrations ([j,g/m3) for Trichloroethylene
Exposure Concentration ([ig/m3)
Percentile
Rural Areas
Urban Areas
Nationwide
5
0.030
0.048
0.038
10
0.034
0.054
0.043
25
0.038
0.065
0.056
50
0.044
0.086
0.076
75
0.053
0.122
0.113
90
0.070
0.189
0.172
95
0.097
0.295
0.262
Mean
0.058
0.130
0.116
Percentiles and mean are based on census tract values.
Source: http://www.epa.gOv/ttn/atw/nata/ted/exporisk.html#indb
These modeled inhalation exposures would have a geographic distribution similar to that of the
modeled air concentrations as shown in Figure 2-3. Table 2-11 indicates that TCE inhalation
exposures in urban areas are generally about twice as high as rural areas. While these modeling
results are useful for understanding the geographic distribution of exposures, they appear to
under estimate actual exposures. This is based on the fact that, as discussed earlier, the modeled
ambient air levels are generally lower than measured values. Also, the modeled exposures do
not consider indoor sources. Indoor sources of TCE make the indoor levels higher than ambient
levels. This is particularly important to consider since people spend about 90% of their time
indoors (U.S. EPA, 1997). A number of measurement studies were presented earlier that showed
higher TCE levels indoors than outdoors. Sexton et al. (2005) measured TCE levels in
Minneapolis/St. Paul area and found means of 0.5 (J,g/m3 indoors (n = 292) and 1.0 (J,g/m3 based
on personal sampling (n = 288). Using 1.0 (J,g/m3 and an average adult inhalation rate of 13 m3
air/day (US EPA, 1997) yields an estimated intake of 13 (J,g/day. This is consistent with ATSDR
(ATSDR, 1997a), which reports an average daily air intake for the general population of 11 to 33
fig/day.
6/22/2009
24
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
2.4.1.2 Ingestion
The California survey of large water utilities in 1984-1985 found a median concentration
of 3.0 [j,g/L (DHS, 1986). The median value from the nationwide survey by USGS for 1985—
2001 is 0.15 [j,g/L which is much lower than the California survey. Several factors contribute to
this lower finding: the USGS survey includes domestic as well as public wells, covers a later
time period and includes a wider geographic area. Therefore, the USGS value is more current
and more representative of the national population. Using this value and an average adult water
consumption rate of 1.4 L/d (EPA, 1997) yields an estimated intake of 0.2 (J,g/day. This is lower
than the ATSDR (1997a) estimate water intake for the general population of 2 to 20 |ig/day. The
use of the USGS survey to represent drinking water is uncertain in two ways. First, the USGS
survey measured only groundwater and some drinking water supplies use surface water. Second,
the USGS measured TCE levels at the well head, not the drinking water tap. Further discussion
about the possible extent and magnitude of TCE exposure via drinking water is presented below.
TCE is the most frequently reported organic contaminant in ground water (ATSDR,
1997a), 93% of the public water systems in the United States obtain water from groundwater
(U.S. EPA, 1995) and between 9 and 34% of the drinking water supply sources tested in the
United States may have some TCE contamination (ATSDR, 1997a). Although commonly
detected in water supplies, the levels are generally low because, as discussed earlier, MCL
violations for TCE in public water supplies are relatively rare for any extended period (U.S.
EPA, 1998). The USGS (2006) survey found that the number of samples exceeding the MCL (5
(j,g/L) was 6 at domestic wells (n = 2,400) and 9 at public wells (n = 1,100). Private wells,
however, are often not closely monitored and if located near TCE disposal/contamination sites
where leaching occurs, may have undetected contamination levels. About 10% of Americans
(27 million people) obtain water from sources other than public water systems, primarily private
wells (U.S. EPA, 1995). TCE is a common contaminant at Superfund sites. It has been
identified in at least 861 of the 1,428 hazardous waste sites proposed for inclusion on the EPA
National Priorities List (NPL) (ATSDR, 1997a). Studies have shown that many people live near
these sites: 41 million people live less than 4 miles from one or more of the nation's NPL sites,
and on average 3,325 people live within 1 mile of any given NPL site (ATSDR, 1996b).
Table 2-12 presents preliminary estimates of TCE intake from food. They are based on
average adult food ingestion rates and food data from Table 2-9. This approach suggests a total
ingestion intake of about 5 j_ig/d. It is important to consider this estimate as preliminary because
it is derived by applying data from very limited food samples to broad classes of food.
6/22/2009
25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 2-12. Preliminary Estimates of TCE Intake from Food Ingestion
Consumption
Rate (g/kg-d)
Consumption
Rate (g/d)
Concentration in
Food (jag/kg)
Intake
(Hg/d)
Fruit
3.4
238
2
0.48
Vegetables
4.3
301
3
0.90
Fish
20
10
0.20
Meat
2.1
147
5
0.73
Dairy Products
8
560
3
1.68
Grains
4.1
287
3
0.86
Sweets
0.5
35
3
0.10
Total
4.96
1. Consumption rates are per capita averages from U.S. E
PA (1997).
2. Consumption rates in g/d assume 70 kg body weight.
2.4.1.3 Dermal
TCE in bathing water and consumer products can result in dermal exposure. A modeling
study has suggested that a significant fraction of the total dose associated with exposure to
volatile organics in drinking water results from dermal absorption (Brown et al., 1984). EPA
(2004) used a prediction model based on octanol-water partitioning and molecular weight to
derive a dermal permeability coefficient for TCE in water of 0.012 cm/hr. EPA used this value
to compute the dermally absorbed dose from a 35 minute shower and compared it to the dose
from drinking 2 L of water at the same concentration. This comparison indicated that the dermal
dose would be 17% of the oral dose. Much higher dermal permeabilities were reported by Nakai
et al. (1999) based on human skin in vitro testing. For dilute aqueous solutions of TCE, they
measured a permeability coefficient of 0.12 cm/hr (26°C). Nakai et al. (1999) also measured a
permeability coefficient of 0.018 cm/hr for tetrachloroethylene in water. Poet et al. (2000)
6/22/2009
26
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
measured dermal absorption of TCE in humans from both water and soil matrices. The absorbed
dose was estimated by applying a physiologically based pharmacokinetic model to TCE levels in
breath. The permeability coefficient was estimated to be 0.015 cm/hr for TCE in water and
0.007 cm/hr for TCE in soil (Poet et al, 2000).
2.4.1.4 Exposure to TCE Related Compounds
Table 2-13 presents adult exposure estimates that have been reported for the TCE related
compounds. This table was originally compiled by Wu and Schaum, 2001. The exposure/dose
estimates are taken directly from the listed sources or derived based on monitoring data
presented in the source documents. They are considered "preliminary" because they are
generally based on very limited monitoring data. These preliminary estimates suggest that
exposures to most of the TCE related compounds are comparable to or greater than TCE itself.
6/22/2009
27
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 2-13. Preliminary intake estimates of TCE and TCE-related chemicals
( hcmii'iil
l*< >l>uLil i<
Mnl in
l\iinM-iiri".
ill!! ilii\ i
l\;lll«!r ill' Vllllll Diim's
(III!! ks ll;l\ i
l):il:i *>iiiirii".
Trichloroethylene
General
Air
11 --33
1.57E-04-4.71E-04
ATSDR (1997a)
General
Water
2-20**
2.86E-05-2.86E-04
ATSDR (1997a)
Occupational
Air
2,232 - 9,489
3.19E-02-1.36E-01
ATSDR (1997a)
Tetrachloroethylene (PERC)
General
Air
80 - 200
1.14E-03-2.86E-03
ATSDR (1997b)
General
Water
0.1 — 0.2
1.43E-06-2.86E-06
ATSDR (1997b)
Occupational
Air
5,897 -- 219,685
8.43E-02-3.14
ATSDR (1997b)
1,1,1-Trichloroethane
General
Air
10.8-- 108
1.54E-04-1.54E-03
ATSDR (1995)
General
Water
0.38-- 4.2
5.5E-06-6.00E-05
ATSDR (1995)
1,2-Dichloroethylene
General
Air
1 --6
1.43E-05-8.57E-05
ATSDR (1996a)
General
Water
2.2
3.14E -05
ATSDR (1996a)
Cis-1,2-Dichloroethylene
General
Air
5.4
7.71E-05
HSDB (1996)
General
Water
0.5 -- 5.4
7.14E-06 -- 7.71E-05
HSDB (1996)
1,1,1,2-Tetrachloroethane
General
Air
142
2.03E-03
HSDB (2002)
1,1-Dichloroethane
General
Air
4
5.71E-05
ATSDR (1990)
General
Water
2.47 - 469.38
3.53E-05 — 6.71E-03
ATSDR (1990)
Chloral
General
Water
0.02 -- 36.4
2.86E-07 -- 5.20E-04
HSDB (1996)
Monochloroacetic Acid
General
Water
2 -- 2.4
2.86E-05 — 3.43E-05
USEPA (1994)
Dichloroacetic Acid
General
Water
10 -- 266
1.43E-04 - 3.80E-03
I ARC (1995)
Trichloroacetic Acid
General
Water
8.56 -- 322
1.22E-03 — 4.60E-03
I ARC (1995)
2 * Originally compiled in Wu and Schaum, 2001
3 ** New data from USGS (2006) suggests much lower water intakes, i.e. 0.2 jxg/d.
4
5 2.4.2 Potentially Highly Exposed Populations
6 Some members of the general population may have elevated TCE exposures. ATSDR
7 (ATSDR, 1997a) has reported that TCE exposures may be elevated for people living near waste
8 facilities where TCE may be released, residents of some urban or industrialized areas, people
6/22/2009
28
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
exposed at work (discussed further below) and individuals using certain products (also discussed
further below). Because TCE has been detected in breast milk samples of the general
population, infants who ingest breast milk may be exposed, as well. Increased TCE exposure is
also a possible concern for bottle-fed infants because they ingest more water on a bodyweight
basis than adults (the average water ingestion rate for adults is 21 mL/kg-d and for infants under
one year old it is 44 mL/kg-d - USEPA, 1997). Also, because TCE can be present in soil,
children may be exposed through activities such as playing in or ingesting soil.
Occupational Exposure: Occupational exposure to TCE in the United States has been
identified in various degreasing operations, silk screening, taxidermy, and electronics cleaning
(IARC, 1995). The major use of trichloroethylene is for metal cleaning or degreasing (IARC,
1995). Degreasing is used to remove oils, greases, waxes, tars, and moisture before galvanizing,
electroplating, painting, anodizing, and coating. The five primary industries using TCE
degreasing are: furniture and fixtures; electronic and electric equipment; transport equipment;
fabricated metal products; and miscellaneous manufacturing industries (IARC, 1995).
Additionally, TCE is used in the manufacture of plastics, appliances, jewelry, plumbing fixtures,
automobile, textiles, paper, and glass (IARC, 1995).
Table 2-13 lists the primary types of industrial degreasing procedures and the years that
the associated solvents were used. Vapor degreasing has the highest potential for exposure
because vapors can escape into the work place. Hot dip tanks, where trichloroethylene is heated
to close to its boiling point of 87°C, are also major sources of vapor that can create exposures as
high as vapor degreasers. Cold dip tanks have a lower exposure potential, but they have a large
surface area which enhances volatilization. Small bench-top cleaning operations with a rag or
brush and open bucket have the lowest exposure potential. In combination with the vapor
source, the size and ventilation of the workroom are the main determinants of exposure intensity
(NRC, 2006).
Occupational exposure to TCE has been assessed in a number of epidemiologic studies.
Studies of aircraft workers show short term peak exposures in the hundreds of ppm (>540
mg/m3) and long term exposures in the low tens of ppm (>54 mg/m3) (Spirtas et al, 1991; Blair et
al, 1998; Garabrant et al., 1988; Morgan et al., 1998; and Boice et al., 1998). Similar exposures
have been reported for cardboard/paperboard workers (Henschler et al., 1995; Sinks et al., 1992)
and uranium processors (Ritz, 1999). ATSDR (1997a) reports that the majority of published
worker exposure data show time weighted average (TWA) concentrations ranging from <50 ppm
to 100 ppm (<270 - 540 mg/m3 ). NIOSH conducted a survey of various industries from 1981 to
1983 and estimated that approximately 401,000 U.S. employees in 23,225 plants in the United
States were potentially exposed to TCE during this timeframe (IARC, 1995; ATSDR, 1997a).
6/22/2009
29
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Occupational exposure to TCE has likely declined since the 1950's and 1960's due to
decreased usage, better release controls and improvements in worker protection. Reductions in
TCE use are illustrated in Table 2-14, which shows that by about 1980 common degreasing
operations had substituted other solvents for TCE.
Table 2-14. Years of Solvent Use in Industrial Degreasing and Cleaning Operations
Years
Vapor Degreasers
Cold Dip Tanks
Rag or Brush and Bucket on Bench
Top
-1934-1954
T richloroethylene
(poorly controlled)
Stoddard solvent*
Stoddard solvent (general use), alcohols
(electronics shop), carbon tetrachloride
(instrument shop).
-1955-1968
T richloroethylene
(poorly controlled,
tightened in 1960s)
T richloroethylene
(replaced some
Stoddard solvent)
Stoddard solvent, trichloroethylene
(replaced some Stoddard solvent),
perchloroethylene, 1,1,1 -trichloroethane
(replaced carbon tetrachloride, alcohols,
ketones).
-1969-1978
Trichloroethylene,
(better controlled)
T richloroethylene,
Stoddard solvent
Trichloroethylene, perchloroethylene,
1,1,1-trichloroethane, alcohols, ketones,
Stoddard solvent.
~1979-1990s
1,1,1 -Trichloroethane
(replaced
trichloroethylene)
1,1,1 -Trichloroethane
(replaced
trichloroethylene),
Stoddard solvent
1,1,1 -Trichloroethane, perchloroethylene,
alcohols, ketones, Stoddard solvent.
* A mixture of straight and branched chain paraffins (48%), naphthenes (38%) and aromatic hydrocarbons (14%).
Source: Stewart and Dosemeci 2005.
Consumer Exposure: Consumer products reported to contain TCE include wood stains,
varnishes, and finishes; lubricants; adhesives; typewriter correction fluids; paint removers; and
cleaners (ATSDR, 1997a). Use of TCE has been discontinued in some consumer products (i.e.,
as an inhalation anesthetic, fumigant, and an extractant for decaffeinating coffee) (ATSDR,
1997a).
6/22/2009
30
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
2 2.4.3 Exposure Standards
3 Table 2-15 summarizes the federal regulations limiting TCE exposure.
4
5 Table 2-15. TCE Standards
Sliiiuliird
\ ill no
Reference
OSHA Permissible Exposure Limit: Table Z-
2 8-hr Time Weighted Average
100 ppm
(538 mg/m3)
29 CFR 1910.1000 (7/1/2000)
OSHA Permissible Exposure Limit: Table Z-
2 Acceptable Ceiling Concentration (this
cannot be exceeded for any time period
during an 8 hour shift except as allowed in the
maximum peak standard below)
200 ppm
(1076 mg/m3)
29 CFR 1910.1000 (7/1/2000)
OSHA Permissible Exposure Limit: Table Z-
2 Acceptable maximum peak above the
acceptable ceiling concentration for an 8-hour
shift. Maximum Duration: 5 minutes in any 2
hours.
300 ppm
(1614 mg/m3)
29 CFR 1910.1000 (7/1/2000)
MCL under the Safe Drinking Water Act
5 ppb (5 (rg/L)
USEPA/Office of Water;
Federal-State Toxicology and
Risk Analysis Committee
(FSTRAC). Summary of State
and Federal Drinking Water
Standards and Guidelines
(11/93)
FDA Tolerances for:
decaffeinated ground coffee
decaffeinated soluble (instant) coffee
extract spice oleoresins
25 ppm (25 (xg/g)
10 ppm (10 (xg/g)
30 ppm (30 (xg/g)
21 CFR 173.290 (4/1/2000)
6
6/22/2009
31
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
2.5 Exposure Summary
TCE is a volatile compound with moderate water solubility. Most TCE produced today
is used for metal degreasing. The highest environmental releases are to the air. Ambient air
monitoring data suggests that levels have remained fairly constant since 1999 at about 0.3 |ig/m3.
Indoor levels are commonly 3 or more times higher than outdoor levels due to releases from
building materials and consumer products. TCE is among the most common groundwater
contaminants and the median level based on a large survey by USGS for 1985-2001 is 0.15 (J,g/L.
It has also been detected in a wide variety of foods in the 1-100 (J,g/kg range. None of the
environmental sampling has been done using statistically based national surveys. However, a
substantial amount of air and groundwater data have been collected allowing reasonably well
supported estimates of typical daily intakes by the general population: inhalation - 13 (j,g/day
and water ingestion - 0.2 (j,g/day. The limited food data suggests an intake of about 5 |ig/day, but
this must be considered preliminary.
Much higher exposures have occurred to various occupational groups. For example, past
studies of aircraft workers have shown short term peak exposures in the hundreds of ppm
(>540,000 (J,g/m3) and long term exposures in the low tens of ppm (>54,000 (J,g/m3).
Occupational exposures have likely decreased in recent years due to better release controls and
improvements in worker protection.
Preliminary exposure estimates were presented for a variety of TCE related compounds
which include metabolites of TCE and other parent compounds that produce similar metabolites.
Exposure to the TCE related compounds can alter or enhance TCE's metabolism and toxicity by
generating higher internal metabolite concentrations than would result from TCE exposure by
itself. The preliminary estimates suggest that exposures to most of the TCE related compounds
are comparable to or greater than TCE itself.
2.6 REFERENCES
JB Andelman (1985). Human exposure to volatile halogenated organic chemicals in indoor and
outdoor air. Environmental Health Perspectives. 62:313-318.
DL Ashley, MA Bonin, FL Cardinali, JM McCraw and JV Wooten (1994). Blood
concentrations of volatile organic compounds in a nonoccupationally exposed US population
and in groups with suspected exposure. Clinical Chemistry, Vol.40, No.7.
ATSDR (1990). Toxicological profile for 1,1-dichloroethane. Atlanta, GA: Agency for Toxic
Substances and Disease Registry.
6/22/2009
32
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
ATSDR(1995). Toxicological profile for 1,1,1-trichloroethane. Update. Atlanta, GA: Agency
for Toxic Substances and Disease Registry.
ATSDR (1996a). Toxicological profile for 1,2-dichloroethene. Update. Atlanta, GA: Agency
for Toxic Substances and Disease Registry.
ATSDR (1996b). Biennial report to congress (1991 and 1992). Atlanta, GA:U.S. Department of
Health and Human Services, Centers for Disease Control and Prevention, ATSDR. Internet site:
www.dhhs.gov.
ATSDR (1997a). Toxicological profile for trichloroethylene. Update. Atlanta, GA: Agency for
Toxic Substances and Disease Registry.
ATSDR (1997b). Toxicological profile for tetrachloroethylene. Update. Atlanta, GA: Agency
for Toxic Substances and Disease Registry.
T Boublik, V Fried, and E Hala (1984). The vapour pressures of pure substances. Second
Revised Edition. Amsterdam: Elsevier.
HS Brown, DR Bishop, and CA Rowan (1984). The role of skin absorption as a route of
exposure for volatile organic compounds in drinking water. American Journal of Public Health.
Vol 74, No.5, 479-484.
NJ Bunce and UA Schneider (1994). Chemical lifetimes of chlorinated aliphatic priority
pollutants in the Canadian troposphere. Journal of Photochemistry and Photobiology A:
Chemistry, 81:93-101.
DE Burmaster (1982). The new pollution-groundwater contamination. Environment, 24, No.2,
6-36.
Chemical Marketing Reporter (1997). Chemical profile trichloroethylene. Dec 8, 1997. NY,
NY: Schnell Pub Co.
6/22/2009
33
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
P Cohn, J Klotz, F Bove, M Berkowitz, J Fagliano (1994). Drinking water contamination and
the incidence of leukemia and non-Hodgkin's lymphoma. Environ. Health Perspectives,
102:556-560.
DHS (1986). Organic Chemical Contamination of Large Public Water Systems in California.
California Department of Health Services, Sacramento, CA.
ME Fleming-Jones and RE Smith. 2003. Volatile organic compounds in foods: A five year
study. J. Agric. Food Chem. 51:8120-8127
C Hansch, A Leo and D. Hoekman. (1995). Exploring QSAR - Hydrophobic, electronic, and
steric constants. Washington, DC: American Chemical Society.
HSDB (2002). Hazardous Substances Data Bank. Trichloroethylene. National Library of
Medicine. National Institutes of Health. Washington, DC: U.S. Department of Health and
Human Services.
AH Hendler and WL Crow (1992). Proc Ann Meet Air Waste Manage Assoc, 85thMeeting,
Kansas City, MS, 92-75.05, 94-08743-2-9.
I Hers, R Zapf-Gilje, L Li and J Atwater. 2001. The use of indoor air measurements to evaluate
intrusion of subsurface VOC vapors into buildings. J Air Waste Manage Assoc. 51:1318-1331.
AL Hotvath, FW Getzen and Z Maczynska. (1999). IUPAC-NIST Solubility Data Series 67.
Halogenated Ethanes and Ethenes With Water. J Phys Chem Ref Data, 28:395-627.
PH Howard, RS Boethling, WF Jarvis, WM Meylan and EM Michalenko (1991). Handbook of
environmental degradation rates. Chelsea, MI: Lewis Publishers.
IARC (1995). IARC monographs on the evaluation of carcinogenic risks to humans: dry
cleaning, some chlorinated solvents and other industrial chemicals. Vol. 63.
DT Leighton Jr and JM Calo (1981). Distribution coefficients of chlorinated hydrocarbons in
dilute air-water systems for groundwater contamination applications. J Chem Eng Data 26: 382-
85.
6/22/2009
34
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
R Lewis (2001). Hawley's Condensed Chemical Dictionary. 14th Edition. John Wiley and Sons.
New York.
MP Ligocki, C Leuenberger, JF Pankow (1985) Trace organic compounds in rain-II. Gas
scavenging of neutral organic compounds. Atmos. Environ., 19:1609-1617.
Little, JC, JM Daisey and WW Nazaroff (1992). Transport of subsurface contaminants into
buildings, an exposure pathway for volatile organics. Environ. Sci. Technol. 26:2058-2066.
Nakai JS, Stathopulos PB, Campbell GL, Chu I, Li-Muller A, Aucoin R.(1999). Penetration of
chloroform, trichloroethylene, and tetrachloroethylene through human skin. J Toxicol Environ
Health A. 58(3): 157-70.
National Research Council of the National Academies (NRC), 2006. Assessing the Human
Health Risks of Trichloroethylene, Key Scientific Issues. The National Academies Press.
Washington, DC.
New York State Department of Environmental Conservation (NYSDEC). 2006a.
http://www.dec.state.nv.us/website/der/proiects/712006.html
New York State Department of Environmental Conservation (NYSDEC). 2006b.
http://www.dec.state.ny.us/website/der/proiects/endicott/endicottsampling.htm
Poet TS, RA Corley, KD Thrall, JA Edwards, H Tanojo, KW Weitz, X Hui, HI Maibach, and RC
Wester (2000). Assessment of the percutaneous absorption of trichloroethylene in rats and
humans using MS/MS real-time breath analysis and physiologically based pharmacokinetic
modeling. Toxicological Sciences 56:61-72.
GV Sabel and TP Clark (1984). Volatile organic compounds as indicators of municipal solid
waste leachate contamination. Waste Management & Research, 2:119-130.
A Sapkota, D Williams, and T Buckley. 2005. Tollbooth workers and mobile source-related
hazardous air pollutants: How protective is the indoor environment? Environ. Sci. Technol.
39:2936-2943.
6/22/2009
35
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
PA Scheff and RA Wadden (1993). Receptor modeling of volatile organic compounds. 1.
Emission inventory and validation. Environ Sci Technol 27: 617-25
K Sexton, JL Adgate, G Ramachandran, GC Pratt, SJ Mongin, TH Stock, and MT Morandi.
2004. Comparison of personal, indoor and outdoor exposures to hazardous air pollutants in three
urban communities. Environ. Sci. Technol. 38:423-430.
JJ Shah and HB Singh (1988). Distribution of volatile organic chemicals in outdoor and indoor
air: a national VOCs data base. Environ Sci Technol 22:1381-1388.
PJ Squillace, MJ Moran and CV Price (2004). VOCs in shallow groundwater in new
residential/commercial areas of the US. Environ Sci Technol 38:5327-5338.
SRI, 1992. 1992 Directory of Chemical Producers-United States of America. Menlo Park, CA:
SRI International, 1032.
Stewart, P., and M. Dosemeci. 2005. Exposure Estimation in Occupational Epidemiology
Studies. Presentation at the Third Meeting on Assessing Human Health Risks of
Trichloroethylene, June 9, 2005, Irvine, CA.
CW Sweet and SJ Vermette (1992). Toxic volatile organic compounds in urban air in Illinois.
Environ Sci Technol 26:165-73.
U.S. Geologic Survey (USGS), 2006. Volatile organic compounds in the nation's ground water
and drinking-water supply wells. Circular 1292. Reston, VA.
U.S. EPA. (1987). The total exposure assessment methodology (TEAM) study: summary and
analysis. Volume 1, EPA/600/6-87/002a.
U.S. EPA. 1995. The national public water systems supervision program. The FY 1994
Compliance Report. Office of Water. EPA 812-R-94-001. Washington: U.S. Environmental
Protection Agency.
U.S. EPA. (1997). Exposure Factors Handbook. EPA/600/P-95/002F
6/22/2009
36
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
U.S. EPA. (1998). EPA Data File. Office of Groundwater and Drinking Water. Washington,
DC.
U.S. EPA (2002). OSWER draft guidance for evaluating the vapor intrusion to indoor air
pathway from groundwater and soils.
http://epa.gov/osw/hazard/correctiveaction/eis/vapor/guidance.pdf
U.S. EPA (2004). Risk assessment guidance for superfund. Volume I: Human health evaluation
manual (Part E, Supplemental guidance for dermal risk assessment). EPA/540/R/99/005.
U.S. EPA (2005). The HAPEM5 User's Guide. Hazardous Air Pollutant Exposure Model,
Version 5. EPA Office of Air Quality Planning and Standards, Research Triangle Park, North
Carolina.
U.S. EPA (2007a). Technology Transfer Network - 1999 National-Scale Air Toxics Assessment.
http://www.epa.gov/ttn/atw/natal999/nsata99.html
U.S. EPA (2007b). EPA's Air Quality System database at the AirData website:
http://www.epa.gov/air/data/index.html. Office of Air Quality Planning & Standards. Research
Triangle Park, NC.
WHO. (1985). Environmental health criteria: trichloroethylene. Geneva, Switzerland: World
Health Organization.
P Williams, L Benton, J Warmerdam and P Sheehan (2002). Comparative risk analysis of six
volatile organic compounds in California drinking water. Environ. Sci. Technol. 36:4721-4728.
C Wu and J Schaum (2000). Exposure assessment of trichloroethylene. Environ Health Perspect
108 (suppl 2):359-363.
C Wu and J Schaum (2001). Sources, emissions and exposures for trichloroethylene (TCE) and
related chemicals. Washington: US Environmental Protection Agency, EPA/600/R-00/099.
B Zielinska, E Fujita, J Sagebiel, GHarshfield, E Uberna, T Hayes and F Keene (1998). Arizona
hazardous air pollutants monitoring program. J Air Waste Manage Assoc 48:1038-1050.
6/22/2009
37
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 J Zhu, R Newhook, L Marro and CC Chan. 2005 accepted. Selected volatile organic compounds
2 in residential air in the city of Ottawa, Canada. Environ. Sci. Technol.
3
4
6/22/2009
38
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3 TOXICOKINETICS
TCE is a lipophilic compound that readily crosses biological membranes. Exposures may
occur via the oral, dermal, and inhalation route, with evidence for systemic availability from
each route. TCE is rapidly and nearly completely absorbed from the gut following oral
administration, and studies with animals indicate that exposure vehicle may impact the
time-course of administration: oily vehicles may delay absorption whereas aqueous vehicles
result in a more rapid increase in blood concentrations.
Following absorption to the systemic circulation, TCE distributes from blood to solid
tissues by each organ's solubility. This process is mainly determined by the blood:tissue partition
coefficients, which are largely established by tissue lipid content. Adipose partitioning is high,
adipose tissue may serve as a reservoir for TCE, and accumulation into adipose tissue may
prolong internal exposures. TCE attains high concentrations relative to blood in the brain,
kidney, and liver—all of which are important target organs of toxicity. TCE is cleared via
metabolism mainly in three organs: the kidney, liver, and lungs.
The metabolism of TCE is an important determinant of its toxicity. Metabolites are
responsible for toxicity—especially for the liver and kidney. Initially, TCE may be oxidized via
cytochrome P450 xenobiotic metabolizing isozymes or conjugated with glutathione by
glutathione-S-transferase enzymes. While CYP2E1 is generally accepted to be the CYP form
most responsible for TCE oxidation at low concentrations, others forms may also contribute,
though their contributions may be more important at higher, rather than lower, environmentally-
relevant exposures.
Once absorbed, TCE is excreted primarily either in breath as unchanged TCE or C02, or
in urine as metabolites. Minor routes of elimination include excretion of metabolites in saliva,
sweat, and feces. Following oral administration or upon cessation of inhalation exposure,
exhalation of unmetabolized TCE is a major elimination pathway. Initially, elimination of TCE
upon cessation of inhalation exposure demonstrates a steep concentration-time profile: TCE is
rapidly eliminated in the minutes and hours post-exposure, and then the rate of elimination via
exhalation decreases. Following oral or inhalation exposure, urinary elimination of parent TCE
is minimal, with urinary elimination of the metabolites trichloroacetic acid and trichloroethanol
accounting for the bulk of the absorbed dose of TCE.
Sections 3.1-3.4 below describe the absorption, distribution, metabolism, and excretion
of TCE and its metabolites in greater detail. Section 3.5 then discusses physiologically based
pharmacokinetic modeling of TCE and its metabolites.
6/22/2009
39
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.1 ABSORPTION
Trichloroethylene is a low-molecular-weight lipophilic solvent; these properties explain
its rapid transfer from environmental media into the systemic circulation after exposure. As
discussed below, it is readily absorbed into the bloodstream following exposure via oral
ingestion and inhalation, with more limited data indicating dermal penetration.
3.1.1 Oral
Available reports on human exposure to TCE via the oral route are largely restricted to
case reports of occupational or intentional (suicidal) ingestions and suggest significant gastric
absorption (e.g. Perbellini et al., 1991, Yoshida et al., 1996, Briining et al., 1998). Clinical
symptoms attributable to TCE or metabolites were observed in these individuals within a few
hours of ingestion (such as lack of consciousness), indicating absorption of TCE. In addition,
TCE and metabolites were measured in blood or urine at the earliest times possible after
ingestion, typically upon hospital admission, while urinary excretion of TCE metabolites was
followed for several days following exposure. Therefore, based on these reports, it is likely that
TCE is readily absorbed in the gastrointestinal tract; however, the degree of absorption cannot be
confidently quantified because the ingested amounts are not known.
Experimental evidence in mice and rats supports rapid and extensive absorption of TCE,
although variables such as stomach contents, vehicle, and dose may affect the degree of gastric
absorption. D'Souza et al. (1985) reported on bioavailability and blood kinetics in fasted and
non-fasted male Sprague-Dawley rats following intra-gastric administration of TCE at 5-25
mg/kg in 50% PEG 400 in water. TCE rapidly appeared in peripheral blood (at the initial 0.5
minutes sampling) of fasted and non-fasted rats with peak levels being attained shortly thereafter
(6-10 minutes), suggesting that absorption is not diffusion limited, especially in fasted animals.
The presence of food in the GI tract, however, seems to influence TCE absorption based on
findings in the non-fasted animals of lesser bioavailability (60-80% vs, 90% in fasted rats),
smaller peak blood levels (2-3 fold lower than non-fasted animals), and a somewhat longer
terminal half-life (ti/2) (174 min vs. 112 min in fasted rats).
Studies by Prout et al. (1985) and Dekant et al. (1986a) have shown that up to 98% of
administered radiolabel was found in expired air and urine of rats and mice following gavage
administration of [14C]TCE. Prout et al. (1985) and Green and Prout (1985) compared the
degree of absorption, metabolites, and routes of elimination among two strains each of male rats
(Osborne-Mendel and Park Wistar) and male mice (B6C3F1 and Swiss-Webster) following a
single oral administration of 10, 500, or 1000, [14C]-TCE. Additional dose groups of Osborne-
Mendel male rats and B6C3F1 male mice also received a single oral dose of 2000 mg/kg [14C]-
6/22/2009
40
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE. At the lowest dose of 10 mg/kg, there were no major differences between rats and mice in
routes of excretion with most of the administered radiolabel (nearly 60-70%) being in the urine.
At this dose, the expired air from all groups contained 1-4% of unchanged TCE and 9-14% C02.
Fecal elimination of the radiolabel ranged from 8.3% in Osborne-Mendel rats to 24.1% in Park
Wistar rats. However, at doses between 500 and 2000 mg/kg, the rat progressively excreted a
higher proportion of the radiolabel as unchanged TCE in expired air such that 78% of the
administered high dose was found in expired air (as unchanged TCE) while only 13% was
excreted in the urine.
Following exposure to a chemical by the oral route, distribution is determined by delivery
to the first organ encountered in the circulatory pathway—the liver (i.e., the first-pass effect),
where metabolism and elimination may limit the proportion that may reach extrahepatic organs.
Lee et al. (1996) evaluated the efficiency and dose-dependency of pre-systemic elimination of
TCE in male Sprague-Dawley rats following administration into the carotid artery, jugular vein,
hepatic portal vein, or the stomach of TCE (0.17, 0.33, 0.71, 2, 8, 16, or 64 mg/kg) in a 5%
aqueous Alkamus emulsion (polyethoxylated vegetable oil) in 0.9% saline. The first-pass
elimination, decreased from 57.5 to <1% with increasing dose (0.17-16 mg/kg) which implied
that hepatic TCE metabolism may be saturated at doses above 16 mg/kg in the male rat. At
doses of 16 mg/kg or higher, hepatic first-pass elimination was almost non-existent indicating
that, at relatively large doses, virtually all of TCE passes through the liver without being
extracted (Lee et al., 1996). In addition to the hepatic first-pass elimination findings, pulmonary
extraction, which was relatively constant (at nearly 5-8% of dose) over the dose range, also
played a role in eliminating TCE.
In addition, oral absorption appears to be affected by both dose and vehicle used. The
majority of oral TCE studies have used either aqueous solution or corn oil as the dosing vehicle.
Most studies that relied on an aqueous vehicle delivered TCE as an emulsified suspension in
Tween 80 or PEG 400 in order to circumvent the water solubility problems. Lee et al. (2000a,b)
used Alkamuls (a polyethoxylated vegetable oil emulsion) to prepare a 5% aqueous emulsion of
TCE that was administered by gavage to male Sprague-Dawley rats. The findings confirmed
rapid TCE absorption but reported decreasing absorption rate constants (i.e., slower absorption)
with increasing gavage dose (2-432 mg/kg). The time to reach blood peak concentrations
increased with dose and ranged between 2 and 26 minutes post-dosing. Other pharmacokinetics
data, including area under the blood concentration time curve (AUC) and prolonged elevation of
blood TCE levels at the high doses, indicated prolonged GI absorption and delayed elimination
due to metabolic saturation occurring at the higher TCE doses.
A study by Withey et al., (1983) evaluated the effect of dosing TCE with corn oil versus
pure water as a vehicle by administering four VOCs separately in each dosing vehicle to male
6/22/2009
41
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Wistar rats. Based on its limited solubility in pure water, the dose for TCE was selected at 18
mg/kg (administered in 5 ml/kg). Times to peak in blood reported for TCE averaged 5.6 minutes
when water was used. In comparison, the time to peak in blood was much longer (approximately
100 minutes) when the oil vehicle was used and the peaks were smaller, below the level of
detection, and not reportable.
Time-course studies reporting times to peak in blood or other tissues have been
performed using both vehicles (Withey et al., 1983; Larson and Bull, 1992 a,b; D'Souza et al.,
1985; Green and Prout, 1985; Dekant et al., 1984). Related data for other solvents (Kim et al.,
1990; Dix et al., 1997; Lilly et al., 1994; Chieco et al., 1981) confirmed differences in TCE
absorption and peak height between the two administered vehicles. One study has also evaluated
the absorption of TCE from soil in rats (Kadry et al., 1991) and reported absorption within 16
hours for clay and 24 hours for sandy soil. In summary, these studies confirm that TCE is
relatively quickly absorbed from the stomach, and that absorption is dependent on vehicle used.
3.1.2 Inhalation
Trichloroethylene is a lipophilic volatile compound, that is readily absorbed from
inspired air. Uptake from inhalation is rapid and the absorbed dose is proportional to exposure
concentration and duration, and pulmonary ventilation rate. Distribution into the body via
arterial blood leaving the lungs is determined by the net dose absorbed and eliminated by
metabolism in the lungs. Metabolic clearance in the lungs will be further discussed in section
3.3, below. In addition to metabolism, solubility in blood is the major determinant of the TCE
concentration in blood entering the heart and being distributed to the each body organ via the
arterial blood. The measure of TCE solubility in each organ is the partition coefficient, or the
concentration ratio between both organ phases of interest. The blood-to-air partition coefficient
(PC) quantifies the resulting concentration in blood leaving the lungs at equilibrium with
alveolar air. The value of the blood-to-air partition coefficient is used in PBPK modeling
(Section 3.5). The blood-to-air partition has been measured in vitro using the same principles in
different studies and found to range between 8.1-11.7 in humans and somewhat higher values in
mice and rats (13.3-25.8) (Table 3.1.la—3.1.1b, and references therein).
6/22/2009
42
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.1.1a. Bloodrair PC values for humans
Species/
BloodrAir Partition
Reference/Notes
Coefficient
HUMANS
8.1 ± 1.8
Fiserova-Bergerova et al. (1984) mean + SD (SD converted from SE
based on n = 5)
8.11
Gargas et al. (1989) («=3-15)
9.13 ± 1.73 [6.47-11]
Fisher et al. (1998) mean + SD [range] of females (n=6)
9.5
Sato and Nakajima (1979) (n= 1)
9.77
Koizumi (1989)
9.92
Sato et al. (1977) (n= 1)
11.15 + 0.74 [10.1-12.1]
Fisher et al. (1998) mean + SD [range] of males (n=7)
11.2+1.8 [7.9-15]
Mahle et al. (2007); mean + SD; 20 male pediatric patients aged 3-7
years [range; USAF, 2004]
11.0+1.6 [6.6-13.5]
Mahle et al. (2007); mean + SD; 18 female pediatric patients aged 3-
17years [range; USAF, 2004]
11.7+1.9 [6.7-16.8]
Mahle et al. (2007); mean + SD; 32 male patients aged 23-82 years
[range; USAF, 2004]
10.6 + 2.3 [3-14.4]
Mahle et al. (2007); mean + SD; 27 female patients aged 23-82
years [range; USAF, 2004]
2
3
6/22/2009
43
-------
1
2
3
4
5
6
7
8
9
10
11
12
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.1.1b. Bloodrair PC values for rats and mice
Species/
BloodrAir Partition
Refer ence/N otes
Coefficient
RAT
15+0.5
Fisher et al. (1989) mean + SD (SD converted from SE based on n=3)
17.5
Rodriguez et al. (2007)
20.5 + 2.4
Barton et al. (1995) mean + SD (SD converted from SE based on n=4)
20.69 + 3.3
Simmons et al. (2002) mean + SD (n=7-10)
21.9
Gargas et al. (1989) («=3-15)
25.8
Koizumi (1989) (pooled n=3)
25.82+ 1.7
Sato et al. (1977) mean + SD («=5)
13.3 + 0.8 [11.6-15]
Mahle et al. (2007); mean + SD; 10 PND 10 male rat pups [range;
USAF, 2004]
13.4+1.8 [11.8-17.2]
Mahle et al. (2007); mean + SD; 10 PND 10 female rat pups [range;
USAF, 2004]
17.5 + 3.6 [11.7-23.1]
Mahle et al. (2007); mean + SD; 9 adult male rats [range; USAF,
2004]
21.8+1.9 [16.9-23.5]
Mahle et al. (2007); mean + SD; 11 aged male rats [range; USAF,
2004]
MOUSE
13.4
Fisher et al. (1991) male
14.3
Fisher et al. (1991) female
15.91
Abbas and Fisher (1997)
TCE enters the human body by inhalation quickly and at high concentrations may lead to
death (Coopman et al., 2003), unconsciousness, and acute kidney damage (Carrieri et al., 2007).
Controlled exposure studies in humans have shown absorption of TCE to approach a steady state
within a few hours after the start of inhalation exposure (Monster et al., 1976,
Fernandez et al., 1977, Vesterberg et al. 1976, Vesterberg and Astrand 1976). Several studies
have calculated the net dose absorbed by measuring the difference between the inhaled
concentration and the exhaled air concentration. Soucek and Vlachova (1959) reported between
58-70% absorption of the amount inhaled for 5-hour exposures between 93-158 ppm.
Bartonicek (1962) obtained an average retention value of 58% after 5 hours of exposure to 186
ppm. Monster et al. (1976) also took into account minute ventilation measured for each
6/22/2009
44
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 exposure, and calculated between 37-49% absorption in subjects exposed to 70 and 140 ppm.
2 The impact of exercise, the increase in workload, and its effect on breathing has also been
3 measured in controlled inhalation exposures. Astrand and Ovrum (1976) reported 50-58%
4 uptake at rest and 25-46 % uptake during exercise from exposure at 100 or 200 ppm (540 or
5 1080 mg/m3, respectively) of TCE for 30 minutes (Table 3.1.2). These authors also monitored
6 heart rate and pulmonary ventilation. In contrast, Jakubowski and Wieczorek (1988) calculated
7 about 40% retention in their human volunteers exposed to TCE at 9.3 ppm (mean ispired
8 concentration of 48-49 mg/m3) for 2 hours at rest, with no change in retention during increase in
9 workload due to exercise (Table 3.1.3).
6/22/2009
45
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.1.2. Air and blood concentrations during exposure to TCE in humans (Astrand
2 and Ovrum, 1976)
TCE
Cone.
(mg/m3)
Work
Load
(Watt)
Exposure
Series
TCE Concentration in
Uptake as
% of
Amount
Available
Amount
Taken Up
(mg)
Alveolar
Air
(mg/m3)
Arterial
Blood
(mg/kg)
Venous
Blood
(mg/kg)
540
0
I
124 + 9
1.1 ±0.1
0.6 + 0.1
53 + 2
79 + 4
540
0
II
127+11
1.3 + 0.1
0.5 + 0.1
52 + 2
81+7
540
50
I
245 + 12
2.7 + 0.2
1.7 + 0.4
40 + 2
160 + 5
540
50
II
218 + 7
2.8 + 0.1
1.8 + 0.3
46+1
179 + 2
540
50
II
234+ 12
3.1 + 0.3
2.2 + 0.4
39 + 2
157 + 2
540
50
II
244+ 16
3.3 + 0.3
2.2 + 0.4
37 + 2
147 + 9
1080
0
I
280+ 18
2.6 + 0.0
1.4 + 0.3
50 + 2
156+9
1080
0
III
212 + 7
2.1 + 0.2
1.2 + 0.1
58 + 2
186 + 7
1080
50
I
459 + 44
6.0 + 0.2
3.3 + 0.8
45 + 2
702 + 31
1080
50
III
407 + 30
5.2 + 0.5
2.9 + 0.7
51 + 3
378 + 18
1080
100
III
542 + 33
7.5 + 0.7
4.8+1.1
36 + 3
418 + 39
1080
150
III
651 + 53
9.0+1.0
7.4+1.1
25 + 5
419 + 84
3 Series I consisted of 30-minute exposure periods of rest, rest, 50W and 50W; Series II consisted
4 of 30-minute exposure periods of rest, 50W, 50W, 50W; Series III consisted of 30-minute
5 exposure periods of rest, 50W, 100W, 150W.
6/22/2009
46
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.1.3. Retention of inhaled TCE vapor in humans (Jakubowski and Wieczorek,
2 1988)
Workload
Inspired Concentration
(mg/m3)
Pulmonary
Ventilation
(m3/hour)
Retention
Uptake
(mg/hour)
Rest
48 + 3a
0.65 + 0.07
0.40 + 0.05
12+1.1
25 W
49+1.3
1.30 + 0.14
0.40 + 0.05
25 + 2.9
50 W
49+1.6
1.53 + 0.13
0.42 + 0.06
31+2.8
75 W
48+1.9
1.87 + 0.14
0.41+0.06
37 + 4.8
3 aMean + S.D., n=6 adult males.
4
5 Environmental or occupational settings may results from a pattern of repeated exposure
6 to TCE. Monster et al. (1979) reported 70-ppm TCE exposures in volunteers for 4 hours for 5
7 consecutive days, averaging a total uptake of 450 mg per 4 hours exposure (Table 3.1.4). In
8 dry-cleaning workers, Skender et al. (1991) reported initial blood concentrations of 0.38 |imol/L,
9 increasing to 3.4 |imol/L 2 days after. Results of these studies support rapid absorption of TCE
10 via inhalation.
11
6/22/2009
47
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.1.4. Uptake of TCE in human volunteers following 4 hour exposure to 70 ppm
(Monster et al., 1979)
BW
(kg)
MV (L/min)
% Retained
Uptake
(mg/day)
Uptake (mg/kg/day)
A
80
9.8 + 0.4
45 + 0.8
404 + 23
5.1
B
82
12.0 + 0.7
44 + 0.9
485 + 35
5.9
C
82
10.9 + 0.8
49+1.2
493 + 28
6.0
D
67
11.8 + 0.8
35 + 2.6
385 + 38
5.7
E
90
11.0 + 0.7
46+1.1
481+25
5.3
Mean
5.6 + 0.4
Direct measurement of retention after inhalation exposure in rodents is more difficult
because exhaled breath concentrations are challenging to obtain. The only available data are
from Dallas et al. (1991), who designed a nose-only exposure system for rats using a facemask
equipped with one-way breathing valves to obtain measurements of TCE in inspired and exhaled
air. In addition, indwelling carotid artery cannulae were surgically implanted to facilitate the
simultaneous collection of blood. After a 1-hour acclimatization period, rats were exposed to 50-
or 500-ppm TCE for 2 hours and the time course of TCE in blood and expired air was measured
during and for 3 hours following exposure. When air concentration data were analyzed to reveal
absorbed dose (minute volume multiplied by the concentration difference between inspired and
exhaled breath), it was demonstrated that the fractional absorption of either concentration was
more than 90% during the initial 5 minutes of exposure. Fractional absorption then decreased to
69 and 71% for the 50 and 500 ppm groups during the second hour of exposure. Cumulative
uptake appeared linear with respect to time over the 2-hour exposure, resulting in absorbed doses
of 8.4 mg/kg and 73.3 mg/kg in rats exposed to 50 and 500 ppm, respectively. Given the 10-fold
difference in inspired concentration and the 8.7-fold difference in uptake, the authors interpreted
this information to indicate that metabolic saturation occurred at some concentration below 500
ppm. In comparing the absorbed doses to those developed for the 70-ppm-exposed human
(see Monster et al., 1979), Dallas et al. (1991) concluded that on a systemic dose (mg/kg) basis,
rats receive a much higher TCE dose from a given inhalation exposure than do humans. In
particular, using the results cited above, the absorption per ppm-hr was 0.084 and
0.073 mg/kg-ppm-hr at 50 and 500 ppm in rats (Dallas et al. 1991) and 0.019 mg/kg-ppm-hr at
70 ppm in humans (Monster et al. 1979)—a difference of around 4-fold. However, rats have
about a 10-fold higher alveolar ventilation rate per unit body weight than humans
(Brown et al. 1997), which more than accounts for the observed increase in absorption.
6/22/2009
48
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Other experiments, such as closed-chamber gas uptake experiments or blood
concentration measurements following open-chamber (fixed concentration) experiments,
measure absorption indirectly but are consistent with significant retention. Closed-chamber
gas-uptake methods (Gargas et al. 1988) place laboratory animals or in vitro preparations into
sealed systems in which a known amount of TCE is injected to produce a predetermined
chamber concentration. As the animal retains a quantity of TCE inside its body, due to
metabolism, the closed-chamber concentration decreases with time when compared to the start of
exposure. Many different studies have made use of this technique in both rats and mice to
calculate total TCE metabolism (i.e., Andersen, 1987; Fisher et al., 1991; Simmons et al., 2002).
This inhalation technique is combined with PBPK modeling to calculate metabolic parameters,
and the results of these studies are consistent with rapid absorption of TCE via the respiratory
tract. Figure 3.1.1 shows and example from Simmons et al. (2002), in Long Evans rats, that
demonstrates an immediate decline in chamber concentrations of TCE indicating absorption,
with multiple initial concentrations needed for each metabolic calculation. At concentrations
below metabolic saturation, a secondary phase of uptake appears, after 1 hour from starting the
exposure, indicative of metabolism. At concentrations above 1000 ppm, metabolism appears
saturated, with time course curves having a flat phase after absorption. At intermediate
concentrations, between 100-1000 ppm, the secondary phase of uptake appears after distribution
as continued decreases in chamber concentration as metabolism proceeds. Using a combination
of experiments that include both metabolic linear decline and saturation obtained by using
different initial concentrations, both components of metabolism can be estimated from the gas
uptake curves, as shown in Figure 3.1.1.
6/22/2009
49
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
10000
E
Q.
Q.
X 1000
o
o
c
o
o
© 100
.Q
E
re
.c
o
111
O
•- 10
r+
++.
'++++
++++.
+ 3000 ppm
~ 1000 ppm
a 500 ppm
o 100 ppm
Aa
^Ay
Aa
~~~
AA/
~ ~l
'Aa
nonr
aa
~ ~~r
AAA
Aa
Dno
oo,
aaa
aAa
>o<>
A A/
Oo
Oo
Oo,
Oo
~ ~
aiA'Aa
«o
Ooo
OOOo
3 4
Time (hr)
Figure 3.1.1. Gas uptake data from closed-chamber exposure of rats to TCE. Symbols represent
measured chamber concentrations. Source: Simmons et al. (2002).
Several other studies in humans and rodents have measured blood concentrations of TCE
or metabolites and urinary excretion of metabolites during and after inhalation exposure
(e.g., Fisher et al. 1998; Filser and Bolt, 1979; Fisher et al. 1990; Fisher et al. 1991). While
qualitatively indicative of absorption, blood concentrations are also determined by metabolism,
distribution, and excretion, so comparisons between species may reflect similarities or
differences in any of the ADME processes.
3.1.3 Dermal
Skin membrane is believed to present a diffusional barrier for entrance of the chemical
into the body, and TCE absorption can be quantified using a permeability rate or permeability
constant, though not all studies performed such a calculation. Absorption through the skin has
been shown to be rapid by both vapor and liquid TCE contact with the skin. Human dermal
absorption of TCE vapors was investigated by Kezic et al. (2000). Human volunteers were
exposed to 3.18xl04 ppm around each enclosed arm for 20 minutes. Adsorption was found to be
6/22/2009
50
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
rapid (within 5 minutes), reaching a peak in exhaled breath around 30 minutes, with a calculated
dermal penetration rate averaging 0.049 cm/hour for TCE vapors.
With respect to dermal penetration of liquid TCE, Nakai et al. (1999) used surgically
removed skin samples exposed to TCE in aqueous solution in a chamber designed to measure the
difference between incoming and outgoing 14C-labelled TCE. The in vitro permeability constant
calculated by these researchers averaged 0.12 cm/hour. In vivo, Sato and Nakajima (1978)
exposed adult male volunteers dermally to liquid TCE for 30 minutes, with exhaled TCE
appearing at the initial sampling time of 5 minutes after start of exposure, with a maximum
observed at 15 minutes. In Kezic et al. (2001), human volunteers were exposed dermally for 3
minutes to neat liquid TCE, with TCE detected in exhaled breath at the first sampling point of 3
minutes, and maximal concentrations observed at 5 minutes. Skin irritancy was reported in all
subjects, which may have increased absorption. A dermal flux of 430 + 295 (mean + SE)
nmol/cm2/minute was reported in these subjects, suggesting high interindividual variability.
Another species where dermal absorption for TCE has been reported is in guinea pigs.
Jakobson et al. (1982) applied liquid TCE to the shaved backs of guinea pigs and reported peak
blood TCE levels at 20 minutes after initiation of exposure. Bogen et al., 1992 estimated
permeability constants for dermal absorption of TCE in hairless guinea pigs between 0.16 - 0.47
mL/cm2/hour across a range of concentrations (19 - 100,000 ppm).
6/22/2009
51
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.2 DISTRIBUTION AND BODY BURDEN
TCE crosses biological membranes and quickly results in rapid systemic distribution to
tissues—regardless of the route of exposure. In humans, in vivo studies of tissue distribution are
limited to tissues taken from autopsies following accidental poisonings or from surgical patients
exposed environmentally, so the level of exposure is typically unknown. Tissue levels reported
after autopsy show wide systemic distribution across all tested tissues, including the brain,
muscle, heart, kidney, lung, and liver (Ford et al., 1995; De Baere et al. 1997; Dehon et al. 2000;
Coopman et al. 2003). However, the reported levels themselves are difficult to interpret because
of the high exposures and differences in sampling protocols. In addition, human populations
exposed environmentally show detectable levels of TCE across different tissues, including the
liver, brain, kidney, and adipose tissues (McConnell et al. 1975; Pellizzari et al. 1982;
Kroneld 1989).
In addition, TCE vapors have been shown to cross the human placenta during childbirth
(Laham , 1970), with experiments in rats confirming this finding (Withey and Karpinski, 1985).
In particular, Laham, (1970) reported determinations of TCE concentrations in maternal and
fetal blood following administration of TCE vapors (concentration unreported) intermittently and
at birth (Table 3.2.1). TCE was present in all samples of fetal blood, with ratios of
concentrations in fetal:maternal blood ranging from approximately 0.5 to approximately 2. The
concentration ratio was less than 1.0 in six pairs, greater than 1 in 3 pairs, and approximately 1 in
1 pair; in general, higher ratios were observed at maternal concentrations below 2.25 mg/100
mL. Because no details of exposure concentration, duration, or time postexposure were given
for samples taken, these results are of minimal quantitative value, but they do demonstrate the
placental transfer of TCE in humans. Withey and Karpinski (1985) exposed pregnant rats to
TCE vapors (302, 1040, 1559, or 2088 ppm for 5 hours) on GD 17 and concentrations of TCE in
maternal and fetal blood were determined. At all concentrations, TCE concentration in fetal
blood was approximately one-third the concentration in corresponding maternal blood. Maternal
blood concentrations approximated 15, 60, 80, and 110 |ig/gram blood. When the position along
the uterine horn was examined, TCE concentrations in fetal blood decreased toward the tip of the
uterine horn.
6/22/2009
52
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.2.1. Concentrations of TCE in maternal and fetal blood at birth
TCE Concentration in
Blood (mg/100 mL)
Ratio of
Concentrations
FetalrMaternal
Maternal
Fetal
4.6
2.4
0.52
3.8
2.2
0.58
8
5
0.63
5.4
3.6
0.67
7.6
5.2
0.68
3.8
3.3
0.87
2
1.9
0.95
2.25
3
1.33
0.67
1
1.49
1.05
2
1.90
Source: Laham (1970).
TCE appears to also distribute to mammary tissues and is excreted in milk.
Pellizzari et al. (1982) conducted a survey of environmental contaminants in human milk using
samples from cities in the northeastern region of the United States and one in the southern
region. No details of times postpartum, milk lipid content, or TCE concentration in milk or
blood are reported, but TCE was detected in 8 milk samples taken from 42 lactating women.
Fisher et al. (1990) exposed lactating rats to 600-ppm TCE for 4 hours and collected milk
immediately following the cessation of exposure. TCE was clearly detectable in milk, and, from
a visual interpretation of the graphic display of their results, concentrations of TCE in milk
approximated 110 |ag/m L milk.
In rodents, detailed tissue distribution experiments have been performed using different
routes of administration (Savolainen et al. 1977; Pfaffenberger et al. 1980; Abbas and Fisher
1997; Greenberg et al. 1999; Simmons et al. 2002; Keys et al. 2003). Savolainen et al. (1977)
exposed adult male rats to 200-ppm TCE for 6 hours/day for a total of 5 days. Concentrations of
TCE in the blood, brain, liver, lung, and perirenal fat were measured 17 hours after cessation of
exposure on the fourth day and after 2, 3, 4, and 6 hours of exposure on the fifth day (Table
3.2.2). TCE appeared to be rapidly absorbed into blood and distributed to brain, liver, lungs, and
perirenal fat. TCE concentrations in these tissues reached near-maximal values within 2 hours of
initiation of exposure on the fifth day. Pfaffenberger et al. (1980) dosed rats by gavage with 1 or
10 mg TCE/kg/d in corn oil for 25 days to evaluate the distribution from serum to adipose tissue.
6/22/2009
53
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
During the exposure period, concentrations of TCE in serum were below the limit of detection (1
|ig/L) and were 280 and 20,000 ng per gram of fat in the 1 and 10 mg/day dose groups,
respectively. Abbas and Fisher (1997) and Greenberg et al. (1999) measured tissue
concentrations in the liver, lung, kidney, and fat of mice administered TCE by gavage (300-2000
mg/kg) and by inhalation exposure (100 or 600 ppm for 4 hours). In a study to investigate the
effects of TCE on neurological function, Simmons et al. (2002) conducted pharmacokinetic
experiments in rats exposed to 200, 2000, or 4000 ppm TCE vapors for 1 hour. Time-course
data were collected on blood, liver, brain, and fat. The data were used to develop a PBPK model
to explore the relationship between internal dose and neurological effect. Keys et al. (2003),
exposed groups of rats to TCE vapors of 50 or 500 ppm for 2 hours and sacrificed at different
time points during exposure. In addition to inhalation, this study also includes oral gavage and
intra-arterial dosing, with the following time course measured: liver, fat, muscle, blood, gastro-
intestinal (GI), brain, kidney, heart, lung, and spleen. These pharmacokinetic data were
presented with an updated PBPK model for all routes.
6/22/2009
54
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.2.2. Distribution of TCE to rat tissues3 following inhalation exposure (Savolainen
et al„ 1977)
Exposure
on 5th Day
Tissue (concentration in nmol/gram tissue)
Cerebrum
Cerebellum
Lung
Liver
Perirenal Fat
Blood
0b
0
0
0.08
0.04
0.23 + 0.09
0.35 + 0.1
2
9.9 + 2.7
11.7 + 4.2
4.9 0.3
3.6
65.9+ 1.2
7.5 + 1.6
3
7.3 + 2.2
8.8 + 2.1
5.5 + 1.4
5.5 + 1.7
69.3 + 3.3
6.6 + 0.9
4
7.2+ 1.7
7.6 + 0.5
5.8+ 1.1
2.5 + 1.4
69.5 + 6.3
6.0 + 0.2
6
7.4 + 2.1
9.5 + 2.5
5.6 + 0.5
2.4 + 0.2
75.4+ 14.9
6.8 + 1.2
aData presented as mean of 2 determinations + range.
bSample taken 17 hours following cessation of exposure on day 4.
Besides the route of administration, another important factor contributing to body
distribution is the individual solubility of the chemical in each organ, as measured by a partition
coefficient. For volatile compounds, partition coefficients are measured in vitro using the vial
equilibration technique to determine the ratio of concentrations between organ and air at
equilibrium. Table 3.2.3 reports values developed by several investigators from mouse, rat, and
human tissues. In humans, partition coefficients in the following tissues have been measured:
brain, fat, kidney, liver, lung, and muscle; but the organ having the highest TCE partition
coefficient is fat (63-70), while the lowest is the lung (0.5-1.7). The adipose tissue also has the
highest measured value in rodents, and is one of the considerations needed to be accounted for
when extrapolating across species. However, the rat adipose partition coefficient value is
smaller (23-36), when compared to humans, that is, TCE is less lipophilic in rats than humans.
For the mouse, the measured fat partition coefficient averages 36, ranging between rats and
humans. The value of the partition coefficient plays a role in distribution for each organ and is
computationally described in computer simulations using a PBPK model. Due to its high
lipophilicity in fat, as compared to blood, the adipose tissue behaves as a storage compartment
for this chemical, affecting the slower component of the chemical's distribution. For example
Monster et al. (1979) reported that, following repeated inhalation exposures to TCE, TCE
concentrations in expired breath post-exposure were highest for the subject with the greatest
amount of adipose tissue (adipose tissue mass ranged 3.5-fold among subjects). The inter-
subject range in TCE concentration in exhaled breath increased from approximately 2-fold at 20
hours to approximately 10-fold 140 hours post-exposure. Notably, they reported that this
difference was not due to differences in uptake, as body weight and lean body mass were most
closely associated with TCE retention. Thus, adipose tissue may play an important role in post-
exposure distribution, but does not affect its rapid absorption.
6/22/2009
55
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.2.3. Tissuerblood partition coefficient values for TCE
Species/
Tissue
TCE Partition Coefficient
References
Tissue:Blood
Tissue:Air
HUMAN
Brain
2.62
21.2
Fiserova-Bergerova et al. (1984)
Fat
63.8-70.2
583-674.4
Sato et al. (1977), Fiserova-Bergerova et al. (1984),
Fisher etal. (1998)
Kidney
1.3-1.8
12-14.7
Fiserova-Bergerova et al. (1984), Fisher et al. (1998)
Liver
3.6-5.9
29.4-54
Fiserova-Bergerova et al. (1984), Fisher et al. (1998)
Lung
0.48-1.7
4.4-13.6
Fiserova-Bergerova et al. (1984), Fisher et al. (1998)
Muscle
1.7-2.4
15.3-19.2
Fiserova-Bergerova et al. (1984), Fisher et al. (1998)
RAT
Brain
0.71-1.29
14.6-33.3
Sato et al. (1977), Simmons et al. (2002), Rodriguez
et al. (2007)
Fat
22.7-36.1
447-661
Gargas et al. 1989, Sato et al. (1977), Simmons et al.
(2002), Rodriguez et al. (2007), Fisher et al. (1989),
Koizumi (1989), Barton et al. (1995)
Heart
1.1
28.4
Sato etal. (1977)
Kidney
1.0-1.55
17.7-40
Sato et al. (1977), Barton et al. (1995), Rodriguez et
al. (2007)
Liver
1.03-2.43
20.5-62.7
Gargas et al. (1989), Sato et al. (1977), Simmons et al.
(2002), Rodriguez et al. (2007), Fisher et al. (1989),
Koizumi (1989), Barton et al. (1995)
Lung
1.03
26.6
Sato etal. (1977)
Muscle
0.46-0.84
6.9-21.6
Gargas et al. (1989), Sato et al. (1977), Simmons et al.
(2002), Rodriguez et al. (2007), Fisher et al. (1989),
Koizumi (1989), Barton et al. (1995)
Spleen
1.15
29.7
Sato etal. (1977)
Testis
0.71
18.3
Sato etal. (1977)
Milk
7.10
N.R.
Fisher etal. (1990)
MOUSE
Fat
36.4
578.8
Abbas and Fisher (1997)
Kidney
2.1
32.9
Abbas and Fisher (1997)
Liver
1.62
23.2
Fisher etal. (1991)
Lung
2.6
41.5
Abbas and Fisher (1997)
Muscle
2.36
37.5
Abbas and Fisher (1997)
2
3 Mahle et al. (2007) reported age-dependent differences in partition coefficients in rats,
4 (Table 3.2.4) that can have implications as to life-stage-dependent differences in tissue TCE
5 distribution. To investigate the potential impact of these differences, Rodriguez et al. (2007)
6 developed models for the postnatal day 10 rat pup; the adult and the aged rat, including
6/22/2009
56
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
age-specific tissue volumes and blood flows; and age-scaled metabolic constants. The models
predict similar uptake profiles for the adult and the aged rat during a 6-hour exposure to 500
ppm; uptake by the PND 10 rat was higher (Table 3.2.5). The effect was heavily dependent on
age-dependent changes in anatomical and physiological parameters (alveolar ventilation rates
and metabolic rates); age-dependent differences in partition coefficient values had minimal
impact on predicted differences in uptake.
Table 3.2.4. Age-dependence of tissuerair partition coefficients in rats
Age
Liver
Kidney
Fat
Muscle
Brain
PND 10 Male
22.1+2.3
15.2 + 1.3
398.7 + 89.2
43.9 + 11.0
11.0 + 0.6
PND10
Female
21.2+1.7
15.0 + 1.1
424.5 + 67.5
48.6 + 17.3
11.6+1.2
Adult Male
20.5 + 4.0
17.6 +3.9a
631.4 + 43.la
12.6 + 4.3
17.4 + 2.6
Aged Male
34.8 + 8.7^
19.9 + 3.4a
757.5 +48.3^
26.4 + 10.3^
25.0 + 2.0a'b
Source: Mahle et al. (2007).
Statistically significant (p < 0.05) difference between either the adult or aged partition
coefficient and the PND 10 male partition coefficient.
bStatistically significant (p < 0.05) difference between aged and adult partition coefficient.
Data are mean + standard deviation; n= 10, adult male and pooled male and female litters; 11,
aged males.
6/22/2009
57
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.2.5. Predicted maximal concentrations of TCE in rat blood following a 6-hour
inhalation exposure (Rodriguez et al., 2007)
Exposure Concentration
50 ppm
500 ppm
Predicted Peak
Predicted
Predicted Peak
Predicted
Age
Concentration
Time to
Concentration
Time to
(mg/L) in:a
Reach 90%
(mg/L) in:a
Reach 90%
Venous
Blood
Brain
of Steady
State
(hour)b
Venous
Blood
Brain
of Steady
State
(hour)b
PND 10
3.0
2.6
4.1
33
28
4.2
Adult
0.8
1.0
3.5
22
23
11.9
Aged
0.8
1.2
6.7
21
26
23.3
aDuring a 6 hour exposure.
bUnder continuous exposure.
Finally, TCE binding to tissues or cellular components within tissues can affect overall
pharmacokinetics. The binding of a chemical to plasma proteins, for example, affects the
availability of the chemical to other organs and the calculation of the total half-life. However,
most studies have evaluated binding using [14C]-radiolabeled TCE, from which one cannot
distinguish binding of TCE from binding of TCE metabolites. Nonetheless, several studies have
demonstrated binding of TCE-derived radiolabel to cellular components (Moslen et al. 1977;
Mazzullo et al. 1992). Bolt and Filser (1977) examined the total amount irreversibly bound to
tissues following 9-, 100-, and 1000-ppm exposures via inhalation in closed chambers. The
largest percent of in vivo radioactivity taken up occurred in the liver; albumin is the protein
favored for binding (Table 3.2.6). Bannerjee and van Duuren (1978) evaluated the in vitro
binding of TCE to microsomal proteins from the liver, lung, kidney, and stomachs in rats and
mice. In both rats and mice, radioactivity was similar in stomach and lung, but about 30% lower
in kidney and liver.
Table 3.2.6. Tissue distribution of TCE metabolites following inhalation exposure
% of Radioactivity Taken Up/g Tissue
TCE =
9 ppm,
TCE =
100 ppm,
TCE = 1000 ppm,
Tissue*
n
=4
n
=4
n
=3
Total
Irreversibly
Total
Irreversibly
Total
Irreversibly
Metabolites
Bound
Metabolites
Bound
Metabolites
Bound
6/22/2009
58
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Lung
0.23 ±0.026
0.06 ±0.002
0.24 ±0.025
0.06 ±0.006
0.22 ±0.055
0.1 ±0.003
Liver
0.77 + 0.059
0.28 + 0.027
0.68 ±0.073
0.27 ±0.019
0.88 ±0.046
0.48 ±0.020
Spleen
0.14 + 0.015
0.05 ±0.002
0.15 ±0.001
0.05 ±0.004
0.15 ±0.006
0.08 ±0.003
Kidney
0.37 + 0.005
0.09 + 0.007
0.40 ±0.029
0.09 ±0.007
0.39 ±0.045
0.14 ± 0.016
Small
Intestine
0.41 + 0.058
0.05 ±0.010
0.38 ±0.062
0.07 ±0.008
0.28 ±0.015
0.09 ±0.015
Muscle
0.11 + 0.005
0.014 ±0.001
0.11 ±0.013
0.012 ±0.001
0.10 ± 0.011
0.027 ±0.003
Source: Bolt and Filser (1977).
*Male Wistar rats, 250g.
n = number of animals.
Values shown are means + SD.
Based on studies of the effects of metabolizing enzyme induction on binding, there is
some evidence that a major contributor to the observed binding is from TCE metabolites rather
than from TCE itself. Dekant et al. (1986a) studied the effect of enzyme modulation on the
binding of radiolabel from [14C]-TCE by comparing tissue binding after administration of 200
mg/kg via oral gavage in corn oil between control (naive) rats and rats pretreated with
phenobarbital (a known inducer of CYP2B family) or arochlor 1254 (a known inducer of both
CYP1A and CYP2B families of isoenzymes) (Table 3.2.7). The results indicate that induction of
total cytochromes P-450 content by 3- to 4-fold resulted in nearly 10-fold increase in
radioactivity (decays per minute; DPM) bound in liver and kidney. By contrast, Mazzullo et al.
(1992) reported that, phenobarbital pretreatment did not result in consistent or marked alterations
of in vivo binding of radiolabel to DNA, RNA, or protein in rats and mice at 22 hours after an ip
injection of [14C]-TCE. On the other hand, in vitro experiments by Mazzullo et al. (1992)
reported reduction of TCE-radiolabel binding to calf thymus DNA with introduction of a CYP
inhibitor into incubations containing rat liver microsomal protein. Moreover, increase/decrease
of GSH levels in incubations containing lung cytosolic protein led to a parallel increase/decrease
in TCE-radiolabel binding to calf thymus DNA.
Table 3.2.7. Binding of 14C from 14C-TCE in rat liver and kidney at 72 hrs. after oral
administration of 200 mg/kg [14C]-TCE (Dekant et al., 1986a)
Tissue
DPM/Gram Tissue
Untreated
Phenobarbital
Arochlor 1254
Liver
850+ 100
9300+ 1100
8700+ 1000
Kidney
680+ 100
5700 + 900
7300 + 800
6/22/2009
59
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.3 METABOLISM
This section focuses on both in vivo and in vitro studies of the biotransformation of
trichloroethylene, identifying metabolites that are deemed significant for assessing toxicity and
carcinogenicity. In addition, metabolism studies may be used to evaluate the flux of parent
compound through the known metabolic pathways. Sex-, species-, and interindividual
differences in the metabolism of TCE are discussed, as are factors that possibly contribute to this
variability. Additional discussion of variability and susceptibility is presented in Section 4.9.
3.3.1 Introduction
The metabolism of TCE has been studied mostly in mice, rats, and humans and has been
extensively reviewed (U.S. EPA, 1985, 2001; Lash et al., 2000a; IARC, 1995). It is now well
accepted that TCE is metabolized in laboratory animals and in humans through at least two
distinct pathways: (1) oxidative metabolism via the cytochrome P450 mixed-function oxidase
system and (2) glutathione (GSH) conjugation followed by subsequent further biotransformation
and processing, either through the cysteine conjugate beta lyase pathway or by other enzymes
(Lash et al., 2000b). While the flux through the conjugative pathway is less, quantitatively, than
the flux through oxidation (Bloemen et al., 2001), GSH conjugation is an important route
toxicologically, giving rise to relatively potent toxic biotransformation products
(Elfarra et al., 1986a,b).
Information about metabolism is important because, as discussed extensively in
Chapter 4, certain metabolites are thought to cause one or more of the same acute and chronic
toxic effects, including carcinogenicity, as TCE. Thus, in many of these cases, the toxicity of
TCE is generally considered to reside primarily in its metabolites rather than in the parent
compound itself.
3.3.2 Extent of Metabolism
TCE is extensively metabolized in animals and humans. The most comprehensive
mass-balance studies are in mice and rats (Dekant et al., 1984; Dekant et al., 1986a,b; Green and
Prout 1985; Prout et al., 1985) in which [14C]-TCE is administered by oral gavage at doses of 2
to 2000 mg/kg, the data from which are summarized in Figure 3.3.1 and Figure 3.3.2. In both
mice and rats, regardless of sex and strain, there is a general trend of increasing exhalation of
unchanged TCE with dose, suggesting saturation of a metabolic pathway. The increase is
smaller in mice (from 1-6% to 10-18%) than in rats (from 1-3% to 43—78%), suggesting greater
overall metabolic capacity in mice. The dose at which apparent saturation occurs appears to be
6/22/2009
60
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
more sex- or strain-dependent in mice than in rats. In particular, the marked increase in exhaled
TCE occurred between 20 and 200 mg/kg in female NMRI mice, between 500 and 1000 mg/kg
in B6C3F1 mice, and between 10 and 500 mg/kg in male Swiss-Webster mice. However,
because only one study is available in each strain, inter-lot or inter-individual variability might
also contribute to the observed differences. In rats, all three strains tested showed marked
increase in unchanged TCE exhaled between 20 and 200 mg/kg or 10 and 500 mg/kg.
Recovered urine, the other major source of excretion, had mainly trichloroacetic acid (TCA),
trichloroethanol (TCOH), and trichloroethanol-glucuronide conjugate (TCOG), but revealed no
detectable TCE. The source of radioactivity in feces was not analyzed, but it is presumed not to
include substantial TCE given the complete absorption expected from the corn oil vehicle.
Therefore, at all doses tested in mice, and at doses <200 mg/kg in rats, the majority of orally
administered TCE is metabolized. Pretreatment of rats with P450 inducers prior to a 200 mg/kg
dose did not change the pattern of recovery, but it did increase the amount recovered in urine by
10-15%, with a corresponding decrease in the amount of exhaled unchanged TCE (Dekant et al.,
1986a).
6/22/2009
61
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
100
90
80
70
60
50
T3
O
5
>
o
o
o
J
¦>
'¦*->
o
rc
o
.2 40
¦a
ro
30
20
10
M
2 mg/kg
20 200
mg/kg mg/kg
F/NMRI
10
mg/kg
500
mg/kg
1000 2000
mg/kg mg/kg
M/B6C3F1
10 500 1000
mg/kg mg/kg mg/kg
M/Swiss-Webster
Mouse Sex/Strain and Dose
Figure 3.3.1. Disposition of [14C]-TCE administered by oral gavage in mice (Dekant et al.,
1984; Dekant et al., 1986a; Green and Prout, 1985; Prout et al., 1985).
6/22/2009
62
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
100
90
80
70
60
£ 50
¦a
£
>
o
o
a>
J
'>
,2 40
¦a
nj
30
20
10
0
2 mg/kg 20 200
mg/kg mg/kg
F/Wistar
1
10 500 1000 2000
mg/kg mg/kg mg/kg mg/kg
M/Osborne-Mendel
10 500 1000
mg/kg mg/kg mg/kg
M/Alderley-Park Wistar
Rat Sex/Strain and Dose
Figure 3.3.2. Disposition of [14C]-TCE administered by oral gavage in rats (Dekant et al., 1984;
Dekant et al., 1986a; Green and Prout, 1985; Prout et al., 1985).
6/22/2009
63
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Comprehensive mass balance studies are not available in humans, but several studies
have measured or estimated recovery of TCE in exhaled breath and/or TCA and TCOH in urine
following controlled inhalation exposures to TCE (Monster et al., 1976; Opdam, 1989; Soucek
and Vlachova, 1960). Opdam (1989) only measured exhaled breath, and estimated that, on
average, 15-20% of TCE uptake (retained dose) was exhaled after exposure to 5.8-38 ppm for
29-62 minutes. Soucek and Vlachova (1960) and Bartonicek (1962) did not measure exhaled
breath but did report 69-73% of the retained dose excreted in urine as TCA and TCOH following
exposure to 93-194 ppm (500-1043 mg/m3) for 5 hours. Soucek and Vlachova (1960)
additionally reported 4% of the retained dose excreted in urine as monochloroacetic acid (MCA).
Monster et al. (1976) reported that an average of 10% of the retained TCE dose was eliminated
unchanged following 6 hour exposures to 70-140 ppm (376-752 mg/m3) TCE, along with an
average of 57% of the retained dose excreted in urine as TCA and free or conjugated TCOH.
The differences among these studies may reflect a combination of inter-individual variability and
errors due to the difficulty in precisely estimating dose in inhalation studies, but in all cases less
than 20% of the retained dose was exhaled unchanged and greater than 50% was excreted in
urine as TCA and TCOH. Therefore, it is clear that TCE is extensively metabolized in humans.
Unlike the rodent studies, no saturation was evident in any of these human recovery studies even
though the metabolic capacity may not have been saturated at the exposure levels that were
tested.
3.3.3 Pathways of Metabolism
As mentioned in Section 3.3.1, TCE metabolism in animals and humans has been
observed to occur via two major pathways: P450-mediated oxidation and GSH conjugation.
Products of the initial oxidation or conjugation step are further metabolized to a number of other
metabolites. For P450 oxidation, all steps of metabolism occur primarily in the liver, although
limited oxidation of TCE has been observed in the lungs of mice, as discussed below. The GSH
conjugation pathway also begins predominantly in the liver, but toxicologically significant
metabolic steps occur extrahepatically—particularly in the kidney (Lash et al., 1995, 1998,
1999b, 2006). The mass-balance studies cited above found that at exposures below the onset of
saturation, >50%) of TCE intake is excreted in urine as oxidative metabolites (primarily as TCA
and TCOH), so TCE oxidation is generally greater than TCE conjugation. This is discussed in
detail in Section 3.3.3.3.
3.3.3.1 Cytochrome P450-Dependent Oxidation
Oxidative metabolism by the cytochrome P450, or CYP-dependent, pathway is
quantitatively the major route of TCE biotransformation (U.S. EPA, 1985; IARC, 1995;
6/22/2009
64
-------
1
2
3
4
5
6
7
8
9
10
11
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Lash et al., 2000a,b). The pathway is operative in humans and rodents and leads to several
metabolic products, some of which are known to cause toxicity and carcinogenicity (U.S. EPA,
1985; IARC, 1995). Although several of the metabolites in this pathway have been clearly
identified, others are speculative or questionable. Figure 3.3.3 depicts the overall scheme of TCE
P450 metabolism.
cw
H (TCE) CI
P450
ohch2 —
(W-(Hydroxyacetyl)- N (CH2)2OH
aminoethanol) |-|
TCE-O-P450
OH
CI2CH
CI3C
CI3C
(CHL)
(DCAC)
ALDH
7\DH or
P450
CI3C
CI3C
P450
ci2ch
(TCOH) QH
(DCA) OH
GST-
zeta-
EHR
UGT
CI3C
co2
CI ch2
(TCOG) o-gluc
(Glyoxylic OH
acid)
HO (OA) OH
(MCA) OH
Figure 3.3.3. Scheme for the oxidative metabolism of TCE.
Adapted from Lash et al. (2000a,b); Clewell et al. (2000); Cummings et al. (2001);
Forkert et al. (2006); Tong et al. (1998).
6/22/2009
65
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
In brief, TCE oxidation via P450, primarily CYP2E1 (Guengerich et al., 1991), yields an
oxygenated TCE-P450 intermediate and TCE oxide. The TCE-P450 complex is a transition state
that goes on to form chloral (CHL). In the presence of water, chloral rapidly equilibrates with
chloral hydrate (CH), which undergoes reduction and oxidation by alcohol dehydrogenase and
aldehyde dehydrogenase or aldehyde oxidase to form TCOH and trichloroacetic acid (TCA),
respectively (Miller and Guengerich 1983, Green and Prout, 1985; Dekant et al., 1986a). Table
3.3.1 summarizes available in vitro measurements of TCE oxidation, as assessed by the
formation of CH, TCOH, and TCA. Glucuronidation of TCOH forms TCOG, which is readily
excreted in urine. Alternatively, TCOG can be excreted in bile and passed to the small intestine
where it is hydrolyzed back to TCOH and reabsorbed (Bull, 2000). TCA is poorly metabolized
but may undergo dechlorination to form dichloroacetic acid (DCA). However, TCA is
predominantly excreted in urine, albeit at a relatively slow rate as compared to TCOG. Like the
TCE-P450 complex, TCE oxide also seems to be a transient metabolite. Recent data suggest that
it is transformed to dichloroactyl chloride, which subsequently decomposes to form DCA (Cai
and Guengerich, 1999). As shown in Figure 3.3.3, several other metabolites, including oxalic
acid and N-(hydroxyacetyl) aminoethanol, may form from the TCE oxide or the TCE-O-P450
intermediate and have been detected in the urine of rodents and humans following TCE
exposure. Pulmonary excretion of carbon dioxide (CO2) has been identified in exhaled breath
from rodents exposed to 14C-labeled TCE and is thought to arise from metabolism of DCA. The
following sections provide details as to pathways of TCE oxidation, including discussion of
inter- and intraspecies differences in metabolism.
6/22/2009
66
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.3.1. In vitro TCE oxidative metabolism in hepatocytes and microsomal fractions
In Vitro
System
Km
V
* max
1000 X
V /K
' max' AVm
Source
jiM in
Medium
nmol TCE
oxidized/min/
mg MSP* or
106
hepatocytes
Human
hepatocytes
210+159
(45-403)
0.268 + 0.215
(0.101-0.691)
2.45 + 2.28
(0.46-5.57)
Lipscomb et al. (1998a)
Human liver
microsomal
protein
16.7 + 2.45
(13.3-19.7)
1.246 + 0.805
(0.490-3.309)
74.1 +44.1
(38.9-176)
Lipscomb et al. (1997) (Low Km)
30.9 + 3.3
(27.0-36.3)
1.442 + 0.464
(0.890-2.353)
47.0+16.0
(30.1-81.4)
Lipscomb et al. (1997) (Mid Km)
51.1+3.77
(46.7-55.7)
2.773 + 0.577
(2.078-3.455)
54.9+14.1
(37.3-69.1)
Lipscomb et al. (1997) (High Km)
24.6
1.44
58.5
Lipscomb et al. (1998b) (pooled)
12 + 3
(9-14)
0.52 + 0.17
(0.37-0.79)
48 + 23
(26-79)
Elfarra et al. (1998) (males, high
affinity)
26+17
(13-45)
0.33 + 0.15
(0.19-0.48)
15 + 10
(11-29)
Elfarra et al. (1998) (females,
high affinity)
Rat liver
microsomal
protein
55.5
4.826
87.0
Lipscomb et al. (1998b) (pooled)
72 + 82
0.96 + 0.65
24 + 21
Elfarra et al. (1998) (males, high
affinity)
42 + 21
2.91 + 0.71
80 + 34
Elfarra et al. (1998) (females,
high affinity)
Rat kidney
microsomal
protein
940
0.154
0.164
Cummings et al. (2001)
Mouse liver
microsomal
protein
35.4
5.425
153
Lipscomb et al. (1998b) (pooled)
378 + 414
8.6 + 4.5
42 + 29
Elfarra et al. (1998) (males)
161+29
26.06 + 7.29
163 + 37
Elfarra et al. (1998) (females)
2 * MSP = Microsomal protein.
3 Notes: Results presented as mean + standard deviation (min-max). Km for human hepatocytes
4 converted from ppm in headspace to [iM in medium using reported hepatocyte:air partition
5 coefficient (Lipscomb et al., 1998a).
6/22/2009
67
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.3.3.1.1 Formation of trichloroethylene oxide
In previous studies of halogenated alkene metabolism, the initial step was the generation
of a reactive epoxide (Anders and Jackobson, 1985). Early studies in anesthetized human
patients (Powell, 1945), dogs (Butler, 1949), and later reviews (e.g., Goeptar et al., 1995) suggest
that the TCE epoxide may be the initial reaction product of TCE oxidation.
Epoxides can form acyl chlorides or aldehydes, which can then form aldehydes,
carboxylic acids, or alcohols, respectively. Thus, the appearance of chloral hydrate (CH), TCA,
and trichloroethanol (TCOH) as the primary metabolites was considered consistent with the
oxidation of TCE to the epoxide intermediate (Powell, 1945; Butler, 1949). Following in vivo
exposures to 1,1-dichloroethylene, a halocarbon very similar in structure to TCE, mouse liver
cytosol and microsomes and lung Clara cells exhibited extensive P450-mediated epoxide
formation (Forkert, 1999a, b; Forkert et al., 1999; Dowsley et al., 1996). Indeed, TCE oxide
inhibits purified CYP2E1 activity (Cai and Guengerich, 2001) similarly to TCE inhibition of
CYP2E1 in human liver microsomes (Lipscomb et al., 1997).
Conversely, cases have been made against TCE oxide as an obligate intermediate. Using
liver microsomes and reconstituted P450 systems (Miller and Guengerich, 1983, 1982) or
isolated rat hepatocytes (Miller and Guengerich, 1983), it has been suggested that chlorine
migration and generation of a TCE-O-P450 complex (via the heme oxygen) would better explain
the observed destruction of the P450 heme, an outcome not likely to be epoxide-mediated.
Miller and Guengerich (1982) found CYP2E1 to generate an epoxide but argued that the
subsequent production of chloral was not likely related to the epoxide. Green and Prout (1985)
argued against epoxide (free form) formation in vivo in mice and rats, suggesting that the
expected predominant metabolites would be CO, CO2, monochloroacetic acid (MCA), and
dichloroacetic acid (DCA), rather than the observed predominant appearance of TCA and TCOH
and its glucuronide (TCOG).
It appears likely that both a TCE-O-P450 complex and a TCE oxide are formed, resulting
in both CH and dichloroacetyl chloride, respectively, though it appears that the former
predominates. In particular, it has been shown that dichloroacetyl chloride can be generated
from TCE oxide, dichloracetyl chloride can be trapped with lysine (Cai and Guengerich, 1999),
and that dichloracetyl-lysine adducts are formed in vivo (Forkert et al., 2006). Together, these
data strongly suggest TCE oxide as an intermediate metabolite, albeit short-lived, from TCE
oxidation in vivo.
6/22/2009
68
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.3.3.1.2 Formation of CH, TCOH and TCA
CH (in equilibrium with chloral) is a major oxidative metabolite produced from TCE as
has been shown in numerous in vitro systems, including human liver microsomes and purified
P450 CYP2E1 (Guengerich et al., 1991) as well as recombinant rat, mouse, and human P450s
including CYP2E1 (Forkert et al., 2005). However, in rats and humans, in vivo circulating CH is
generally absent from blood following TCE exposure. In mice, CH is detectable in blood and
tissues but is rapidly cleared from systemic circulation (Abbas and Fisher, 1997). The low
systemic levels of CH are because of its rapid transformation to other metabolites.
CH is further metabolized predominantly to TCOH (Sellers et al., 1972), a reaction
thought to be catalyzed by alcohol dehydrogenase (Shultz and Weiner, 1979) and/or CYP2E1
(Ni et al., 1996). The role for alcohol dehydrogenase was suggested by the observation that
ethanol inhibited CH reduction to TCOH (Larson and Bull, 1989; Miiller et al., 1975; Sellers et
al., 1972). For instance, Sellers et al. (1972) reported that coexposure of humans, to ethanol and
CH resulted in a higher percentage of urinary TCOH (24% of CH metabolites) compared to TCA
(19%). When ethanol was absent, 10 and 11% of CH was metabolized to TCOH and TCA,
respectively. However, because ethanol can be oxidized by both alcohol dehydrogenase and
CYP2E1, there is some ambiguity as to whether these observations involve competition with one
or the other of these enzymes. For instance, Ni et al. (1996) reported that CYP2E1 expression
was necessary for metabolism of CH to mutagenic metabolites in a human lymphoblastoid cell
line, suggesting a role for CYP2E1. Furthermore, Ni et al. (1996) reported that cotreatment of
mice with CH and pyrazole, a specific CYP2E1 inducer, resulted in enhanced liver microsomal
lipid peroxidation, while treatment with DPEA, an inhibitor of CYP2E1, suppressed lipid
peroxidation, suggesting CYP2E1 as a primary enzyme for CH metabolism in this system.
Lipscomb et al. (1996) suggested that two enzymes are likely responsible for CH reduction to
TCOH based on observation of bi-phasic metabolism for this pathway in mouse liver
microsomes. This behavior has also been observed in mouse liver cytosol, but was not observed
in rat or human liver microsomes. Moreover, CH metabolism to TCOH increased significantly
both in the presence of NADH in the 700 x g supernatant of mouse, rat, and human liver
homogenate as well as with the addition of NADPH in human samples, suggesting two enzymes
may be involved (Lipscomb et al., 1996).
TCOH formed from CH is available for oxidation to TCA (see below) or glucuronidation
via UDP-glucuronyltransferase to TCOG, which is excreted in urine or in bile (Stenner et al.,
1997). Biliary TCOG is hydrolyzed in the gut and available for reabsorption to the liver as
TCOH, where it can be glucuronidated again or metabolized to TCA. This enterohepatic
circulation appears to play a significant role in the generation of TCA from TCOH and in the
observed lengthy residence time of this metabolite, compared to TCE. Using jugular-, duodenal -
6/22/2009
69
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
, and bile duct-cannulated rats, Stenner et al. (1997) showed that enterohepatic circulation of
TCOH from the gut back to the liver and subsequent oxidation to TCA was responsible for 76%
of TCA measured in the systemic blood.
Both CH and TCOH can be oxidized to TCA, and has been demonstrated in vivo in mice
(Larson and Bull, 1992a; Dekant et al., 1986a; Green and Prout, 1985), rats (Stenner et al., 1997;
Pravecek et al., 1996; Templin et al., 1995b; Larson and Bull, 1992a; Dekant et al., 1986a; Green
and Prout, 1985), dogs (Templin et al., 1995a), and humans (Sellers et al., 1978). Urinary
metabolite data in mice and rats exposed to 200 mg/kg TCE (Larson and Bull, 1992a;
Dekant et al., 1986a) and humans following oral CH exposure (Sellers et al., 1978) show greater
TCOH production relative to TCA production. However, because of the much longer urinary
half-life in humans of TCA relative to TCOH, the total amount of TCA excreted may be similar
to TCOH (Monster et al., 1976; Fisher et al., 1998). This is thought to be primarily due to
conversion of TCOH to TCA, either directly or via "back-conversion" of TCOH to CH, rather
than due to the initial formation of TCA from CH (Marshall and Owens, 1955).
In vitro data are also consistent with CH oxidation to TCA being much less than CH
reduction to TCOH. For instance, Lipscomb et al. (1996) reported 1832-fold differences in Km
values and 10-195-fold differences in clearance efficiency (Vmax/Km) for TCOH and TCA in all
three species (Table 3.3.2). Clearance efficiency of CH to TCA in mice is very similar to
humans but is 13-fold higher than rats. Interestingly, Bronley-DeLancey et al. (2006) recently
reported that similar amounts of TCOH and TCA were generated from CH using cryopreserved
human hepatocytes. However, the intersample variation was extremely high, with measured
Vmax ranging from 8-fold greater TCOH to 5-fold greater TCA and clearance (Vmax/Km) ranging
from 13-fold greater TCOH to 17-fold greater TCA. Moreover, because a comparison with fresh
hepatocytes or microsomal protein was not made, it is not clear to what extent these differences
are due to population heterogeneity or experimental procedures.
6/22/2009
70
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.3.2. In vitro kinetics of trichloroethanol and trichloroacetic acid formation from
chloral hydrate in rat, mouse, and human liver homogenates
Species
TCOH
TCA
Kma
V b
~ max
Vmax/Kmc
Kma
V b
~ max
Vmax/Kmc
Rat
0.52
24.3
46.7
16.4
4
0.24
Moused
0.19
11.3
59.5
3.5
10.6
3.0
High affinity
0.12
6.3
52.5
nae
na
na
Low affinity
0.51
6.1
12.0
na
na
na
Human
1.34
34.7
25.9
23.9
65.2
2.7
aKm presented as mM CH in solution.
bVmax presented as nmoles/mg supernatant protein/min.
cClearance efficiency represented by Vmax/Km.
dMouse kinetic parameters derived for observations over the entire range of CH exposure as well
as discrete, bi-phasic regions for CH concentrations below (high affinity) and above (low
affinity) 1.0 mM.
ena = not applicable.
Source: Lipscomb et al. (1996).
The metabolism of CH to TCA and TCOH involves several enzymes including CYP2E1,
alcohol dehydrogenase, and aldehyde dehodrogenase enzymes (Guengerich et al., 1991; Miller
and Guengerich, 1983; Ni et al., 1996; Shultz and Weiner, 1979; Wang et al., 1993). Because
these enzymes have preferred cofactors (NADPH, NADH, NAD+), cellular cofactor ratio and
redox status of the liver may have an impact on the preferred pathway (Kawamoto et al., 1988;
Lipscomb et al., 1996).
3.3.3.1.3 Formation of DCA and other products
As discussed above, DCA could hypothetically be formed via multiple pathways. The
work reviewed by Guengerich (2004) has suggested that one source of DCA may be through a
TCE oxide intermediary. Miller and Guengerich (1983) report evidence of formation of the
epoxide, and Cai and Guengerich (1999) report that a significant amount (about 35%) of DCA is
formed from aqueous decomposition of TCE oxide via hydrolysis in an almost pH-independent
manner. Because this reaction forming DCA from TCE oxide is a chemical process rather than a
process mediated by enzymes, and because evidence suggests that some epoxide was formed
from TCE oxidation, Guengerich (2004) notes that DCA would be an expected product of TCE
oxidation (see also Yoshioka et al. 2002). Alternatively, dechlorination of TCA and oxidation of
TCOH have been proposed as sources of DCA (Lash et al., 2000a). Merdink et al. (2000)
6/22/2009
71
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
investigated dechlorination of TCA and reported trapping a DCA radical with the spin-trapping
agent phenyl-tert-butyl nitroxide, identified by gas chromatography/mass spectroscopy, in both a
chemical Fenton system and rodent microsomal incubations with TCA as substrate. Dose-
dependent catalysis of TCA to DCA was observed in cultured microflora from B6C3F1 mice
(Moghaddam et al., 1996). However, while antibiotic-treated mice lost the ability to produce
DCA in the gut, plasma DCA levels were unaffected by antibiotic treatment, suggesting that the
primary site of murine DCA production is other than the gut (Moghaddam et al., 1997).
However, direct evidence for DCA formation from TCE exposure remains equivocal. In
vitro studies in human and animal systems have demonstrated very little DCA production in the
liver (James et al., 1997). In vivo, DCA was detected in the blood of mice (Templin et al., 1993;
Larson and Bull, 1992a) and humans (Fisher et al., 1998; but not detected by
Bloemen et al., 2001) and in the urine of rats and mice (Larson and Bull, 1992b) exposed to TCE
by aqueous oral gavage. However, the use of strong acids in the analytical methodology
produces artifactual conversion of TCA to DCA in mouse blood (Ketcha et al., 1996). This
method may have resulted in the appearance of DCA as an artifact in human plasma (Fisher et
al., 1998) and mouse blood in vivo (Templin et al., 1995b). Evidence for the artifact is suggested
by DCA areas under the curve (AUCs) that were larger than would be expected from the
available TCA (Templin et al., 1995a). After the discovery of these analytical issues, Merdink et
al. (1998) reevaluated the formation of DCA from TCE, TCOH, and TCA in mice, with
particular focus on the hypothesis that DCA is formed from dechlorination of TCA. They were
unable to detect blood DCA in naive mice after administration of TCE, TCOH, or TCA. Low
levels of DCA were detected in the blood of children administered therapeutic doses of CH
(Henderson et al., 1997), suggesting TCA or TCOH as the source of DCA. Oral TCE exposure
in rats and dogs failed to produce detectable levels of DCA (Templin et al., 1995a).
Another difficulty in assessing the formation of DCA is its rapid metabolism at low
exposure levels. Degradation of DCA is mediated by GST-zeta (Saghir and Schultz, 2002;
Tong et al., 1998), apparently occurring primarily in the hepatic cytosol. DCA metabolism
results in suicide inhibition of the enzyme, evidenced by decreased DCA metabolism in DCA-
treated animals (Gonzalez-Leon et al., 1999) and humans (Shroads et al., 2008) and loss of DCA
metabolic activity and enzymatic protein in liver samples from treated animals (Schultz et al.,
2002). This effect has been noted in young mice exposed to DCA in drinking water at doses
approximating 120 mg/kg-day (Schultz et al., 2002). The experimental data and
pharmacokinetic model simulations of several investigators (Jia et al., 2006; Keys et al., 2004; Li
et al., 2008; Merdink et al., 1998; Shroads et al., 2008) suggest that several factors prevent the
accumulation of measurable amounts of DCA: 1) its formation as a short-lived intermediate
metabolite, and 2) its rapid elimination relative to its formation from TCA. While DCA
6/22/2009
72
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
elimination rates appear approximately one order of magnitude higher in rats and mice than in
humans (James et al., 1997) (Table 3.3.3), they still may be rapid enough so that even if DCA
were formed in humans, it would be metabolized too quickly to appear in detectable quantities in
blood.
Table 3.3.3. In vitro kinetics of DCA metabolism in hepatic cytosol of mice, rats, and
humans
Species
V max
(nmol/min/mg protein)
Km
(HM)
Vmax/Km
Mouse
13.1
350
37.4
Rat
11.6
280
41.4
Human
0.37
71
5.2
Source: James et al. (1997).
A number of other metabolites, such as oxalic acid (OA), MCA, glycolic acid, and
glyoxylic acid, are formed from DCA (Lash et al., 2000a; Saghir and Schultz, 2002). Unlike
other oxidative metabolites of TCE, DCA appears to be metabolized primarily via hepatic
cytosolic proteins. Since P450 activity resides almost exclusively in the microsomal and
mitochondrial cell fractions, DCA metabolism appears to be independent of P450. Rodent
microsomal and mitochondrial metabolism of DCA was measured to be <10% of cytosolic
metabolism (Lipscomb et al., 1995). DCA in the liver cytosol from rats and humans is
transformed to glyoxylic acid via a GSH-dependent pathway (James et al., 1997). In rats, the Km
for GSH was 0.075 mM with a Vmax for glyoxylic acid formation of 1.7 nmol/mg protein/minute.
While this pathway may not involve GST (as evidenced by very low GST activity in this study),
Tong et al. (1998) showed GST-zeta, purified from rat liver, to be involved in metabolizing DCA
to glyoxylic acid, with a Vmax of 1334 nmol/mg protein/minute and Km of 71.4 |iM for glyoxylic
acid formation and a GSH Km of 59 [xM.
3.3.3.1.4 Tissue distribution of oxidative metabolism and metabolites
Oxidative metabolism of TCE, irrespective of the route of administration, occurs
predominantly in the liver, but TCE metabolism via the P450 (CYP) system also occurs at other
sites because CYP isoforms are present to some degree in most tissues of the body. For
example, both the lung and kidneys exhibit cytochrome P450 enzyme activities
(Green et al., 1997a,b; Forkert et al., 2005; Cummings et al., 2001). Green et al. (1997b)
detected TCE oxidation to chloral in microsomal fractions of whole-lung homogenates from
mice, rats, and humans, with the activity in mice the greatest and in humans the least. The rates
6/22/2009
73
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
were slower than in the liver (which also has a higher microsomal protein content as well as
greater tissue mass) by 1.8-, 10-, and >10-fold in mice, rats, and humans, respectively. While
qualitatively informative, these rates were determined at a single concentration of about 1 mM
TCE. A full kinetic analysis was not performed, so clearance and maximal rates of metabolism
could not be determined. The situation is similar with the kidney where Cummings et al. (2001)
performed a full kinetic analysis using kidney microsomes, and found clearance rates (Vmax/Km)
for oxidation were more than 100-fold smaller than average rates that were found in the liver
(Table 3.3.1). In humans, Cummings and Lash (2000) reported detecting oxidation of TCE in
only one of 4 samples, and only at the highest tested concentration of 2 mM, with a rate of 0.13
nmol/min/mg protein. This rate contrasts with the Vmax values for human liver microsomal
protein of 0.19-3.5 nmol/min/mg protein reported in various experiments (Table 3.3.1, above).
Thus, the lower rates of oxidation combined with lower microsomal protein content as well as
the relatively smaller organ mass mean that TCE oxidation in the lung and kidney is not expected
to contribute substantially to the total oxidation of TCE. However, while quantitatively minor in
terms of total systemic metabolism, extra-hepatic oxidation of TCE may play an important role
for generation of toxic metabolites in situ. The roles of local metabolism in kidney and lung
toxicity are discussed in detail in Sections 4.3 and 4.6, respectively.
With respect to further metabolism beyond oxidation of TCE, CH has been shown to be
metabolized to TCA and TCOH in lysed whole blood of mice and rats and fractionated human
blood (Lipscomb et al., 1996) (Table 3.3.4). TCOH production is similar in mice and rats and is
approximately 2-fold higher in rodents than in human blood. However, TCA formation in
human blood is 2- or 3-fold higher than in mouse or rat blood, respectively. In human blood,
TCA is formed only in the erythrocytes. TCOH formation occurs in both plasma and
erythrocytes, but 4-fold more TCOH is found in plasma than in an equal volume of packed
erythrocytes. While blood metabolism of CH may contribute further to its low circulating levels
in vivo., the metabolic capacity of blood (and kidney) may be substantially lower than liver.
Regardless, any CH reaching the blood may be rapidly metabolized to TCA and TCOH.
Table 3.3.4. TCOH and TCA formed from CH in vitro in lysed whole blood of rats and
mice or fractionated blood of humans (nmoles formed in 400 jiL samples over 30 minutes)
Rat
Mouse
Human
Erythrocytes
Plasma
TCOH
45.4 + 4.9
46.7 + 1.0
15.7+1.4
4.48 + 0.2
TCA
0.14 + 0.2
0.21+0.3
0.42 + 0.0
not detected
Source: Lipscomb et al. (1996).
6/22/2009
74
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
DCA and TCA are known to bind to plasma proteins. Schultz et al. (1999) measured
DC A binding in rats at a single concentration of about 100 |iM and found a binding fraction of
less than 10%. However, these data are not greatly informative for TCE exposure in which DCA
levels are significantly lower, and limitation to a single concentration precludes fitting to
standard binding equations from which the binding at low concentrations could be extrapolated.
Templin et al. (1993, 1995a,b), Schultz et al. (1999), Lumpkin et al. (2003), and Yu et al. (2003)
all measured TCA binding in various species and at various concentration ranges. Of these,
Templin et al. (1995a,b) and Lumpkin et al. (2003) measured levels in humans, mice, and rats.
Lumpkin et al. (2003) studied the widest concentration range, spanning reported TCA plasma
concentrations from experimental studies. Table 3.3.5 shows derived binding parameters.
However, these data are not entirely consistent among researchers; 2- to 5-fold differences in
Bmax and Kd are noted in some cases, although some differences existed in the rodent strains and
experimental protocols used. In general, however, at lower concentrations, the bound fraction
appears greater in humans than in rats and mice. Typical human TCE exposures, even in
controlled experiments with volunteers, lead to TCA blood concentrations well below the
reported Kd (Table 3.3.5, below), so the TCA binding fraction should be relatively constant.
However, in rats and mice, experimental exposures may lead to peak concentrations similar to,
or above, the reported Kd (e.g., Templin et al., 1993; Yu et al., 2000), meaning that the bound
fraction should temporarily decrease following such exposures.
6/22/2009
75
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.3.5. Reported TCA plasma binding parameters
A
Bmax
(pM)
K„
(pM)
A+
Bmax/Kd
Concentration
Range (jiM
bound+free)
Human
Templin et al. (1995a)
-
1020
190
5.37
3-1224
Lumpkin et al. (2003)
-
708.9
174.6
4.06
0.06-3065
Rat
Templin et al. (1995a)
-
540
400
1.35
3-1224
Yu et al. (2000)
0.602
312
136
2.90
3.8-1530
Lumpkin et al. (2003)
-
283.3
383.6
0.739
0.06-3065
Mouse
Templin et al. (1993)
-
310
248
1.25
3-1224
Lumpkin et al. (2003)
-
28.7
46.1
0.623
0.06-1226
Notes: Binding parameters based on the equation Cbound = A * Cfree + Bmax * Cfree / (Kd + Cfree),
where Cb0Und is the bound concentration, Cfree is the free concentration, and A = 0 for
Templin et al. (1993, 1995a) and Lumpkin et al. (2003). The quantity A+ Bmax/Kd is the ratio of
bound-to-free at low concentrations.
Limited data is available on tissue:blood partitioning of the oxidative metabolites CH,
TCA, TCOH and DC A, as shown in Table 3.3.6. As these chemicals are all water soluble and
not lipophilic, it is not surprising that their partition coefficients are close to 1 (within about
2-fold). It should be noted that the TCA tissue:blood partition coefficients reported in
Table 3.3.6 were measured at concentrations 1.6-3.3 M, over 1000-fold higher than the reported
Kd. Therefore, these partition coefficients should reflect the equilibrium between tissue and free
blood concentrations. In addition, only one in vitro measurement has been reported of
blood:plasma concentration ratios for TCA: Schultz et al. (1999) reported a value of 0.76 in rats.
6/22/2009
76
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.3.6. Partition coefficients for TCE oxidative metabolites
Species/Tissue
TissuerBlood Partition Coefficient
CH
TCA
TCOH
DCA
HUMAN3
Kidney
-
0.66
2.15
-
Liver
-
0.66
0.59
-
Lung
-
0.47
0.66
-
Muscle
-
0.52
0.91
-
MOUSEb
Kidney
0.98
0.74
1.02
0.74
Liver
1.42
1.18
1.3
1.08
Lung
1.65
0.54
0.78
1.23
Muscle
1.35
0.88
1.11
0.37
a Fisher et al. (1998).
b Abbas and Fisher (1997).
Note: TCA and TCOH partition coefficients have not been reported for rats.
3.3.3.1.5 Species-, Sex-, and age-dependent differences of oxidative metabolism
The ability to describe species- and sex-dependent variations in TCE metabolism is
important for species extrapolation of bioassay data and identification of human populations that
are particularly susceptible to TCE toxicity. In particular, information on the variation in the
initial oxidative step of CH formation from TCE is desirable, because this is the rate-limiting
step in the eventual formation and distribution of the putative toxic metabolites TCA and DCA
(Lipscomb et al., 1997).
Inter- and intraspecies differences in TCE oxidation have been investigated in vitro using
cellular or subcelluar fractions, primarily of the liver. The available in vitro metabolism data on
TCE oxidation in the liver (Table 3.3.1) show substantial inter and intraspecies variability.
Across species, microsomal data show that mice apparently have greater capacity (Vmax) than rat
or humans, but the variability within species can be 2- to 10-fold. Part of the explanation may be
related to CYP2E1 content. Although liver P450 content is similar across species, mice and rats
exhibit higher levels of CYP2E1 content (0.85 and 0.89 nmol/mg protein, respectively)
(Nakajima et al., 1993; Davis et al., 2002) than humans (approximately 0.25-0.30 nmol/mg
protein) (Elfarra et al., 1998; Davis et al., 2002). Thus, the data suggest that rodents would have
a higher capacity than humans to metabolize TCE, but this is difficult to verify in vivo because
6/22/2009
77
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
very high exposure concentrations in humans would be necessary to assess the maximum
capacity of TCE oxidation.
With respect to the Km of liver microsomal TCE oxidative metabolism, where Km is
indicative of affinity (the lower the numerical value of Km. the higher the affinity), the trend
appears to be mice and rats have higher Km values (i.e., lower affinity) than humans, but with
substantial overlap due to inter-individual variability. Note that, as shown in Table 3.3.1, the
ranking of rat and mouse liver microsomal Km values between the two reports by Lipscomb et al.
(1998b) and Elfarra et al. (1998) is not consistent. However, both studies clearly show that Km is
the lowest (i.e., affinity is highest) in humans. Because clearance at lower concentrations is
determined by the ratio Vmax to Km, the lower apparent Km in humans may partially offset the
lower human Vmax, and lead to similar oxidative clearances in the liver at environmentally
relevant doses. However, differences in activity measured in vitro may not translate into in vivo
differences in metabolite production, as the rate of metabolism in vivo depends also on the rate of
delivery to the tissue via blood flow (e.g., Lipscomb et al., 2003). The interaction of enzyme
activity and blood flow is best investigated using PBPK models and is discussed, along with
descriptions of in vivo data, in Section 3.5.
Data on sex- and age-dependence in oxidative TCE metabolism are limited but suggest
relatively modest differences in humans and animals. In an extensive evaluation of
CYP-dependent activities in human liver microsomal protein and cryopreserved hepatocytes,
Parkinson et al. (2004) identified no age or gender-related differences in CYP2E1 activity. In
liver microsomes from 23 humans, the Km values for females was lower than males, but Vmax
values were very similar (Lipscomb et al., 1997). Appearance of total trichloro compounds in
urine following intrapertoneal dosing with TCE was 28% higher in female rats than in males
(Verma and Rana, 2003). The oxidation of TCE in male and female rat liver microsomes was
not significantly different; however, pregnancy resulted in a decrease of 27-39% in the rate of
CH production in treated microsomes from females (Nakajima et al., 1992b). Formation of CH
in liver microsomes in the presence of 0.2 or 5.9 mM TCE exhibited some dependency on age of
rats, with formation rates in both sexes of 1.1-1.7 nmol/mg protein/minute in 3-week-old animals
and 0.5-1.0 nmol/mg protein/minute in 18-week old animals (Nakajima et al., 1992b).
Fisher et al. (1991) reviewed data available at that time on urinary metabolites to
characterize species differences in the amount of urinary metabolism accounted for by TCA
(Table 3.3.7). They concluded that TCA seemed to represent a higher percentage of urinary
metabolites in primates than in other mammalian species, indicating a greater proportion of
oxidation leading ultimately to TCA relative to TCOG.
6/22/2009
78
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.3.7. Urinary excretion of trichloroacetic acid by various species exposed to
2 trichloroethylene (based on data reviewed in Fisher et al., 1991)
Species
Percentage of
Urinary
Excretion of
TCA
Dose Route
TCE Dose
References, comments
Male
Female
Baboon3'0
16
—
Intramuscular
injection
50 mg TCE/Kg
Mueller et al. (1982)
Chimpanzee3
24
22
Intramuscular
injection
50 mg TCE/Kg
Mueller et al. (1982)
Monkey,
Rhesus3,0
19
—
Intramuscular
injection
50 mg TCE/Kg
Mueller et al. (1982)
Mice, NMRIb
—
8-20
Oral intubation
2-200 mg
TCE/Kg
Dekant et al. (1986a)
Mice, B6C3F13
7-12
—
Oral intubation
10-2000 mg
TCE/Kg
Green and Prout (1985)
Rabbit,
Japanese
White30
0.5
Intraperitoneal
injection
200 mg
TCE/Kg
Nomiyama and Nomiyama (1979)
Rat, Wistarb
—
14-17
Oral intubation
2-200 mg
TCE/Kg
Dekant et al. (1986a)
Rat, Osborne-
Mendel3
6-7
—
Oral intubation
10-2000 mg
TCE/Kg
Green and Prout (1985)
Rat, Holt/man3
7
—
Intraperitoneal
injection
10 mg TCE/rat
Nomiyama and Nomiyama (1979)
3 Percentage urinary excretion determined from accumulated amounts of TCOH and TCA in urine 3 to 6 days
4 postexposure.
5 Percentage urinary excretion determined from accumulated amounts of TCOH, dichloroacetic acid, oxalic acid, and
6 iV-(hydroxyacetyl)aminoethanol in urine 3 days postexposure.
7 °Sex is not specified.
8 Note: Human data tabulated in Fisher et al. (1991) from Nomiyama and Nomiyama (1971) was not included here
9 because it was relative to urinary excretion of total trichloro-compounds, not as fraction of intake as was the case for
10 the other data included here.
11 3.3.3.1.6 CYP isoforms and genetic polymorphisms
12 A number of studies have identified multiple P450 isozymes as having a role in the
13 oxidative metabolism of TCE. These isozymes include CYP2E1 (Nakajima et al., 1992a;
14 Guengerich and Shimada, 1991; Guengerich et al., 1991; Nakajima et al., 1990;
15 Nakajima et al., 1988), CYP3A4 (Shimada et al., 1994), CYP1A1/2, CYP2C11/6
16 (Nakajima et al., 1993, 1992a), CYP2F, and CYP2B1 (Forkert et al., 2005). Recent studies in
17 CYP2E1-knockout mice have shown that in the absence of CYP2E1, mice still have substantial
18 capacity for TCE oxidation (Kim and Ghanayem 2006; Forkert et al., 2006). However, CYP2E1
19 appears to be the predominant (i.e., higher affinity) isoform involved in oxidizing TCE
20 (Nakajima et al., 1992a; Guengerich and Shimada, 1991; Guengerich et al., 1991;
21 Forkert et al., 2005). In rat liver, CYP2E1 catalyzed TCE oxidation more than CYP2C11/6
6/22/2009
79
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(Nakajima et al., 1992a). In rat recombinant-derived P450s, the CYP2E1 had a lower Km (higher
affinity) and higher Vmax/Km ratio (intrinsic clearance) than CYP2B1 or CYP2F4 (Forkert et al.,
2005). Interestingly, there was substantial differences in Km between rat and human CYP2Els
and between rat CYP2F4 and mouse CYP2F2, suggesting that species-specific isoforms have
different kinetic behavior (Table 3.3.8).
Table 3.3.8. P450 isoform kinetics for metabolism of TCE to CH in human, rat, and mouse
recombinant P450s
Experiment
Km
\iM
Vmax
pmol/min/pmol P450
V max/Km
Human rCYP2El
196 + 40
4 + 0.2
0.02
Rat rCYP2El
14 + 3
11 + 0.3
0.79
Rat rCYP2B 1
131 + 36
9 + 0.5
0.07
Rat rCYP2F4
64 + 9
17 + 0.5
0.27
Mouse rCYP2F2
114+17
13 + 0.4
0.11
Source: Forkert et al. (2005)
The presence of multiple P450 isoforms in human populations affects the variability in
individuals' ability to metabolize TCE. Studies using microsomes from human liver or from
human lymphoblastoid cell lines expressing CYP2E1, CYP1A1, CYP1A2, or CYP3A4 have
shown that CYP2E1 is responsible for greater than 60% of oxidative TCE metabolism
(Lipscomb et al., 1997). Similarities between metabolism of chlorzoxazone (a CYP2E1
substrate) in liver microsomes from 28 individuals (Peter et al., 1990) and TCE metabolism
helped identify CYP2E1 as the predominant (high affinity) isoform for TCE oxidation.
Additionally, Lash et al. (2000a) suggested that, at concentrations above the Km value for
CYP2E1, CYP1A2 and CYP2A4 may also metabolize TCE in humans; however, their
contribution to the overall TCE metabolism was considered low compared to that of CYP2E1.
Given the difference in expression of known TCE-metabolizing P450 isoforms (Table 3.3.9) and
the variability in P450-mediated TCE oxidation (Lipscomb et al., 1997), significant variability
may exist in individual human susceptibility to TCE toxicity.
Table 3.3.9. P450 isoform activities in human liver microsomes exhibiting different
affinities for TCE
Affinity Group
CYP Isoform Activity (pmol/min/mg protein)
CYP2E1
CYP1A2
CYP3A4
6/22/2009
80
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Low Km
520 + 295
241+ 146
2.7 + 2.7
Mid Km
820 + 372
545 + 200
2.9 + 2.8
High Km
1317 + 592
806 + 442
1.8+1.1
Activities of CYP1A2, CYP2E1, and CYP3A4 were measured with phenacetin, chlorzoxazone,
and testosterone as substrates, respectively. Data are means + standard deviation from 10, 9, and
4 samples for the low-, mid-, and high-Km groups, respectively. Only CYP3 A4 activities are not
significantly different (p < 0.05) from one another by Kruskal-Wallis one-way analysis of
variance.
Source: Lash et al. (2000a).
Differences in content and/or intrinsic catalytic properties (Km, Vmax) of specific enzymes
among species, strains, and individuals may play an important role in the observed differences in
TCE metabolism and resulting toxicities. Lipscomb et al. (1997) reported observing three
statistically distinct groups of Km values for TCE oxidation using human microsomes. The mean
± SD ([xM TCE) for each of the three groups was 16.7 + 2.5 (n = 10), 30.9 + 3.3 (n = 9), and 51.1
+ 3.8 (n = 4). Within each group, there were no significant differences in sex or ethnicity.
However, the overall observed Km values in female microsomes (21.9 + 3.5 [xM, n= 10) were
significantly lower than males (33.1 + 3.5 [xM, n = 13). Interestingly, in human liver
microsomes, different groups of individuals with different affinities for TCE oxidation appeared
to also have different activities for other substrates not only with respect to CYP2E1 but also
CYP1A2 (Lash et al., 2000a) (Table 3.3.9). Genetic polymorphisms in humans have been
identified in the CYP isozymes thought to be responsible for TCE metabolism (Pastino et al.,
2000), but no data exist correlating these polymorphisms with enzyme activity. It is relevant to
note that repeat polymorphism (Hu et al., 1999) or polymorphism in the regulatory sequence
(McCarver et al., 1998) were not involved in the constitutive expression of human CYP 2E1;
however, it is unknown if these types of polymorphisms may play a role in the inducibility of the
respective gene.
Individual susceptibilities to TCE toxicity may also result from variations in enzyme
content, either at baseline or due to enzyme induction/inhibition, which can lead to alterations in
the amounts of metabolites formed. Certain physiological and pathological conditions or
exposure to other chemicals (e.g., ethanol and acetominophen) can induce, inhibit, or compete
for enzymatic activity. Given the well established (or characterized) role of the liver to
oxidatively metabolize TCE (by CYP2E1), increasing the CYP2E1 content or activity (e.g., by
enzyme induction) may not result in further increases in TCE oxidation. Indeed, Kaneko et al.
(1994) reported that enzyme induction by ethanol consumption in humans increased TCE
metabolism only at high concentrations (500 ppm, 2687 mg/m3) in inspired air. However, other
6/22/2009
81
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
interactions between ethanol and the enzymes that oxidatively metabolize TCE metabolites can
result in altered metabolic fate of TCE metabolites. In addition, enzyme inhibition or
competition can decrease TCE oxidation and subsequently alter the TCE toxic response via, for
instance, increasing the proportion undergoing GSH conjugation (Lash et al., 2000a). TCE itself
is a competitive inhibitor of CYP2E1 activity (Lipscomb et al., 1997), as shown by reducedp-
nitrophenol hydroxylase activity in human liver microsomes, and so may alter the toxicity of
other chemicals metabolized through that pathway. On the other hand, suicidal CYP heme
destruction by the TCE-oxygenated P-450 intermediate has also been shown (Miller and
Guengerich, 1983).
3.3.3.2 GSH Conjugation Pathway
Historically, the conjugative metabolic pathways have been associated with xenobiotic
detoxification. This is true for GSH conjugation of many compounds. However, several
halogenated alkanes and alkenes, including TCE, are bioactivated to cytotoxic metabolites by the
GSH conjugate processing pathway (mercapturic acid) pathways (Elfarra et al., 1986a,b). In the
case of TCE, production of reactive species several steps downstream from the initial GSH
conjugation is believed to cause cytotoxicity and carcinogenicity, particularly in the kidney.
Since the GSH conjugation pathway is in competition with the P450 oxidative pathway for TCE
biotransformation, it is important to understand the role of various factors in determining the flux
of TCE through each pathway. Figure 3.3.4 depicts the present understanding of TCE
metabolism via GSH conjugation.
6/22/2009
82
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
CI
CI
r
Cl2 C2 H
V.
O ^
II NH,+
Mf
O- y
(DCVCS)
H (TCE) CI
CI
H
GST
SG
CI
GGT
Cl2 c2 H
FMO-3
CGDP
(DCVC)
B-lyase
(DCVT)
NAT
q' Acylase ^'2 ^2 ^
(NAcDCVC)
Cl2 C2 H
CYP3A
(NAcDCVCS)
Figure 3.3.4. Scheme for glutathione-dependent (GSH) metabolism of TCE
Adapted from: Lash et al. (2000a); Cummings and Lash (2000); NRC (2006).
3.3.3.2.1 Formation of DCVG
The conjugation of TCE to GSH produces S-(l,2-dichlorovinyl)glutathione (DCVG).
There is some uncertainty as to which glutathione-S-transferase (GST) isoforms mediate TCE
conjugation. Lash and colleagues studied TCE conjugation in renal tissue preparations, isolated
renal tubule cells from male F344 rats and purified GST alpha-class isoforms 1-1, 1-2 and 2-2
(Cummings et al., 2000a; Cummings and Lash 2000; Lash et al., 2000b). The results
demonstrated high conjugative activity in renal cortex and in proximal tubule cells. Although
the isoforms studied had similar Vmax values, the Km value for GST 2-2 was significantly lower
6/22/2009
83
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
than the other forms, indicating that this form will catalyze TCE conjugation at lower (more
physiologically relevant) substrate concentrations. In contrast, using purified rat and human
enzymes, Hissink et al. (2002) reported in vitro activity for DCVG formation only for mu- and
pi-class GST isoforms, and none towards alpha-class isoforms; however, the rat mu-class GST 3-
3 was several folds more active than the human mu-class GST Ml-1. Although GSTs are
present in tissues throughout the body, the majority of TCE GSH conjugation is thought to occur
in the liver (Lash et al., 2000a). Using in vitro studies with renal preparations, it has been
demonstrated that GST catalyzed conjugation of TCE is increased following the inhibition of
CYP-mediated oxidation (Cummings et al., 2000b).
In F344 rats, following gavage doses of 263-1971 mg/kg TCE in 2 ml corn oil, DCVG
was observed in the liver and kidney of females only, in blood of both sexes (Lash et al., 2006),
and in bile of males (Dekant et al., 1990). The data from Lash et al. (2006) are difficult to
interpret because the time courses seem extremely erratic, even for the oxidative metabolites
TCOH and TCA. Moreover, a comparison of blood levels of TCA and TCOH with other studies
in rats at similar doses reveals differences of over 1000-fold in reported concentrations. For
instance, at the lowest dose of 263 mg/kg, the peak blood levels of TCE and TCA in male F344
rats were 10.5 and 1.6 |ig/L, respectively (Lash et al., 2006). By contrast, Larson and Bull
(1992a) reported peak blood TCE and TCA levels in male Sprague-Dawley rats over 1000-fold
higher—around 10 and 13 mg/L, respectively—following oral doses of 197 mg/kg as a
suspension in 1% aqueous Tween 80. The results of Larson and Bull (1992a) are similar to Lee
et al. (2000a), who reported peak blood TCE levels of 20-50 mg/L after male Sprague-Dawley
rats received oral doses of 144-432 mg/kg in a 5% aqueous Alkamus emulsion (polyethoxylated
vegetable oil), and to Stenner et al. (1997), who reported peak blood levels of TCA in male F344
rats of about 5 mg/L at a slightly lower TCE oral dose of 100 mg/kg administered to fasted
animals in 2% Tween 80. Thus, while useful qualitatively as an indicator of the presence of
DCVG in rats, the quantitative reliability of reported concentrations, for metabolites of either
oxidation or GSH conjugation, may be questionable.
In humans, DCVG was readily detected at in human blood following onset of a 4-hour
TCE inhalation exposure to 50 or 100 ppm (269 or 537 mg/m3; Lash et al., 1999a). At 50 ppm,
peak blood levels ranged from 2.5 to 30 [xM, while at 100 ppm, the mean (+ SE, n=8) peak blood
levels were 46.1 + 14.2 [xM in males and 13.4 + 6.6 [jM in females. While on average, male
subjects had 3-fold higher peak blood levels of DCVG than females, in half of the male subjects,
DCVG blood levels were similar to or lower than those of female subjects. This suggests a
polymorphism in GSH conjugation of TCE rather than a true gender difference (Lash et al.,
1999a) as also has been indicated by Hissink et al. (2002) for the human mu-class GST Ml-1
enzyme. Interestingly, as shown in Table 3.3.10, the peak blood levels of DCVG are similar on a
6/22/2009
84
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 molar basis to peak levels of TCE, TCA, and TCOH in the same subjects, as reported in
2 Fisher et al. (1998).
3
4 Table 3.3.10. Comparison of peak blood concentrations in humans exposed to 100 ppm
5 (537 mg/m3) TCE for 4 hours (Fisher et al., 1998; Lash et al., 1999a)
Chemical Species
Peak Blood Concentration (mean + SD, jiM)
Males
Females
TCE
23 + 11
14 + 4.7
TCA
56 + 9.8
59 + 12
TCOH
21 + 5.0
15 + 5.6
DCVG
46.1 + 14.2
13.4 + 6.6
6
7 Tables 3.3.11 and 3.3.12 summarize DCVG formation from TCE conjugation from in
8 vitro studies of liver and kidney cellular and subcellular fractions in mouse, rat, and human.
9 Tissue-distribution and species-and gender-differences in DCVG formation are discussed below.
6/22/2009
85
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
2 Table 3.3.11. GSH conjugation of TCE (at 1-2 mM) in liver and kidney cellular fractions in
3 humans, male F344 rats, and male B6C3F1 mice
Species and Cellular/Sub-Cellular Fraction
(TCE concentration)
DCVG Formation
(nmol/hour/mg protein or 106 cells)
Male
Female
Human
Hepatocytes (0.9 mM) [pooled]
11+3
Liver cytosol (1 mM) [individual samples]
156+16
174+13
Liver cytosol (2 mM) [pooled]
346
Liver microsomes (1 mM) [individual samples]
108 + 24
83 + 11
Liver microsomes (1 mM) [pooled]
146
Kidney cytosol (2 mM) [pooled]
42
Kidney microsomes (1 mM) [pooled]
320
Rat
Liver cytosol (2 mM)
7.30 + 2.8
4.86 + 0.14
Liver microsomes (2 mM)
10.3 + 2.8
7.24 + 0.24
Kidney cortical cells (2 mM)
0.48 + 0.02
0.65 + 0.15
Kidney cytosol (2 mM)
0.45 + 0.22
0.32 + 0.02
Kidney microsomes (2 mM)
not detected
0.61+0.06
Mouse
Liver cytosol (2 mM)
24.5 + 2.4
21.7 + 0.9
Liver microsomes (2 mM)
40.0 + 3.1
25.6 + 0.8
Kidney cytosol (2 mM)
5.6 + 0.24
3.7 + 0.48
Kidney microsomes (2 mM)
5.47+1.41
16.7 + 4.7
4 Mean + SE. Source: Lash et al. (1999a, 1998, 1995); Cummings and Lash (2000);
5 Cummings et al. (2000b).
6
7
6/22/2009 86
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.3.12. Kinetics of TCE metabolism via GSH conjugation in male F344 rat kidney
and human liver and kidney cellular and subcellular fractions
V max
(nmol
Tissue and Cellular Fraction
Km
(jiM TCE)
DCVG/min/
mg protein
or 106
hepatocytes)
1000 X
V /K
' max' lvm
Rat
Kidney proximal tubular cells: low affinity
2910
0.65
0.22
Kidney proximal tubular cells: high
460
0.47
1.0
affinity
Human
Liver hepatocytesa
37-106
0.16-0.26
2.4-4.5
Liver cytosol: low affinity
333
8.77
2.6
Liver cytosol: high affinity
22.7
4.27
190
Liver microsomes: low affinity
250
3.1
12
Liver microsomes: high affinity
29.4
1.42
48
Kidney proximal tubular cells: low affinity
29,400
1.35
0.046
Kidney proximal tubular cells: high
580
0.11
0.19
affinity
Kidney cytosol
26.3
0.81
31
Kidney microsomes
167
6.29
38
Source: Lash et al. (1999a); Cummings and Lash (2000); Cummings et al. (2000b).
a Kinetic analyses of first 6 to 9 (out of 10) data points from Fig 1. from Lash et al. (1999a) using
Lineweaver-Burk or Eadie-Hofstee plots and linear regression (R2 = 0.50-0.95). Regression
with best R2 used first 6 data points and Eadie-Hofstee plot, with resulting Km and Vmax of 106
and 0.26, respectively.
3.3.3.2.2 Formation of DCVC
The cysteine conjugate, S-(l,2-dichlorovinyl) cysteine (DCVC), is formed from DCVG
in a two-step sequence. DCVG is first converted to the cysteinylglycine conjugate
S-(l,2-dichlorovinyl)-L-cysteinylglycine (DCVCG) by y-glutamyltransferase (GGT) in the renal
brush border (Elfarra and Anders, 1984; Lash et al., 1988).
6/22/2009
87
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Cysteinylglycine dipeptidases in the renal brush border and basolateral membrane
convert DCVG to DCVC via glycine cleavage (Goeptar et al., 1995; Lash et al., 1998). This
reaction can also occur in the bile or gut, as DCVG excreted into the bile is converted to DCVC
and reabsorbed into the liver where it may undergo further acetylation.
3.3.3.2.3 Formation of NAcDCVC
N-acetylation of DCVC can either occur in the kidney, as demonstrated in rat kidney
microsomes (Duffel and Jakoby, 1982), or in the liver (Birner et al., 1997). Subsequent release
of DCVC from the liver to blood may result in distribution to the kidney resulting in increased
internal kidney exposure to the acetylated metabolite over and above what the kidney already is
capable of generating. In the kidney, NAcDCVC may undergoe deacetylation, which is
considered a rate-limiting-step in the production of proximal tubule damage (Wolfgang et al.,
1989; Zhang and Stevens, 1989). As a polar mercapturtae, NAcDCVC may be excreted in the
urine as evidenced by findings in mice (Birner et al., 1993), rats (Bernauer et al., 1996;
Commandeur and Vermeulen, 1990), and humans who were exposed to TCE (Bernauer et al.,
1996; Birner et al., 1993), suggesting a common glutathione-mediated metabolic pathway for
DCVC among species.
3.3.3.2.4 Beta lyase metabolism of DCVC
The enzyme cysteine conjugate B-lyase catalyzes the breakdown of DCVC to reactive
nephrotoxic metabolites (Goeptar et al., 1995). This reaction involves removal of pyruvate and
ammonia and production of S-(l,2-dichlorovinyl) thiol (DCVT), an unstable intermediate, which
rearranges to other reactive alkylation metabolites that form covalent bonds with cellular
nucleophiles (Goeptar et al., 1995; Dekant et al., 1988). The rearrangement of DCVT to
enethiols and their acetylating agents has been described in trapping experiments
(Dekant et al., 1988) and proposed to be responsible for nucleophilic adduction and toxicity in
the kidney. The quantification of acid-labile adducts was proposed as a metric for TCE flux
through the GSH pathway. However, the presence of analytical artifacts precluded such
analysis. In fact, measurement of acid-labile adduct products resulted in higher values in mice
than in rats (Eyre et al., 1995a, b).
DCVC metabolism to reactive species via a P-lyase pathway has not been directly
observed in vivo in animals or humans. However, P-lyase activity in humans and rats (reaction
rates were not reported) was demonstrated in vivo using a surrogate substrate,
2-(fluoromethoxy)-l,l,3,3,3-pentafluoro-l-propene (Iyer et al., 1998). P-lyase -mediated
reactive adducts have been described in several extra-renal tissues, including rat and human liver
and intestinal microflora (Larsen and Stevens, 1986; Tomisawa et al., 1984, 1986; Stevens,
6/22/2009
88
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1985a; Stevens and Jakoby, 1983; Dohn and Anders, 1982; Tateishi et al., 1978) and rat brain
(Alberati-Giani et al., 1995; Malherbe et al., 1995).
In the kidneys, glutamine transaminase K appears to be primarily responsible for P-lyase
metabolism of DCVC (Perry et al., 1993; Lash et al., 1990a; Jones et al., 1988;
Stevens et al., 1988; Stevens et al., 1986; Lash et al., 1986). P-lyase transformation of DCVC
appears to be regulated by 2-keto acids. DCVC toxicity in isolated rat proximal tubular cells was
significantly increased with the addition of a-keto-y-methiolbutyrate or phenylpyruvate (Elfarra
et al., 1986b). The presence of a-keto acid cofactors is necessary to convert the inactive form of
the P-lyase enzyme (containing pyridoxamine phosphate) to the active form (containing
pyridoxal phosphate) (Goeptar et al., 1995).
Both low- and high-molecular-weight enzymes with P-lyase activities have been
identified in rat kidney cytosol and mitochondria (Abraham et al., 1995a, b; Stevens et al., 1988;
Lash et al., 1986). While glutamine transaminase K and kynureninase-associated P-lyase
activities have been identified in rat liver (Alberati-Giani et al., 1995; Stevens, 1985a), they are
quite low compared to renal glutamine transaminase K activity and do not result in
hepatotoxicity in DCVG- or DCVC-treated rats (Elfarra and Anders, 1984). Similar isoforms of
P-lyase have also been reported in mitochondrial fractions of brain tissue (Cooper, 2004).
The kidney enzyme L-a-hydroxy (L-amino) acid oxidase is capable of forming an
iminium intermediate and keto acid analogues (pyruvate or S-(l,2-dichlorovinyl)-2-oxo-3-
mercaptopropionate) of DCVC, which decomposes to DCVSH (Lash et al., 1990b; Stevens et al.,
1989). In rat kidney homogenates, this enzyme activity resulted in as much as 35% of GSH
pathway-mediated bioactivation. However, this enzyme is not present in humans, an important
consideration for extrapolation of renal effects across species.
3.3.3.2.5 Sulfoxidation of DCVC and NAcDCVC
A second pathway for bioactivation of TCE ^-conjugates involves sulfoxidation of either
the cysteine or mercapturic acid conjugates (Sausen and Elfarra, 1990; Park et al., 1992;
Lash et al., 1994, 2003; Werner et al., 1995a, b, 1996; Birner et al., 1998; Krause et al., 2003).
Sulfoxidation of DCVC was mediated mainly by flavin monooxygenase (FM03), rather than
CYP450, in rabbit liver microsomes (Ripp et al, 1997) and human liver microsomes (Krause et
al., 2003). Krause et al. (2003) was not able to detect sulfoxidation in human kidney
microsomes, and the authors attributed the lack of metabolic actibvity to low and variable FM03
expression in the kidney when compared to liver.
Sulfoxidation of NAcDCVC, by contrast, was found to be catalyzed predominantly, if not
exclusively, by CYP3A enzymes (Werner et al., 1996), whose expressions are highly
polymorphic in humans. Sulfoxidation of other haloalkyl mercapturic acid conjugates has also
been shown to be catalyzed by CYP3A (Werner et al., 1995a, b; Altuntas et al., 2004). While
6/22/2009
89
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Lash et al. (2000a) suggested that this pathway would be quantitatively minor because of the
relatively low CYP3 A levels in the kidney, no direct data exist to establish the relative
toxicological importance of this pathway relative to bioactivation of DCVC by P-lyase or FMO.
However, the contribution of CYP3 A in S-conjugate sulfoxidation to nephrotoxicity in vivo was
recently demonstrated by Sheffels et al. (2004) with fluoromethyl-2,2-difluoro-l-
(trifluoromethyl)vinyl ether (FDVE). In particular, in vivo production and urinary excretion of
FDVE-mercapturic acid sulfoxide metabolites were unambiguously established by mass
spectrometry, and P450 inducers/inhibitors increased/decreased nephrotoxicity in vivo while
having no effect on urinary excretion of metabolites produced through P-lyase (Sheffels et al.,
2004). These data suggest that, by analogy, sulfoxidation of NAcDCVC may be an important
bioactivating pathway.
3.3.3.2.6 Tissue distribution of GSH metabolism
The sites of enzymatic metabolism of TCE to the various GSH pathway-mediated
metabolites are significant in determining target tissue toxicity along this pathway. Figure 3.3.5
presents a schematic of interorgan transport and metabolism of TCE along the glutathione
pathway. TCE is taken up either by the liver or kidney and conjugated to DCVG. The primary
factors affecting TCE flux via this pathway include high hepatic GST activity, efficient transport
of DCVG from the liver to the plasma or bile, high renal brush border and low hepatic GGT
activities, and the capability for GSH conjugate uptake into the renal basolateral membranes with
limited or no uptake into liver cell plasma membranes.
6/22/2009
90
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE
DCVG
DCVC
¦¦~Blood flow
-~Bile flow
~Glomerular filtration
DCVT
^~Metabolism
NAcDCVC
Blood/
Blood/
Blood/
Blood/
Plasma
Plasma
Plasma
Plasma
Rest of
Rest of
Rest of
Rest of
Small
Intestine
Small
Intestine
Small
Intestine
Small
Intestine
Kidney
Kidney
Kidney
Kidney
DCVCS NAcDCVCS
Figure 3.3.5. Interorgan TCE transport and metabolism via the GSH pathway. See Figure 3.3.4
for enzymes involved in metabolic steps. Source: Lash et al. (2000a,b); NRC (2006).
As discussed previously, GST activity is present in many different cell types. However,
the liver is the major tissue for GSH conjugation. GST activities in rat and mouse cytosolic
fractions were measured using l-chloro-2,4-dinitrobenzene, a GST substrate that is non-specific
for particular isoforms (Lash et al., 1998). Specific activities (normalized for protein content) in
whole kidney cytosol were slightly less than those in the liver (0.64 compared to 0.52 mU/mg
protein for males and females). However, the much larger mass of the liver compared to the
kidney indicates that far more total GST activity resides in the liver. This is consistent with in
vitro data on TCE conjugation to DCVG, discussed previously (Table 3.3.11 and Table 3.3.12).
For instance, in humans, rats, and mice, liver cytosol exhibits greater DCVG production than
kidney cytosol. Distinct high- and low-affinity metabolic profiles were observed in the liver but
6/22/2009
91
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
not in the kidney (Table 3.3.12). In microsomes, human liver and kidney had similar rates of
DCVG production, while for rats and mice, the production in the liver was substantially greater.
According to studies by Lash et al. (1998, 1999b), the activity of GGT, the first step in
the conversion of DCVG to DCVC, is much higher in the kidney than the liver of mice, rats, and
humans, with most of the activity being concentrated in the microsomal, rather than the
cytosolic, fraction of the cell (Table 3.3.13). In rats, this activity is quite high in the kidney but is
below the level of detection in the liver while the relative kidney to liver levels in humans and
mice were higher by 18- and up to 2300- fold, respectively. Similar qualitative findings were
also reported in another study (Hinchman and Ballatori, 1990) when total organ GGT levels were
compared in several species (Table 3.3.14). Cysteinylglycine dipeptidase was also preferentially
higher in the kidney than the liver of all tested species although the inter-organ differences in this
activity (1-9 folds) seemed to be less dramatic than for GGT (Table 3.3.14). High levels of both
GGT and dipeptidases have also been reported in the small intestine of rat (Kozak and Tate,
1982) and mouse (Habib et al., 1996, 1998), as well as GGT in the human jejunum (Fairman et
al., 1977). No specific human intestinal cysteinylglycine dipeptidase has been identified;
however, a related enzyme (EC 3.4.13.11) from human kidney microsomes has been purified and
studied (Adachi et al., 1989) while several human intestinal dipeptidases have been characterized
including a membrane dipeptidase (MDP; EC 3.4.13.19) which has a wide dipeptide substrate
specificity including cysteinylglycine (Hooper et al, 1994; Ristoff and Larsson, 2007).
6/22/2009
92
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.3.13. GGT activity in liver and kidney subcellular fractions of mice, rats, and
humans
Species
Sex
Tissue
Cellular Fraction
Activity (mU/mg)
Mouse
Male
Liver
Cytosol
0.07 + 0.04
Microsomes
0.05 + 0.04
Kidney
Cytosol
1.63 + 0.85
Microsomes
92.6+15.6
Female
Liver
Cytosol
0.10 + 0.10
Microsomes
0.03 + 0.03
Kidney
Cytosol
0.79 + 0.79
Microsomes
69.3 + 14.0
Rat
Male
Liver
Cytosol
<0.02
Microsomes
<0.02
Kidney
Cytosol
<0.02
Microsomes
1570+ 100
Female
Liver
Cytosol
<0.02
Microsomes
<0.02
Kidney
Cytosol
<0.02
Microsomes
1840 + 40
Human
Male
Liver
Cytosol
8.89 + 3.58
Microsomes
29
Kidney
Cytosol
13.2+1.0
Microsomes
960 + 77
Source: Lashetal. (1998, 1999b).
3.3.3.2.7 Sex- and Species-dependent differences in GSH metabolism
Diverse sex and species differences appear to exist in TCE metabolism via the
glutathione pathway. In rodents, rates of TCE conjugation to GSH in male rats and mice are
higher than females (Table 3.3.11). Verma and Rana (2003) reported 2-fold higher GST activity
values in liver cytosol of female rats, compared to males, given 15 intraperitoneal injections of
TCE over 30 days period. This effect may be due to sex-dependent variation in induction, as
GST activities in male and female controls were similar. DCVG formation rates by liver and
kidney subcellular fractions were much higher in both sexes of mice than in rats and, except for
mouse kidney microsomes, the rates were generally higher in males than in females of the same
species(Table 3.3.11).
6/22/2009
93
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
In terms of species differences, comparisons at 1-2 mM TCE concentrations (Table
3.3.11) suggest that, in liver and kidney cytosol, the greatest DCVG production rate was in
humans, followed by mice and then rats. However, different investigators have reported
considerably different rates for TCE conjugation in human liver and kidney cell fractions . For
instance, values in Table 3.3.11 from Lash et al. (1999a) are between two and five orders of
magnitude higher than those reported by Green et al. (1997a). [The rates of DCVG formation by
liver cytosol from male F344 rat, male B6C3F1 mouse, and human were 1.62, 2.5, and 0.19
pmol/min/mg protein, respectively, while there were no measurable activity in liver microsomes
or subcellular kidney fractions (Green et al., 1997a)]. The reasons for such discrepancies are
unclear but may be related to different analytical methods employed such as detection of
radiolabled substrate vs. derivatized analytes (Lash et al., 2000a).
Expression of GGT activity does not appear to be influenced by sex (Table 3.3.13); but
species differences in kidney GGT activity are notable with rat subcellular fractions exhibiting
the highest levels and mice and humans exhibiting about 4-6% and 50%, respectively, of rat
levels (Lash et al., 1999a, 1998). Table 3.3.14 shows measures of whole-organ GGT and
dispeptidase activities in rats, mice, guinea pigs, rabbits, pigs, and monkeys. These data show
that the whole kidney possesses higher activities than liver for these enzymes, despite the
relatively larger mass of the liver.
Table 3.3.14. Multi-species comparison of whole-organ activity levels of
y-glutamyltransferase (GGT) and dispeptidase
Species
Whole Organ Enzyme Activity (^imol substrate/organ)
Kidney
Liver
GGT
Dispeptidase
GGT
Dispeptidase
Rat
1010 + 41
20.2 + 1.1
7.1 + 1.4
6.1 + 0.4
Mouse
60.0 + 4.2
3.0 + 0.3
0.47 + 0.05
1.7 + 0.2
Rabbit
1119 + 186
112+17
71.0 + 9.1
12.6+1.0
Guinea pig
148 + 13
77+10
46.5 + 4.2
13.2+1.5
Pig
3800 + 769
2428 + 203
1600 + 255
2178 + 490
Macaque
988
136
181
71
Source: Hinchman and Ballatori (1990).
As discussed above, the three potential bioactivating pathways subsequent to the
formation of DCVC are catalyzed by P-lyase, FMO-3 or CYP3A. Lash et al. (2000a) compared
in vitro P-lyase activities and kinetic constants (when available) for kidney of rats, mice, and
6/22/2009
94
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
humans. They reported that variability of these values spans up to two orders of magnitude
depending on substrate, analytical method used, and research group. Measurements of rat,
mouse, and human P-lyase activities collected by the same researchers following
tetrachloroethylene exposure (Green et al., 1990) resulted in higher Km and lower Vmax values for
mice and humans than rats. Further, female rats exhibited higher Km and lower Vmax values than
males
With respect to FMO-3, Ripp et al. (1999) found that this enzyme appeared catalytically
similar across multiple species, including humans, rats, dogs, and rabbits, with respect to several
substrates, including DCVC, but that there were species differences in expression. Specifically,
in male liver microsomes, rabbits had 3-fold higher methionine S-oxidase activity than mice and
dogs had 1.5-fold higher activity than humans and rats. Species differences were also noted in
male and female kidney microsomes; rats exhibited 2- to 6-fold higher methionine S-oxidase
activity than the other species. Krause et al. (2003) detected DCVC sulfoxidation in incubations
with human liver microsomes but did not in an incubation with a single sample of human kidney
microsomes. However, FMO-3 expression in the 26 human kidney samples was found to be
highly variable, with a range of 5-6-fold (Krause et al., 2003). These data suggest that for a
given amount of DCVC, the rat kidney may bioactivate more through FMO-3 than the human
kidney, but in vivo data is lacking.
No data on species differences in CYP3A-mediated sulfoxidation of NAcDCVC are
available. However, Altuntas et al. (2004) examined sulfoxidation of cysteine and mercapturic
acid conjugates of FDVE (fluoromethyl-2,2-difluoro-l-(trifluoromethyl)vinyl ether) in rat and
human liver and kidney microsomes. They reported that the formation of sulfoxides from the
mercapturates N-Ac-FFVC and (Z)-N-Ac-FFVC (FFVC is (A",Z)-S-(1 -fluoro-2-fluoromethoxy-2-
(trifluoromethyl)vinyl)-Lcysteine) were greatest in rat liver microsomes, and 2- to 30-fold higher
than in human liver microsomes (which had high variability). Sulfoxidation of N-Ac-FFVC
could not be detected in neither rat nor human kidney microsomes, but sulfoxidation of (Z)-N-
Ac-FFVC was detected in both rat and human kidney microsomes at rates comparable to human
liver microsomes. Using human- and rat-expressed CYP3A, Altuntas et al. (2004) reported that
rates of sulfoxidation of (Z)-N-Ac-FFVC were comparable in human CYP3 A4 and rat CYP3 Al
and CYP3 A2., but that only rat CYP3 Al and A2 catalyzed sulfoxidation of N-Ac-FFVC. As the
presence or absence of the species differences in mercapturate sulfoxidation appear to be highly
chemical-specific, no clear inferences can be made as to whether species differences exist for
sulfoxidation of NAcDCVC
6/22/2009
95
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Also relevant to assess the flux through the various pathways are the rates of N-
acetylation and de-acetylation of DCVC. This is demonstrated by the results of Elfarra and
Hwang (1990) using using S-(2-benzothiazolyl)-L-cysteine (BTC) as a marker for P-lyase
metabolism in rats, mice, hamsters, and guinea pigs. Guinea pigs exhibited about 2-fold greater
flux through the P-lyase pathway, but this was not attributable to higher P-lyase activity. Rather,
guinea pigs have relatively low N-acetylation and high deacetylation activities, leading to a high
level of substrate recirculation (Lau et al., 1995). Thus, a high N-deacetylase:N-acetylase
activity ratio may favor DCVC recirculation and subsequent metabolism to reactive species. In
human, Wistar rat, Fischer rat, and mouse cytosol, deacetylation rates for NAcDCVC varied less
than 3-fold (0.35, 0.41, 0.61, and 0.94 nmol DCVC formed/min/mg protein in humans, rats, and
mice) (Birner et al., 1993). However, similar experiments have not been carried out for
N-acetylation of DCVC, so the balance between its N-acetylation and de-acetylation has not
been established.
3.3.3.2.8 Human variability and susceptibility in GSH conjugation
Knowledge of human variability in metabolizing TCE through the glutathione pathway is
limited to in vitro comparisons of variance in GST activity rates. Unlike P450-mediated
oxidation, quantitative differences in the polymorphic distribution or activity levels of GST
isoforms in humans are not presently known. However, the available data (Lash et al., 1999a, b)
do suggest that significant variation in GST-mediated conjugation of TCE exists in humans. In
particular, at a single substrate concentration of 1 mM, the rate of GSH conjugation of TCE in
human liver cytosol from 9 male and 11 females spanned a range of 2.4-fold (34.7-83.6 nmol
DCVG formed/20 min/mg protein) (Lash et al., 1999b). In liver microsomes from 5 males and
15 females, the variation in activity was 6.5-fold (9.9-64.6 nmol DCVG formed/20 min/mg
protein). No sex-dependent variation was identified. Despite being less pronounced than the
known variability in human P450-mediated oxidation, the impact on risk assessment of the
variability in GSH conjugation to TCE is currently unknown especially in the absence of data on
variability for N-acetylation and bioactivation via P-lyase, FMO, or CYP3 A in the human
kidney.
3.3.3.3 Relative Roles of the CYP and GSH Pathways
In vivo mass balance studies in rats and mice, discussed above, have shown
unequivocally that in these species, P450 oxidation of TCE predominates over GSH conjugation.
In these species, at doses from 2 to 2000 mg/kg of [14C] TCE, the sum of radioactivity in exhaled
TCE, urine, and exhaled CO2 constitutes 69-94% of the dose, with the vast majority of the
radioactivity (95-99%) in urine attributable to oxidative metabolites (Dekant et al., 1984; Dekant
et al., 1986a; Green and Prout 1985; Prout et al., 1995). The rest of the radioactivity was found
6/22/2009
96
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
mostly in feces and the carcass. More rigorous quantitative limits on the amount of GSH
conjugation based on in vivo data such as these can be obtained using PBPK models, discussed
in Section 3.5.
Comprehensive mass-balance studies are unavailable in humans. DCVG and DCVC in
urine have not been detected in any species, while the amount of urinary NAcDCVC from
human exposures is either below detection limits or very small from a total mass balance point of
view (Birner et al., 1993; Bernauer et al., 1996; Lash et al., 1999b; Bloemen et al., 2001). For
instance, the ratio of primary oxidative metabolites (TCA + TCOH) to NAcDCVC in urine of
rats and humans exposed to 40-160 ppm (215 to 860 mg/m3) TCE heavily favored oxidation,
resulting in ratios of 986-2562:1 in rats and 3292-7163:1 in humans (Bernauer et al., 1996).
Bloemen et al. (2001) reported that at most 0.05% of an inhaled TCE dose would be excreted as
NAcDCVC, and concluded that this suggested TCE metabolism by GSH conjugation was of
minor importance. Therefore, while it is a useful biomarker of exposure and an indicator of GSH
conjugation, NAcDCVC may capture only a small fraction of TCE flux through the GSH
conjugation pathway due to the dominance of bioactivating pathways (Lash et al., 2000a).
A number of lines of evidence suggest that the amount of TCE conjugation to GSH in
humans, while likely smaller than the amount of oxidation, may be much more substantial than
analysis of urinary mercapturates would suggest. In Table 3.3.15, in vitro estimates of the Vmax,
Km, and clearance (Vmax/Km) for hepatic oxidation and conjugation of TCE are compared in a
manner that accounts for differences in cytosolic and microsomal partitioning and protein
content. Surprisingly, the range of in vitro kinetic estimates for oxidation and conjugation of
TCE substantially overlap, suggesting similar flux through each pathway, though with high
inter-individual variation. The microsomal and cytosolic protein measurements of GSH
conjugation should be caveated by the observation by Lash et al. (1999a) that GSH conjugation
of TCE was inhibited by -50% in the presence of oxidation. Note that this comparison cannot be
made in rats and mice because in vitro kinetic parameters for GSH conjugation in the liver are
not available in those species (only activity at 1 or 2 mM have been measured).
Table 3.3.15. Comparison of hepatic in vitro oxidation and conjugation of TCE
Cellular or
Sub-
Cellular
Fraction
V
* max
(nmol TCE
metabolized/min/g tissue)
Km
(jiM in blood)
V IK
* max' 1 vm
(mL/min/g tissue)
Oxidation
GSH
Conjugation
Oxidation
GSH
Conjugation
Oxidation
GSH
Conjugation
Hepatocytes
10.0-68.4
16-25
22.1-198
16-47
0.087-1.12
0.55-1.0
Liver
microsomes
6.1-111
45
2.66-11. V
5.9a
1.71-28.2a
7.6a
71.0-297b
157b
0.064-1.06b
0.29b
6/22/2009
97
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Liver cytosol
-
380
-
4.5a
-
84a
-
-
22.7b
-
16.7b
Note: When triphasic metabolism was reported, only high affinity pathway is shown here.
Conversion assumptions for Vmax:
Hepatocelluarity of 99 million cells/g liver (Barter et al., 2007);
Liver microsomal protein content of 32 mg protein/g tissue (Barter et al., 2007); and
Liver cytosolic protein content of 89 mg protein/g tissue (based on rats: Prasanna et al., 1989; van
Bree et al., 1990).
Conversion assumptions for Km:
For hepatocytes, Km in headspace converted to Km in blood using blood:air partition coefficient of 9.5
(reported range of measured values 6.5-12.1, Table 3.1.1a);
For microsomal protein, option (a) assumes Km in medium is equal to Km in tissue, and converts to
Km in blood by using a liverblood partition coefficient of 5 (reported ranges of measured values
3.6-5.9, Table 3.2.3), and option (b) converts Km in medium to Km in air using the measured
microsomal protein:air partition coefficient of 1.78 (Lipscomb et al., 1997), and then converts to Km
in blood by using the blood:air partition coefficient of 9.5; and
For cytosolic protein, option (a) assumes Km in medium is equal to Km in tissue, and converts to Km in
blood by using a liverblood partition coefficient of 5 (reported ranges of measured values 3.6-5.9,
Table 3.2.3), and option (b) assumes Km in medium is equal to Km in blood, so no conversion is
necessary.
Furthermore, as shown earlier in Table 3.3.10, the human in vivo data of
Lash et al. (1999a) show blood concentrations of DCVG similar, on a molar basis, to that of
TCE, TCA, or TCOH, suggesting substantial conjugation of TCE. In addition, these data give a
lower limit as to the amount of TCE conjugated. In particular, by multiplying the peak blood
concentration of DCVG by the blood volume, a minimum amount of DCVG in the body at that
time can be derived (i.e., assuming the minimal empirical distribution volume equal to the blood
volume). As shown in Table 3.3.16, this lower limit amounts to about 0.4-3.7% of the inhaled
TCE dose. Since this is the minimum amount of DCVG in the body at a single time point, the
total amount of DCVG formed is likely to be substantially greater owing to possible distribution
outside of the blood as well as the metabolism and/or excretion of DCVG. Lash et al. (1999)
found levels of urinary mercapturates were near or below the level of detection of 0.19 uM,
results that are consistent with those of Bloemen et al. (2001), who reported urinary
concentrations below 0.04 uM at 2- to 4-fold lower cumulative exposures. Taken together, these
results confirm the suggestion by Lash et al. (2000a) that NAcDCVC is a poor quantitative
marker for the flux through the GSH pathway.
6/22/2009
98
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.3.16. Estimates of DCVG in blood relative to inhaled TCE dose in humans exposed
to 50 and 100 ppm (269 and 537 mg/m3; Fisher et al., 1998; Lash et al., 1999)
Sex
Exposure
Estimated Inhaled TCE Dose
(mmol)a
Estimated Peak Amount of DCVG
in Blood (mmol)1
Males
50 ppm x 4 hours
3.53
0.11 ±0.08
100 ppm x 4 hours
7.07
0.26 + 0.08
Females
50 ppm x 4 hours
2.36
0.010 + 0
100 ppm x 4 hours
4.71
0.055 + 0.027
aInhaled dose estimated by (50 or 100 ppm)/(24,450 ppm/mM)*(240 min)*QP, where alveolar
ventilation rate Qp is 7.2 L/min for males and 4.8 1/min for females. Qp is calculated as
(Vx-VD)*fR with the following respiratory parameters: tidal volume Vx (0.75 L for males, 0.46 L
for females), dead space VD (0.15 L for males, 0.12 L for females), and respiration frequency fR
(12 min"1 for males, 14 min"1 for females) (assumed sitting, awake from ICRP , 2002)
bPeak amount of DCVG in blood estimated by multiplying the peak blood concentration by the
estimated blood volume: 5.6 L in males and 4.1 L in females (ICRP, 2002).
In summary, TCE oxidation is likely to be greater quantitatively than conjugation with
GSH in mice, rats, and humans. However, the flux through the GSH pathway, particularly in
humans, may be greater by an order of magnitude or more than the <0.1% typically excreted of
NAcDCVC in urine. This is evidenced both by a direct comparison of in vitro rates of oxidation
and conjugation, as well as by in vivo data on the amount of DCVG in blood. PBPK models can
be used to more quantitatively synthesize these data and put more rigorous limits on relative
amount TCE oxidation and conjugation with GSH. Such analyses are discussed in Section 3.5.
3.4 TCE EXCRETION
This section discusses the major routes of excretion of TCE and its metabolites in
exhaled air, urine, and feces. Unmetabolized TCE is eliminated primarily via exhaled air. As
discussed in Section 3.3, the majority of TCE absorbed into the body is eliminated by
metabolism. With the exception of CO2, which is eliminated solely via exhalation, most TCE
metabolites have low volatility and, therefore, are excreted primarily in urine and feces. Though
6/22/2009
99
-------
1
2
3
4
5
6
7
8
9
10
11
12
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
trace amounts of TCE metabolites have also been detected in sweat and saliva (Bartonicek et al.,
1962), these excretion routes are likely to be relatively minor.
3.4.1 Exhaled Air
In humans, pulmonary elimination of unchanged trichloroethylene and other volatile
compounds is related to ventilation rate, cardiac output, and the solubility of the compound in
blood and tissue, which contribute to final exhaled air concentration of TCE. In their study of
the impact of workload on TCE absorption and elimination, Astrand and Ovrum (1976)
characterized the post-exposure elimination of TCE in expired breath. TCE exposure (540 or
1080 mg/m3; 100 or 200 ppm) was for a total of 2 hours, at workloads from 0 to 150 Watts.
Elimination profiles were roughly equivalent among groups, demonstrating a rapid decline in
TCE concentrations in expired breath post-exposure (Table 3.4.1).
6/22/2009
100
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.4.1. Concentrations of TCE in expired breath from inhalation-exposed humans
(Astrand and Ovrum, 1976)
Time
Postexposure
Alveolar Air
I*
II
III
0
459 + 44
244+ 16
651 +53
30
70 + 5
51 + 3
105 + 18
60
40 + 4
28 + 2
69 + 8
90
35 + 9
21 + 1
55+2
120
31+8
16+ 1
45 + 1
300
8 + 1
9 + 2
14 + 2
420
5 + 0.5
4 + 0.5
8 + 1.3
19 hours
2 + 0.3
2 + 0.2
4 + 0.5
* Roman numerals refer to groups assigned different workloads.
Concentrations are in mg/m3 for expired air.
The lung clearance of TCE represents the volume of air from which all TCE can be
removed per unit time, and is a measure of the rate of excretion via the lungs.
Monster et al. (1976) reported lung clearances ranging from 3.8 to 4.9 1/min in four adults
exposed at rest to 70 ppm and 140 ppm of trichloroethylene for four hours. Pulmonary
ventilation rates in these individuals at rest ranged from 7.7-12.3 1/min. During exercise, when
ventilation rates increased to 29-30 1/min, lung clearance was correspondingly higher, 7.7-12.3
1/min. Under single and repeated exposure conditions, Monster et al. (1976, 1979) reported from
7%-17% of absorbed TCE excreted in exhaled breath.
Pulmonary elimination of unchanged trichloroethylene at the end of exposure is a
first-order diffusion process across the lungs from blood into alveolar air, and it can be thought
of as the reversed equivalent of its uptake from the lungs. Exhaled pulmonary excretion occurs
in several distinct (delayed) phases corresponding to release from different tissue groups, at
different times. Sato et al. (1977) detected 3 first-order phases of pulmonary excretion in the
first 10 hours after exposure to 100 ppm for 4 hours, with fitted half-times of pulmonary
elimination of 0.04 hr, 0.67 hr, and 5.6 hr, respectively. Opdam (1989) sampled alveolar air up
to 20-310 hours after 29-62 minute exposures to 6-38 ppm, and reported terminal half-lives of
8-44 hr at rest. Chiu et al. (2007) sampled alveolar air up to 100 hr after 6-hour exposures to 1
ppm and reported terminal half-lives of 14-23 hr. The long terminal half-time of TCE
pulmonary excretion indicates that a considerable time is necessary to completely eliminate the
compound, primarily due to the high partitioning to adipose tissues (see Section 3.2).
As discussed above, several studies (Dekant et al. 1984, Dekant et al. 1986a, Green and
Prout 1985, Prout et al. 1985) have investigated the disposition of [14CJ-TCE in rats and mice
following gavage administrations (see Section 3.3.2). These studies have reported C02 as an
6/22/2009
101
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
exhalation excretion product in addition to unchanged TCE. With low doses, the amount of TCE
excreted unchanged in exhaled breath is relatively low. With increasing dose in rats, a
disproportionately increased amount of radiolabel is expired as unchanged TCE. This may
indicate saturation of metabolic activities in rats at doses 200 mg/kg and above, which is perhaps
only minimally apparent in the data from mice. In addition, exhaled air TCE concentration has
been measured after constant inhalation exposure for 2 hours to 50 or 500 ppm in rats
(Dallas et al., 1991), and after dermal exposure in rats and humans (Poet, 2000). Exhaled TCE
data from rodents and humans have been integrated into the PBPK model presented in Section
3.5.
Finally, TCOH is also excreted in exhaled breath, though at a rate about 10,000-fold
lower than unmetabolized TCE (Monster et al. 1976, 1979).
3.4.2 Urine
Urinary excretion after TCE exposure consists predominantly of the metabolites TCA
and TCOH, with minor contributions from other oxidative metabolites and GSH conjugates.
Measurements of unchanged TCE in urine have been at or below detection limits (e.g.,
Fisher et al. 1998, Chiu et al. 2007). The recovery of urinary oxidative metabolites in mice, rats,
and humans was addressed earlier (see section 3.3.2) and will not be discussed here.
Because of their relatively long elimination half-life, urinary oxidative metabolites have
been used as an occupational biomarker of TCE exposure for many decades
(Ikeda and Imamura 1973, Carrieri 2007). Ikeda and Imamura (1973) measured total trichloro
compounds (TTC), TCOH and TCA, in urine over three consecutive post-exposure days for 4
exposure groups totaling 24 adult males and one exposure group comprising 6 adult females.
The elimination half-life for TTC ranged 26.1 to 48.8 hours in males and was 50.7 hours in
females. The elimination half-life for TCOH was 15.3 hours in the only group of males studied
and was 42.7 hours in females. The elimination half-life for TCA was 39.7 hours in the only
group of males studied and was 57.6 hours in females. These authors compared their results to
previously published elimination half-lives for TTC, TCOH, and TCA. Following experimental
exposures of groups of 2 to 5 adults, elimination half-lives ranged 31-50 hours for TTC; 19-29
hours for TCOH; and 36-55 hours for TCA (Bartonicek, 1962; Stewart et al., 1970; Nomiyama
and Nomiyama, 1971; Ogata et al., 1971). The urinary elimination half-life of TCE metabolites
in a subject who worked with and was addicted to sniffing TCE for 6-8 years approximated 49.7
hours for TCOH, 72.6 hours for TCA, and 72.6 hours for TTC (Ikeda et al., 1971).
The quantitative relationship between urinary concentrations of oxidative metabolites and
exposure in an occupational setting was investigated by Ikeda (1977). This study examined the
6/22/2009
102
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
urinary elimination of TCE and metabolites in urine of 51 workers from 10 workshops. The
concentration of TCA and TCOH in urine demonstrated a marked concentration-dependence,
with concentrations of TCOH being approximately twice as high as those for TCA. Urinary
half-life values were calculated for 6 males and 6 females from 5 workshops; males were
intermittently exposed to 200 ppm and females were intermittently exposed to 50 ppm (269
mg/m3). Urinary elimination half-lives for TTC, TCOH and TCA were 26.1, 15.3, and 39.7
hours; and 50.7, 42.7 and 57.6 hours in males and females, respectively, which were similar to
the range of values previously reported. These authors estimated that urinary elimination of
parent TCE during exposure might account for one-third of the systemically absorbed dose.
Importantly, urinary TCA exhibited marked saturation at exposures higher than 50 ppm.
Because TTC nor urinary TCOH (in the form of the glucuronide TCOG) showed such an effect,
this saturation cannot be due to TCE oxidation itself, but must rather be from one of the
metabolic processes forming TCA from TCOH. Unfortunately, since biological monitoring
programs usually measure only urinary TCA, rather than TTC, urinary TCA levels above around
150 mg/1 cannot distinguish between exposures at 50 ppm and at much higher concentrations.
It is interesting to attempt to extrapolate on a cumulative exposure basis the Ikeda (1977)
results for urinary metabolites obtained after occupational exposures at 50 ppm to the controlled
exposure study by Chiu et al. (2007) at 1.2 ppm for 6 hours (the only controlled exposure study
for which urinary concentrations, rather than only cumulative excretion, are available). Ikeda
(1977) reported that measurements were made during the second half of the week, so one can
postulate a cumulative exposure duration of 20-40 hours. At 50 ppm, Ikeda (1977) report a
urinary TCOH concentration of about 290 mg/1, so that per ppm-hr, the expected urinary
concentration would be 290/(50 x 20-40) = 0.145-0.29 mg/l-ppm-hr. The cumulative exposure
in Chiu et al. (2007) is 1.2 x 6 = 7.2 ppm-hr, so the expected urinary TCOH concentration would
be 7.2 x (0.145-0.29) = 1.0-2.1 mg/1. This estimate is somewhat surprisingly consistent with the
actual measurements of Chiu et al. (2007) during the first day post-exposure, which ranged from
0.8-1.2 mg/1 TCOH in urine.
On the other hand, extrapolation of TCA concentrations was less consistent. At 50 ppm,
Ikeda (1977) report a urinary TCA concentration of about 140 mg/1, so that per ppm-hr, the
expected urinary concentration would be 140/(50 x 20-40) = 0.07-0.14 mg/l-ppm-hr. The
cumulative exposure in Chiu et al. (2007) is 1.2 x 6 = 7.2 ppm-hr, so the expected urinary TCA
concentration would be 7.2 x (0.07-0.14) = 0.5-1.0 mg/1, whereas Chiu et al. (2007) reported
urinary TCA concentrations on the first day after exposure of 0.03-0.12 mg/1. However, as
noted in Chiu et al. (2007), relative urinary excretion of TCA was 3- to 10-fold lower in Chiu et
al. (2007) than other studies at exposures 50-140 ppm, which may explain part of the
discrepancies. However, this may be due in part to saturation of many urinary TCA
6/22/2009
103
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
measurements, and, furthermore, inter-individual variance, observed to be substantial in Fisher et
al. (1998), cannot be ruled out.
Urinary elimination kinetics have been reported to be much faster in rodents than in
humans. For instance, adult rats were exposed to 50, 100, or 250 ppm (269, 537, or 1344 mg/m3)
via inhalation for 8 hours or were administered an i.p. injection (1.47 g/kg) and the urinary
elimination of total trichloro compounds was followed for several days (Ikeda and Imamura,
1973). These authors calculated urinary elimination half-lives of 14.3-15.6 hours for female rats
and 15.5-16.6 hours for male rats; the route of administration did not appear to influence half-
life value. In other rodent experiments using orally administered radiolabeled TCE, urinary
elimination was complete within one or two days after exposure (Dekant et al. 1984, Dekant et
al. 1986a, Green and Prout 1985, Prout et al. 1985).
3.4.3 Feces
Fecal elimination accounts for a small percentage of TCE as shown by limited
information in the available literature. Bartonicek (1962) exposed 7 human volunteers to 1.042
mg TCE/L air for 5 hours and examined TCOH and TCA in feces on the third and seventh day
following exposure. The mean amount of TCE retained during exposure was 1107 mg,
representing 51-64% (mean 58%) of administered dose. On the third day following TCE
exposure, TCOH and TCA in feces demonstrated mean concentrations of 17.1 and 18.5 mg/100
grams feces, similar to concentrations in urine. However, because of the 10-fold smaller daily
rate of excretion of feces relative to urine, this indicates fecal excretion of these metabolites is
much less significant than urinary excretion. Neither TCOH nor TCA was detected in feces on
the seventh day following exposure.
In rats and mice, total radioactivity has been used to measure excretion in feces after oral
gavage TCE administration in corn oil, but since the radiolabel was not characterized it is not
possible to determine whether the fecal radiolabel in feces represented unabsorbed parent
compound, excreted parent compound, and/or excreted metabolites. Dekant et al. (1984)
reported mice eliminated 5% of the total administered TCE, while rats eliminated 2% after oral
gavage. Dekant et al., 1986a reported a dose response related increase in fecal elimination with
dose, ranging between 0.8-1.9%) in rats and 1.6-5% in mice after oral gavage in corn oil. Due to
the relevant role of CYP2E1 in the metabolism of TCE (Section 3.3.3.1.6), Kim and Ghanayem,
2006 compared fecal elimination in both wild type and CYP2E1 knockouts mice and reported
fecal elimination ranging between 4.1-5.2%) in wild type and 2.1-3.8%) in knockout mice
exposed by oral gavage in aqueous solution.
6/22/2009
104
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 References for 3.1-3.4
2 Abbas, R; Fisher, JW. (1997) A physiologically based pharmacokinetic model for
3 trichloroethylene and its metabolites, chloral hydrate, trichloroacetate, dichloroacetate,
4 trichloroethanol, and trichloroethanol glucuronide in B6C3F1 mice. Toxicol Appl Pharm
5 147:15-30.
6 Abraham, D.G., P.P. Patel and A.J.L. Cooper. 1995a. Isolation from rat kidney of a cytosolic
7 high molecular weight cysteine-»Y-conjugate P-lyase with activity toward leukotriene E4. J. Biol.
8 Chem. 270(1): 180-188.
9 Abraham, D.G., R.J. Thomas and A.J.L. Cooper. 1995b. Glutamine transaminase K is not a
10 major cysteine ^-conjugate P-lyase of rat kidney mitochondria: Evidence that a high-molecular
11 weight enzyme fulfills this role. Am. Soc. J Pharmacol. Exp. Ther 48:855-860.
12 Adachi, H., Kubota, I., Okamura, N., Iwata, H., Tsujimoto, M., Nakazato, H., Nishihara, T., and
13 Noguchi, T. (1989). Purification and characterization of human microsomal dipeptidase. J
14 Biochem. 105(6):957-61.
15
16 Alberati-Giani, D., P. Malherbe, C. Kohler et al. 1995. Cloning and characterization of a
17 soluble kynurenine aminotransferase from rat brain: Identity with kidney cysteine conjugate
18 P-lyase. J. Neurochem. 64:1448-1455.
19 Altuntas, TG; Park, SB; Kharasch, ED. (2004) Sulfoxidation of cysteine and mercapturic acid
20 conjugates of the sevoflurane degradation product fluoromethyl-2,2-difluoro-l-
21 (trifluoromethyl)vinyl ether (compound A). Chem Res Toxicol 17(3):435-445.
22 Anders, M.W. and I. Jakobson. 1985. Biotransformation of halogenated solvents. Scand. J.
23 Work. Env. Hea. ll(Suppl. l):23-32.
24 Andersen, M. E., Gargas, M. L., Clewell, H. J., 3rd, and Severyn, K. M. (1987). Quantitative
25 evaluation of the metabolic interactions between trichloroethylene and 1,1-dichloroethylene in
26 vivo using gas uptake methods. Toxicol Appl Pharmacol 89, 149-157.
27 Astrand, I; Ovrum, P. (1976) Exposure to trichloroethylene uptake and distribution in man.
28 Scand J Work Env Hea 2(4): 199-211.
6/22/2009
105
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Banki, K; Elfarra, AA; Lash, LH; Anders, MW. (1986) Metabolism of S-(2-chloro-1,1,2-
2 trifluoroethyl)-l-cysteine to hydrogen sulfide and the role of hydrogen sulfide in »S-(2-chloro-
3 l,l,2-trifluoroethyl)-l-cysteine-induced mitochondrial toxicity. Biochem Bioph Res Co
4 138:707-713.
5 Banneijee, S; van Duuren, BL (1978) Covalent binding of the carcinogen trichloroethylene to
6 hepatic microsomal proteins and to exogenous DNA in vitro. Cancer Res 38:776-780.
7 Barter, Z. E., Bayliss, M. K., Beaune, P. H., Boobis, A. R., Carlile, D. J., Edwards, R. J.,
8 Houston, J. B., Lake, B. G., Lipscomb, J. C., Pelkonen, O. R., Tucker, G. T., and Rostami-
9 Hodjegan, A. (2007). Scaling factors for the extrapolation of in vivo metabolic drug clearance
10 from in vitro data: reaching a consensus on values of human microsomal protein and
11 hepatocellularity per gram of liver. Curr DrugMetab 8, 33-45.
12 Bartonicek V. (1962) Metabolism and excretion of trichloroethylene after inhalation by human
13 subjects. Brit J Ind Med. 19:134-141.
14 Barton, HA; Creech, JR; Godin, CS; et al. (1995) Chloroethylene mixtures: pharmacokinetic
15 modeling and in vitro metabolism of vinyl chloride, trichloroethylene, and
16 trans-1,2-dichloroethylene in rat. Toxicol Appl Pharm 130(2):237-247.
17 Bartonicek, V. (1962) Metabolism and excretion of trichloroethylene after inhalation by human
18 subjects. Brti JIndustrMed 19:134-141.
19 Bernauer, U; Birner, G; Dekant, W; et al. (1996) Biotransformation of trichloroethene: dose-
20 dependent excretion of 2,2,2-trichloro-metabolites and mercapturic acids in rats and humans
21 after inhalation. Arch Toxicol 70(6):338-346.
22 Birner, G; Vamvakas, S; Dekant, W; et al. (1993) Nephrotoxic and genotoxic
23 A-acety 1 -»S'-dichloroviny 1 -L-cysteine is a urinary metabolite after occupational
24 1,1,2-trichloroethene exposure in humans: implications for the risk of trichloroethene exposure.
25 Environ Health Persp 99:281-284.
26 Birner, G; Richling, C; Henschler, D; Anders, MW; Dekant, W. (1994) Metabolism of
27 tetrachloroethene in rats: identification of N epsilon-(dichloroacetyl)-L-lysine and N epsilon-
28 (trichloroacetyl)-L-lysine as protein adducts. Chem Res Toxicol. 7(6):724-32.
6/22/2009
106
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Birner, G., Rutkowska, A., and Dekant, W. (1996). N-acetyl-S-(l,2,2-trichlorovinyl)-L-cysteine
2 and 2,2,2-trichloroethanol: two novel metabolites of tetrachloroethene in humans after
3 occupational exposure. DrugMetab Dispos 24, 41-48.
4 Birner, G., Bernauer, U., Werner, M., and Dekant, W. (1997). Biotransformation, excretion and
5 nephrotoxicity of haloalkene-derived cysteine S-conjugates. Arch Toxicol. 72(1): 1-8.
6
7 Birner, G; Werner, M; Rosner, E; Mehler, C; Dekant, W. (1998) Biotransformation, excretion,
8 and nephrotoxicity of the hexachlorobutadiene metabolite (E)-N-acetyl-S-(l,2,3,4, 4-
9 pentachlorobutadienyl)-L-cysteine sulfoxide. Chem Res Toxicol 11(7):750-777.
10 Bloemen, LJ; Monster, AC; Kezic, S; et al. (2001) Study on the cytochrome P-450- and
11 glutathione-dependent biotransformation of trichloroethylene in humans. Int Arch Occ Env Hea
12 74(2): 102-108.
13 Bogen, KT; Colston, BW, Jr; MaChicao, LK. (1992) Dermal absorption of dilute aqueous
14 chloroform, trichloroethylene, and tetrachloroethylene in hairless guinea pigs. Fund Appl
15 Toxicol 18(l):30-39.
16 Bolt, HM; Filser, JG. (1977) Irreversible binding of chlorinated ethylenes to macromolecules.
17 Environ Health Persp 21:107-112.
18 Bronley-DeLancey, A; McMillan, DC; McMillan, JM; et al. (2006) Application of cryopreserved
19 human hepatocytes in trichloroethylene risk assessment: relative disposition of chloral hydrate to
20 trichloroacetate and trichloroethanol. Environ Health Persp 114(8): 1237-1242.
21 Brown, R. P., Delp, M. D., Lindstedt, S. L., Rhomberg, L. R., and Beliles, R. P. (1997).
22 Physiological parameter values for physiologically based pharmacokinetic models. ToxicolInd
23 Health 13, 407-484.
24 Bruning, T., Vamvakas, S., Makropoulos, V., and Birner, G. (1998). Acute intoxication with
25 trichloroethene: clinical symptoms, toxicokinetics, metabolism, and development of biochemical
26 parameters for renal damage. ToxicolSci 41, 157-165.
27 Bull, RJ. (2000) Mode of action of liver tumor induction by trichloroethylene and its metabolites,
28 trichloroacetate and dichloroacetate. Environ Health Persp 108(Suppl 2):241-259.
6/22/2009
107
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Butler, T.C. 1949. Metabolic transformations of trichloroethylene. Presented in part at the 39th
2 Annual Meeting of the American Society for Pharmacology and Experimental Therapeutics,
3 April 22, 1949. Fed. Proc. 8:278.
4 Cai, H; Guengerich, FP. (1999) Mechanisms of aqueous decomposition of trichloroethylene
5 oxide. J Am Chem Soc 121:11656-11663.
6 Cai, H. and F.P. Guengerich. 2001. Reaction of trichloroethylene and trichloroethylene oxide
7 with cytochrome P450 enzymes: Inactivation and sites of modification. Chem. Res. Toxicol.
8 14:451-458.
9 Carrieri, M., Magosso, D., Piccoli, P., Zanetti, E., Trevisan, A., and Bartolucci, G. B. (2007).
10 Acute, nonfatal intoxication with trichloroethylene. Arch Toxicol 81, 529-532.
11 Chieco, P, Moslen, MT, and Reynolds, ES. 1981. Effect of administration vehicle on oral
12 1,1-dichloroethane toxicity. Toxicol Appl Pharm 57:146-155.
13 Chiu, W. A., Okino, M. S., Lipscomb, J. C., and Evans, M. V. (2006). Issues in the
14 pharmacokinetics of trichloroethylene and its metabolites. Environ Health Perspect 114, 1450-
15 1456.
16 Chiu, W. A., Micallef, S., Monster, A. C., and Bois, F. Y. (2007). Toxicokinetics of inhaled
17 trichloroethylene and tetrachloroethylene in humans at 1 ppm: empirical results and comparisons
18 with previous studies. Toxicol Sci 95, 23-36.
19 Clewell, HJ III; Gentry, PR; Covington, TR; Gearhart, JM. (2000) Development of a
20 physiologically based pharmacokinetic model of trichloroethylene and its metabolites for use in
21 risk assessment. Environ Health Persp 108(Suppl 2):283-305.
22 Commandeur, JN; Vermeulen, NP. (1990) Identification of A-acetyl(2,2-dichlorovinyl)- and
23 A-acetyl(1,2-dichloroviny 1 )-L-cysteine as two regioisomeric mercapturic acids of
24 trichloroethylene in the rat. Chem Res Toxicol 3(3):212-218.
25 Cooper, AJ. (2004) The role of glutamine transaminase k (GTK) in sulfur and alpha-keto acid
26 metabolism in the brain, and in the possible bioactivation of neurotoxicants. Neurochem Int
27 44(8):557-577.
6/22/2009
108
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Coopman, VAE; Cordonnier, JACM; De Letter, EA; Piette, MHA. (2003) Tissue distribution of
2 trichloroethylene in a case of accidental acute intoxication by inhalation. Forensic Sci Int
3 134:115-119.
4 Cummings, BS; Lash, LH. (2000) Metabolism and toxicity of trichloroethylene and
5 ^-(l^-dichloroviny^-L-cysteine in freshly isolated human proximal tubular cells. Toxicol Sci
6 53(2):458-466.
7 Cummings, BS; Parker, JC; Lash, LH. (2000a) Role of cytochrome P450 and glutathione
8 ^-transferase alpha in the metabolism and cytotoxicity of trichloroethylene in rat kidney.
9 Biochem Pharmacol 59(5): 531-543.
10 Cummings, BS; Zangar, RC; Novak, RF; et al. (2000b) Cytotoxicity of trichloroethylene and
11 £-(1, 2-dichlorovinyl)-L-cysteine in primary cultures of rat renal proximal tubular and distal
12 tubular cells. Toxicology 150(1-3) :83-98.
13 Cummings, B.S., J.C. Parker and L.H. Lash. 2001. Cytochrome P450-dependent metabolism of
14 trichloroethylene in rat kidney. Toxicol. Sci. 60:11-19.
15 D'Souza, RW; Bruckner, JV; Feldman, S. (1985) Oral and intravenous trichloroethylene
16 pharmacokinetics in the rat. J Toxicol Env Health 15(5): 587-601.
17 Dallas, CE; Gallo, JG; Ramanathan R, Muralidhara, S; Bruckner, JV. 1991. Physiological
18 pharmacokinetic modeling of inhaled trichloroethylene in rats. Toxicol Appl Pharm 110:303-
19 314.
20 Damani, LA; Houdi, AA. (1988) Cytochrome P-450 and FAD-monoxygenase mediated S- and
21 N-oxygenations. Drug Metabol Drug Interact 6(3-4):235-244.
22 Davis, RR; Murphy, WJ; Snawder, JE; Striley, CA; Henderson, D; Khan, A., Krieg, EF. (2002)
23 Susceptibility to the ototoxic properties of toluene is species specific. Hearing Res 166(l-2):24-
24 32.
25 De Baere, S; Meyer, E; Dirinck, I; Lambert, W; Piette, M; Van Peteghem, C; De Leenheer, A.
26 (1997) Tissue distribution of trichloroethylene and its metabolites in a forensic case. J Anal
27 Toxicol 21:223-227.
6/22/2009
109
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Dehon, B; Humbert, L; Devisme, M; Stievenart, M; Mathieu, D; Houdret, N; Lhermitte, M.
2 (2000) Tetrachloroethylene and trichloroethylene fatality: case report and simple headspace
3 SPME-capillary gas chromatographic determination in tissues. J Anal Toxicol 24:22-26.
4 Dekant, W; Metzler, M; Henschler, D. (1984) Novel metabolites of trichloroethylene through
5 dechlorination reactions in rats, mice and humans. Biochem Pharmacol 33(13):2021 -2027.
6 Dekant, W; Schulz, A; Metzler, M; et al. (1986a) Absorption, elimination and metabolism of
7 trichloroethylene: a quantitative comparison between rats and mice. Xenobiotica 16(2): 143-152.
8 Dekant, W; Metzler, M; Henschler, D. (1986b) Identification of ^-l^-dichlorovinyl-iV-acetyl-
9 cysteine as a urinary metabolite of trichloroethylene: a possible explanation for its
10 nephrocarcinogenicity in male rats. Biochem Pharmacol 35(15):2455-2458.
11 Dekant, W., K. Berthold, S. Vamvakas, D. Hencschler and M.W. Anders. 1988. Thioacylating
12 intermediates as metabolites of »S'-(1,2-dichlorovinyl)-L-cysteine and S-( 1,2,2-trichlorovinyl)-L-
13 cysteine formed by cysteine conjugate P-lyase. Chem. Res. Toxicol. 1:175-178.
14 Dekant, W., M. Koob and D. Henschler. 1990. Metabolism of trichloroethene - in vivo and in
15 vitro evidence for activation by glutathione conjugation. Chem. Biol. Interact. 73:89-101.
16 Dix, K. J., Kedderis, G. L., and Borghoff, S. J. (1997). Vehicle-dependent oral absorption and
17 target tissue dosimetry of chloroform in male rats and female mice. Toxicol Lett 91, 197-209.
18 Dohn, D.R. and M.W. Anders. 1982. Assay of cysteine conjugate P-lyase activity with
19 £-(2-benzothiazolyl)cysteine as the substrate. Anal. Biochem. 120:379-386.
20 Dowsley, TF; Ulreich, JB; Bolton, JL; et al. (1996) CYP2E1-dependent bioactivation of
21 1,1-dichloroethylene in murine lung: formation of reactive intermediates and glutathione
22 conjugates. Toxicol Appl Pharm 139(l):42-48.
23 Duffell, M. W., and Jakobi, W. B. (1982). Cysteine S-Conjugate N-Acetyltransferase from Rat
24 Kidney Microsomes. Molec. Pharmacol. 21, 444-448.
25 Elfarra, A. A. and M.W. Anders. 1984. Commentary. Renal processing of glutathione
26 conjugates role in nephrotoxicity. Biochem. Pharmacol. 33(23):3729-3732.
6/22/2009
110
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Elfarra, A.A. and I.Y. Hwang. 1990. In vivo metabolites of »S'-(2-benzothiozolyl)-L-cysteine as
2 markers of in vivo cysteine conjugate P-lyase and thiol glucuronosyl transferase activities. Drug
3 Metab. Dispos. 18(6):917-922.
4 Elfarra, AA; Jakobson, I; Anders, MW. (1986a) Mechanism of »S'-(1,2-dichlorovinyl)glutathione-
5 induced nephrotoxicity. Biochem Pharmacol 35(2) :283-288.
6 Elfarra, A.A., L.H. Lash and M.W. Anders. 1986b. a-Ketoacids stimulate rat renal cysteine
7 conjugate P-lyase activity and potentiate the cytotoxicity of »S'-(1,2-dichlorovinyl)-L-cysteine.
8 Mol Pharmacol 31:208-212.
9 Elfarra, AA; Krause, RJ; Last, AR; et al. (1998) Species- and sex-related differences in
10 metabolism of trichloroethylene to yield chloral and trichloroethanol in mouse, rat, and human
11 liver microsomes. Drug Metab Dispos 26(8):779-785.
12 Eyre, RJ; Stevens, DK; Parker, JC; et al. (1995a) Acid-labile adducts to protein can be used as
13 indicators of the cysteine ^-conjugate pathway of trichloroethene metabolism. J Toxicol Env
14 Health 46(4):443-464.
15 Eyre, R.J., D.K. Stevens, J.C. Parker and R.J. Bull. 1995b. Renal activiation of trichloroethene
16 and »S'-(1,2-dichlorovinyl)-L-cysteine and cell proliferative responses in the kidneys of F344 rats
17 and B6C3F1 mice. J. Toxicol. Environ. Health. 46:465-481.
18 Fairman, MJ, Scott, BB, Toothill, C, and Losowsky, MS. (1977). Fairman MJ, Scott BB,
19 Toothill C, Losowsky MS. Gut. 18(6):484-7.
20
21 Filser, JG; Bolt, HM. (1979) Pharmacokinetics of halogenated etylenes in rats. Arch Toxicol
22 42:123-136.
23 Fernandez, J. G., Droz, P. O., Humbert, B. E., and Caperos, J. R. (1977). Trichloroethylene
24 exposure. Simulation of uptake, excretion, and metabolism using a mathematical model. Br JInd
25 Med 34, 43-55.
26 Fiserova-Bergerova, V; Tichy, M; Di Carlo, FJ. (1984) Effects of biosolubility on pulmonary
27 uptake and disposition of gases and vapors of lipophilic chemicals. Drug Metab Rev 15(5—
28 6): 1033-1070.
6/22/2009
111
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Fisher, JW; Whittaker, TA; Taylor, DH, Clewell, HJ 3rd; Andersen, ME. (1989) Physiologically
2 based pharmacokinetic modeling of the pregnant rat: a multiroute exposure model for
3 trichloroethylene and its metabolite, trichloroacetic acid. Toxicol Appl Pharm 99(3):395-414.
4 Fisher, JW; Whittaker, TA; Taylor, DH; Clewell, HJ, III; Andersen, ME. (1990) Physiologically
5 based pharmacokinetic modeling of the lactating rat and nursing pup: a multiroute exposure
6 model for trichloroethylene and its metabolite, trichloroacetic acid. Toxicol Appl Pharm
7 102:497-513.
8 Fisher, JW; Gargas, ML; Allen, BC; Andersen, ME. 1991. Physiologically based
9 pharmacokinetic modeling with trichloroethylene and its metabolite, trichloroacetic acid, in the
10 rat and mouse. Toxicol Appl Pharm 109:183-195.
11 Fisher, JW; Mahle, D; Abbas, R. (1998) A human physiologically based pharmacokinetic model
12 for trichloroethylene and its metabolites, trichloroacetic acid and free trichloroethanol. Toxicol
13 Appl Pharm 152:339-359.
14 Ford, ES; Rhodes, S; McDiarmid, M; Schwartz, SL; Brown, J. (1995) Deaths from acute
15 exposure to trichloroethylene. J Occup Environ Med 37(6):750-754.
16 Forkert, PG. (1999a) 1,1-Dichloroethylene-induced Clara cell damage is associated with in situ
17 formation of the reactive epoxide. Immunohistochemical detection of its glutathione conjugate.
18 Am JResp Cell Mol Biol 20(6): 1310-1318.
19 Forkert, PG. (1999b) In vivo formation and localization of 1,1-dichloroethylene epoxide in
20 murine liver: identification of its glutathione conjugate 2-»S-glutathionyl acetate. J Pharmacol
21 Exp Ther 290(3): 1299-1306. Please verify a and b labeled correctly vs. citations in text
22 Forkert, PG; Malkinson, AM; Rice, P; et al. (1999) Diminished CYP2E1 expression and
23 formation of 2-»S'-glutathionyl acetate, a glutathione conjugate derived from 1,1-dichloroethylene
24 epoxide, in murine lung tumors. Drug Metab Dispos 27(l):68-73.
25 Forkert, P-G., R.M. Baldwin, B. Millen et al. 2005. Pulmonary bioactivation of
26 trichloroethylene to chloral hydrate: Relative contributions of CYP2E1, CYP2F, and CYP2B1.
27 Drug Metab. Disp. 33(10):1429-1437.
28 Forkert, PG; Millen, B; Lash, LH; Putt, DA, Ghanayem, BI. (2006) Pulmonary bronchiolar
29 cytotoxicity and formation of dichloroacetyl lysine protein adducts in mice treated with
30 trichloroethylene. J Pharmacol Exp Ther 316(2):520-529.
6/22/2009
112
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Gargas, M. L., Seybold, P. G., and Andersen, M. E. (1988). Modeling the tissue solubilities and
2 metabolic rate constant (Vmax) of halogenated methanes, ethanes, and ethylenes. Toxicol Lett
3 43,235-256.
4 Gargas, ML; Burgess, RJ; Voisard, DE; Cason, GH; Andersen, ME. (1989) Partition coefficients
5 of low-molecular-weight volatile chemicals in various liquids and tissues. Toxicol Appl Pharm
6 98(l):87-99.
7 Goeptar, AR; Commandeur, JN; van, OB; et al. (1995) Metabolism and kinetics of
8 trichloroethylene in relation to toxicity and carcinogenicity. Relevance of the mercapturic acid
9 pathway. Chem Res Toxicol 8(1):3-21.
10 Gonzalez-Leon, A; Merdink, JL; Bull, RJ; Schultz, IR. (1999) Effect of pre-treatment with
11 dichloroacetic or trichloroacetic acid in drinking water on the pharmacokinetics of a subsequent
12 challenge dose in B6C3F1 mice. Chem Biol Interact 123(3):239-253.
13 Green, T; Prout, MS. (1985) Species differences in response to trichloroethylene. II.
14 Biotransformation in rats and mice. Toxicol Appl Pharm 79(3):401-411.
15 Green, T., J. Odum, J.A. Nash and J.R. Foster. 1990. Perchloroethylene-induced rat kidney
16 tumors: An investigation of the mechanisms involved and their relevance to humans. Toxicol.
17 Appl. Pharm. 103:77-89.
18 Green, T; Dow, J; Ellis, MK; et al. (1997a) The role of glutathione conjugation in the
19 development of kidney tumours in rats exposed to trichloroethylene. Chem Biol Interact
20 105(2):99-117.
21 Green, T., G.W. Mainwaring and J.R. Foster. 1997b. Trichloroethylene-induced mouse lung
22 tumors: Studies of the mode of action and comparisons between species. Fund. Appl. Toxicol.
23 37:125-130.
24 Greenberg, M. S., Burton, G. A., and Fisher, J. W. (1999). Physiologically based
25 pharmacokinetic modeling of inhaled trichloroethylene and its oxidative metabolites in B6C3F1
26 mice. Toxicol Appl Pharmacol 154, 264-278.
27 Guengerich, FP. (2004) Metabolism of trichloroethylene and covalent binding of reaction
28 products. In: Proceedings of symposium on new scientific research related to the health effects
29 of trichloroethylene, Feb. 26-27, 2004, Washington, DC. Available online at
3 0 http ://cfpub. epa.gov/ncea/cfm/recordi splay. cfm? deid=75 93 4.
6/22/2009
113
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Guengerich, F.P. and T. Shimada. 1991. Oxidation of toxic and carcinogenic chemicals by
2 human cytochrome P-450 enzymes. Chem. Res. Toxicol. 4(4):391-407.
3 Guengerich, FP; Kim, DH; Iwasaki, M. (1991) Role of human cytochrome P-450 IIE1 in the
4 oxidation of many low molecular weight cancer suspects. Chem Res Toxicol 4(2): 168-179.
5 Habib, GM, Barrios, R, Shi, ZZ, and Lieberman, MW. (1996). Four distinct membrane-bound
6 dipeptidase RNAs are differentially expressed and show discordant regulation with gamma-
7 glutamyl transpeptidase. J Biol Chem. 271(27): 16273-80.
8
9 Habib, GM, Shi, ZZ, Cuevas, AA, Guo, Q, Matzuk, MM, and Lieberman, MW. (1998).
10 Leukotriene D4 and cystinyl-bis-glycine metabolism in membrane-bound dipeptidase-deficient
11 mice. Proc Natl Acad Sci U SA. 95(9):4859-63.
12
13 Henderson, G.N., Z. Yan, M.O. James, N. Davydova and P.W. Stacpoole. 1997. Kinetics and
14 metabolism of chloral hydrate in children: Identification of dichloroacetate as a metabolite.
15 235:695-698.
16 Hinchman, C.A. and N. Ballatori. 1990. Glutathione-degrading capacities of liver and kidney in
17 different species. Biochem. Pharmacol. 40(5): 1131-1135.
18 Hissink, E. M., Bogaards, J. J. P., Freidg, A. P., Commandeur, J. N. M., Vermeulen, N. P. E.,
19 and van Bladeren, P. J. (2002). The use of in vitro metabolic parameters and physiologically
20 based pharmacokinetic (PBPK) modeling to explore the risk assessment of trichloroethylene.
21 Environmental Toxicology and Pharmacology 11, 259-271.
22 Hooper, N.M., Hesp, R.J., and Tieku, S.(1994). Metabolism of aspartame by human and pig
23 intestinal microvillar peptidases. Biochem J. 298 Pt 3:635-9.
24
25 Hu, Y., Hakkola, J., Oscarson, M., and Ingelman-Sundberg, M. (1999). Structural and functional
26 characterization of the 5'-flanking region of the rat and human cytochrome P450 2E1 genes:
27 identification of a polymorphic repeat in the human gene. Biochem Biophys Res Commun.
28 263,286-293.
29
30 IARC (International Agency for Research on Cancer). 1995. Dry Cleaning, Some Chlorinated
31 Solvents and Other Industrial Chemicals. Summary of Data Reported and Evaluation. IARC
6/22/2009
114
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Monographs on the Evaluation of Carcinogenic Risks to Humans. Volume 63. International
2 Agency for Research on Cancer, World Health Organization, Lyon, France.
3 ICRP (The International Commission on Radiological Protection). (2002) Basic anatomical and
4 physiological data for use in radiological protection: reference values. A report of age- and
5 gender-related differences in the anatomical and physiological characteristics of reference
6 individuals. ICRP Publication 89. Ann ICRP. 2002;32(3-4):5-265.
7 Ikeda, M; Imamura, T. (1973) Biological half-life of trichloroethylene and tetrachloroethylene in
8 human subjects. Int Arch Arbeitsmed 31:209-224.
9 Ikeda, M; Ohtsuji, H; Kawai, H; Kuniyoshi, M. (1971) Excretion kinetics of urinary metabolites
10 in a patient addicted to trichloroethylene. Br J Ind Med 28(2):203-206.
11 Ikeda, M. (1977) Metabolism of trichloroethylene and tetrachloroethylene in human subjects.
12 Environ Health Persp 21:239-245.
13 Iyer, R.A., E.J. Frink, T.J. Ebert and M.W. Anders. 1998. Cysteine conjugate P-lyase-dependent
14 metabolism of compound A (2-[fluoromethoxy]-l, 1,3,3,3-pentafluoro-l-propene) in human
15 subjects anesthetized with sevoflurane and in rats given compound A. Anesthesiology.
16 88:611-618.
17 Jakobson, I; Wahlberg, JE; Holmberg, B; Johansson, G. (1982) Uptake via the blood and
18 elimination of 10 organic solvents following epicutaneous exposure of anesthetized guinea pigs.
19 Toxicol Appl Pharm 63:181-187.
20 Jakubowski, M; Wieczorek, H. (1988) The effects of physical effort on pulmonary uptake of
21 selected organic compound vapours. Pol J Occup Med 1(1):62-71.
22 James, M.O., R. Cornett, Z. Yan, G.N. Henderson and P.W. Stacpoole. 1997. Glutathione-
23 dependent conversion to glyoxylate, a major pathway of dichloroacetate biotransformation in
24 hepatic cytosol from humans and rats, is reduced in dichloroacetate-treated rats. Drug Metab.
25 Disp. 25(11): 1223-1227.
26 Jia, M.; Coats, B; Chadha, M; Frentzen, B;Perez-Rodriguez, J; Chadik, PA; Yost, RA;
27 Henderson, GN; Stacpoole, PW. (2006) Human kinetics of orally and intravenously administered
28 low-dose l,2-13C-dichloroacetate. J Clin Pharmacol 46:1449-1459.
6/22/2009
115
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Jones, T.W., C. Qin, V.H. Schaeffer and J.L. Stevens. 1988. Immunohistochemical localization
2 of glutamine transaminase K, a rat kidney cysteine conjugate P-lyase, and the relationship to the
3 segment specificity of cysteine conjugate nephrotoxicity. Mol Pharmacol 34:621-627.
4 Kadry, AM; Turkall, RM; Skowronski, GA; Abdel-Rahman, MS. (1991) Soil adsorption alters
5 kinetics and bioavailability of trichloroethylene in orally exposed female rats. Toxicol Lett
6 58:337-346.
7 Kaneko, T; Wang, P-Y; Sato, A. (1994) Enzymes induced by ethanol differently affect the
8 pharmacokinetics of trichloroethylene and 1,1,1 -trichloroethane. Occup Environ Med 51:113—
9 Kawamoto, T.,T. Hobara, T. Aoki et al. 1988. Metabolism of chloral hydrate in the anoxic
10 perfused liver. Toxicol. Lett. 40:225-231.
11 Ketcha, M.M., D.K. Stevens, D.A. Warren, C.T. Bishop and W.T. Brashear. 1996. Conversion
12 of trichloroacetic acid to dichloroacetic acid in biological samples. J. Anal. Toxicol. 20:236-
13 241.
14 Keys, D. A., Bruckner, J. V., Muralidhara, S., and Fisher, J. W. (2003). Tissue dosimetry
15 expansion and cross-validation of rat and mouse physiologically based pharmacokinetic models
16 for trichloroethylene. Toxicol Sci 76, 35-50.
17 Keys, DA; Schultz, IR; Mahle, DA; Fisher, JW. (2004) A quantitative description of suicide
18 inhibition of dichloroacetic acid in rats and mice. Toxicol Sci. 82(2):381-393.
19 Kezic, S; Monster, AC; Kriise, J; Verberk, MM. (2000) Skin absorption of some vaporous
20 solvents in volunteers. Int Arch Occ Env Hea 73:415-422.
21 Kezic, S; Monster, AC; van de Gevel IA; Kruse, J; Opdam, JJG; Verberk, MM. (2001) Dermal
22 absorption of neat liquid solvents on brief exposures in volunteers. Am Ind Hyg Assoc J
23 62(1): 12—18.
24 Kim, D; Ghanayem, BI; (2006) Comparative metabolism and disposition of trichloroethylene in
25 Cyp2el-l- and wild-type mice. Drug Metab Dispos 34(12):2020-2027.
26 Kim, HJ, Bruckner, JV, Dallas, CE and Gallo, JM. 1990. Effect of dosing vehicles on the
27 pharmacokinetics of orally administered carbon tetrachloride in rats. Toxicol Appl Pharm
28 102:50-60.
6/22/2009
116
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Koizumi, A. (1989) Potential of physiologically based pharmacokinetics to amalgamate kinetic
2 data of trichloroethylene and tetrachloroethyolene obtained in rats and man. Br J Ind Med
3 46(4):23 9-249.
4 Kozak, EM, and Tate, SS. (1982). Glutathione-degrading enzymes of microvillus membranes. J
5 BiolChem. 257(11):6322-7.
6
7 Krause, RJ; Lash, LH; Elfarra, AA. (2003) Human kidney flavin-containing monooxygenases
8 and their potential roles in cysteine s-conjugate metabolism and nephrotoxicity. J Pharmacol
9 Exp Ther 304(1): 185—191.
10 Kroneld R. (1989) Volatile pollutants in the environment and human tissues. B Environ Contam
11 Tox 42:873-877.
12 Laham S. 1970. Studies on placental transfer: Trichloroethylene. Ind Med 39:46-49.
13 Larson, JL; Bull, RJ. (1989) Effect of ethanol on the metabolism of trichloroethylene. J Toxicol
14 Env Health 28(4):395-406.
15 Larson, JL; Bull, RJ. (1992a) Species differences in the metabolism of trichloroethylene to the
16 carcinogenic metabolites trichloroacetate and dichloroacetate. Toxicol Appl Pharm 115(2):278-
17 285.
18 Larson, J.L. and R.J. Bull. 1992b. Metabolism and lipoperoxidative activity of trichloroacetate
19 and dichloroacetate in rats and mice. Toxicol. Appl. Pharm. 115:268-277.
20 Larsen, G. L., and Stevens, J. L. (1986). Cysteine conjugate beta-lyase in the gastrointestinal
21 bacterium Eubacterium limosum. MolPharmacol 29, 97-103.
22 Lash, L.H. and M.W. Anders. 1986. Cytotoxicity of »S'-(1,2-dichlorovinyl)glutathione and
23 ^-(l^-dichloroviny^-L-cysteine in isolated rad kidney cells. J. Biol. Chem.
24 261(28): 13076-13081.
25 Lash, L.H., A.A. Elfarra and M.W. Anders. 1986. Renal cysteine conjugate P-lyase.
26 Bioactivation of nephrotoxic cysteine ^-conjugates in mitochondrial outer membrane. J. Biol.
27 Chem. 261(13):5930-5935.
6/22/2009
117
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Lash, LH; Jones, DP; Anders, MW. (1988) Glutathione homeostasis and glutathione ^-conjugate
2 toxicity in kidney. Rev Biochem Toxicol 9:29-67.
3 Lash, L.H., R.M. Nelson, R.A. Van Dyke and M.W. Anders. 1990a. Purification and
4 characterization of human kidney cytosolic cysteine conjugate P-lyase activity. Drug Metab.
5 Disp. 18(l):50-54.
6 Lash, L.H., A. A. Elfarra, D. Rakiewicz-Nemeth and M.W. Anders. 1990b. Bioactivation
7 mechanism of cytotoxic homocysteine ^-conjugates. Arch. Biochem. Biophys. 276(2):322-330.
8 Lash, LH; Sausen, PJ; Duescher, RJ; Cooley, AJ; Elfarra, AA. (1994) Roles of cysteine
9 conjugate beta-lyase and S-oxidase in nephrotoxicity: studies with S-(l,2-dichlorovinyl)-L-
10 cysteine and S-(l,2-dichlorovinyl)-L-cysteine sulfoxide. J Pharmacol Exp Ther 269(l):374-383.
11 Lash, LH; Xu, Y; Elfarra, AA; et al. (1995) Glutathione-dependent metabolism of
12 trichloroethylene in isolated liver and kidney cells of rats and its role in mitochondrial and
13 cellular toxicity. Drug Metab Dispos 23(8):846-853.
14 Lash, LH; Qian, W; Putt, DA; et al. (1998) Glutathione conjugation of trichloroethylene in rats
15 and mice: sex-, species-, and tissue-dependent differences. Drug Metab Dispos 26(1): 12-19.
16 Lash, LH; Lipscomb, JC; Putt, DA; et al. (1999a) Glutathione conjugation of trichloroethylene in
17 human liver and kidney: kinetics and individual variation. Drug Metab Dispos 27(3):351-359.
18 Lash, LH; Putt, DA; Brashear, WT; et al. (1999b) Identification of £-(1,2-
19 dichlorovinyl)glutathione in the blood of human volunteers exposed to trichloroethylene. J
20 Toxicol Env Health A 56(1): 1-21.
21 Lash, LH; Fisher, JW; Lipscomb, JC; et al. (2000a) Metabolism of trichloroethylene. Environ
22 Health Persp 108(Suppl. 2): 177-200.
23 Lash, L. H., Parker, J. C., and Scott, C. S. (2000b). Modes of action of trichloroethylene for
24 kidney tumorigenesis. Environ Health Perspect 108 Suppl 2, 225-240.
25 Lash, LH; Putt, DA; Hueni, SE; Krause, RJ; Elfarra, AA. (2003) Roles of necrosis, apoptosis,
26 and mitochondrial dysfunction in S-(l,2-dichlorovinyl)-L-cysteine sulfoxide-induced
27 cytotoxicity in primary cultures of human renal proximal tubular cells. J Pharmacol Exp Ther
28 305(3): 1163—1172.
6/22/2009
118
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Lash, LH; Putt, DA; Parker, JC. (2006) Metabolism and tissue distribution of orally administered
2 trichloroethylene in male and female rats: identification of glutathione- and cytochrome P-450-
3 derived metabolites in liver, kidney, blood, and urine. J Toxicol Env Health A
4 69(13): 1285-1309.
5 Lau, S.S., H.E. Kleiner and T.J. Monks. 1995. Metabolism as a determinant of sepecies
6 susceptibility to 2,3,5-(triglutathion-tS'-YL)hydroquinone-mediated nephrotoxicity. The role of
7 /V-acetylation and/V-deacetylation. Drug. Metab. Disp. 23(10): 1136-1142.
8 Lee, K. M., Bruckner, J. V., Muralidhara, S., and Gallo, J. M. (1996). Characterization of
9 presystemic elimination of trichloroethylene and its nonlinear kinetics in rats. Toxicol Appl
10 Pharmacol 139, 262-271.
11 Lee, KM; Muralidhara, S; White, CA; et al. (2000a) Mechanisms of the dose-dependent kinetics
12 of trichloroethylene: oral bolus dosing of rats. Toxicol Appl Pharm 164(l):55-64.
13 Lee, K. M., Muralidhara, S., Schnellmann, R. G., and Bruckner, J. V. (2000b). Contribution of
14 direct solvent injury to the dose-dependent kinetics of trichloroethylene: portal vein
15 administration to rats. Toxicol Appl Pharmacol 164, 46-54.
16 Li, T; Schultz, I; Keys, DA; Campbell, JL; Fisher, JW. (2008) Quantitative evaluation of
17 dichloroacetic acid kinetics in human—a physiologically based pharmacokinetic modeling
18 investigation. Toxicology 245:35-48.
19 Lilly, P. D., Simmons, J. E., and Pegram, R. A. (1994). Dose-dependent vehicle differences in
20 the acute toxicity of bromodichloromethane. Fundam Appl Toxicol 23, 132-140.
21 Lipscomb, J.C., D.A. Mahle, W.T. Brashear and H.A. Barton. 1995. Dichloroacetic acid:
22 Metabolism in cytosol. Drug Metab. Disp. 23(11):1202-1205.
23 Lipscomb, J.C., D.A. Mahle, W.T. Brashear and C.M. Garrett. 1996. A species comparison of
24 chloral hydrate metabolism in blood and liver. Biochem. Biophys. Res. Comm. 227:340-350.
25 Lipscomb, JC; Garrett, CM; Snawder, JE. (1997) Cytochrome P450-dependent metabolism of
26 trichloroethylene: interindividual differences in humans. Toxicol Appl Pharm 142(2):311 -318.
27 Lipscomb, JC; Fisher, JW; Confer, PD; et al. (1998a) In vitro to in vivo extrapolation for
28 trichloroethylene metabolism in humans. Toxicol Appl Pharm 152(2):376-387.
6/22/2009
119
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Lipscomb, J.C., C.M. Garrett and J.E. Snawder. 1998b. Use of kinetic and mechanistic data in
2 species extrapolation of bioactivation: Cytochrome P-450 dependent trichloroethylene
3 metabolism at occupationally relevant concentrations. J. Occup. Health. 40:110-117.
4 Lipscomb, JC; Teuschler, LK; Swartout, J; et al. (2003) The impact of cytochrome P450 2E1-
5 dependent metabolic variance on a risk-relevant pharmacokinetic outcome in humans. Risk Anal
6 23(6): 1221-1238.
7 Lumpkin, MH; Bruckner, JV; Campbell, JL; Dallas, CE; White, CA; Fisher, JW. (2003) Plasma
8 binding of trichloroacetic acid in mice, rats, and humans under cancer bioassay and
9 environmental exposure conditions. Drug Metab Dispos 31(10): 1203-1207.
10 MacFarlane, M; Foster, JR; Gibson, GG; King, LJ; Lock, EA. (1989) Cysteine conjugate beta-
11 lyase of rat kidney cytosol: characterization, immunocytochemical localization, and correlation
12 with hexachlorobutadiene nephrotoxicity. Toxicol Appl Pharm 98(2): 185-197.
13 Mahle, DA; Gearhart, JM; Grigsby, CC; Mattie, DR; Barton, HA; Lipscomb, JC; Cook, RS.
14 (2007) Age-dependent partition coefficients for a mixture of volatile organic solvents in
15 Sprague-Dawley rats and humans. J Toxicol Env Health A 70:1745-1751.
16 Malherbe, P., D. Alberati-Giani, C. Kohler and A.M. Cesura. 1995. Identification of a
17 mitochondrial form of kynurenine aminotransferase/glutamine transaminase K from rat brain.
18 FEBSLett. 367:141-144.
19 Marshall, EK; Owens, AH. (1955) Absorption, excretion and metabolic fate of chloral hydrate
20 and trichloroethanol. Bull Johns Hopkins Hosp 95:1-18.
21 Mazzullo, M; Bartoli, S; Bonora, B; Colacci, A; Lattanzi, G.; Niero, A; Silingardi, P. Grilli, S.
22 (1992) In vivo and in vitro interaction of trichloroethylene with macromolecules from various
23 organs of rat and mouse. Res Commun Chem Path 76(2): 192-208.
24 McCarver, DG, Byun, R, Hines, RN, Hichme, M, Wegenek, W. (1998). A genetic polymorphism
25 in the regulatory sequences of human CYP2E1: association with increased chlorzoxazone
26 hydroxylation in the presence of obesity and ethanol intake. Toxicol Appl Pharmacol.
27 152(1):276-81.
28
29 McConnell, G; Ferguson, DM; Pearson, CR. (1975) Chlorinated hydrocarbons and the
3 0 environment. Endeavour 34:13-18.
6/22/2009
120
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Merdink, JL; Gonzalez-Leon, A; Bull, RJ; et al. (1998) The extent of dichloroacetate formation
2 from trichloroethylene, chloral hydrate, trichloroacetate, and trichloroethanol in B6C3F1 mice.
3 Toxicol Sci 45(1):33-41.
4 Merdink, JL; Bull, RJ; Schultz, IR. (2000) Trapping and identification of the dichloroacetate
5 radical from the reductive dehalogenation of trichloroacetate by mouse and rat liver microsomes.
6 Free Radic Biol Med 29(2): 125-130.
7 Miller, R.E. and F.P. Guengerich. 1982. Oxidation of trichloroethylene by liver microsomal
8 cytochrome P-450: Evidence for chlorine migration in a transition state not involving
9 trichloroethylene oxide. Biochemistry. 21:1090-1097.
10 Miller, RE; Guengerich, FP. (1983) Metabolism of trichloroethylene in isolated hepatocytes,
11 microsomes, and reconstituted enzyme systems containing cytochrome P-450. Cancer Res
12 43(3): 1145-1152.
13 Moghaddam, A.P., R. Abbas, J.W. Fisher, S. Stavrou and J.C. Lipscomb. 1996. Formation of
14 dichloroacetic acid by rat and mouse gut microflora, an in vitro study. Biochem. Biophys. Res.
15 Comm. 228:639-645.
16 Moghaddam, A.P., R. Abbas, J.W. Fisher and J.C. Lipscomb. 1997. The role of mouse
17 intestinal microflora in the metabolism of trichloroethylene, an in vivo study. Human Exp.
18 Toxicol. 16:629-635.
19 Monster, AC; Boersma, G; Duba, WC (1976) Pharmacokinetics of trichloroethylene in
20 volunteers, influence of workload and exposure concentration. Int Arch Occ Env Hea 38(2):87-
21 102.
22 Monster AC, Boersma G, Duba WC. 1979. Kinetics of trichloroethylene in repeated exposure of
23 volunteers. Int Arch Occ Env Hea 42:283-292.
24 Moslen, MT; Reynolds, ES; Boor, PJ; Bailey, K; Szabo, S. (1977) Trichloroethylene-induced
25 deactivation of cytochrome P-450 and loss of liver glutathione in vivo. Res Commun Chem Path
26 16(1): 109—120.
27 Miiller G, Spassowski M, Henschler D. 1975. Metabolism of trichloroethylene in man. III.
28 Interaction of trichloroethylene and ethanol. Arch Toxicol 33: 173-189.
6/22/2009
121
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Mueller, W. F., Coulston, F., and Korte, F. (1982) Comparative metabolism of 14C TCE in
2 chimpanzees, baboons, and rhesus monkeys. Chemosphere 11:215-218.
3 Nakai, JS; Stathopulos, PB; Campbell, GL; Chu, I; Li-Muller, A; Aucoin, R. (1999) Penetration
4 of chloroform, thrichloroethylene, and tetrachloroethylene through human skin. J Toxicol Env
5 Health A 58:157-170.
6 Nakajima, T., T. Okino, S. Okuyama, T. Kaneko, I. Yonekura and A. Sato. 1988. Ethanol-
7 induced enhancement of trichloroethylene metabolism and hepatotoxicity: Difference from the
8 effect of phenobarbital. Toxicol Appl Pharm 94:227-237.
9 Nakajima, T., R-S. Wang, N. Murayama and A. Sato. 1990. Three forms of trichloroethylene-
10 metabolizing enzymes in rat liver induced by ethanol, phenobarbital, and 3-methylcholanthrene.
11 Toxicol Appl Pharm 102:546-552.
12 Nakajima, T., R-S. Wang, E. Elovaara, S.S. Park, H.V. Gelboin and H. Vainio. (1992a). A
13 comparative study on the contribution of cytochrome P450 isozymes to metabolism of benzene,
14 toluene and trichloroethylene in rat liver. Biochem. Pharmacol. 43(2):251-257.
15 Nakajima, T; Wang, RS; Katakura, Y; et al. (1992b) Sex-, age- and pregnancy-induced changes
16 in the metabolism of toluene and trichloroethylene in rat liver in relation to the regulation of
17 cytochrome P450TTE1 and P450IIC11 content. J Pharmacol Exp Ther 261 (3): 869-874.
18 Nakajima, T; Wang, RS; Elovaara, E; et al. (1993) Cytochrome P450-related differences
19 between rats and mice in the metabolism of benzene, toluene and trichloroethylene in liver
20 microsomes. Biochem Pharmacol 45(5): 1079-1085.
21 Ni, Y-C., T-Y. Wong, R.V. Lloyd et al. 1996. Mouse liver microsomal metabolism of chloral
22 hydrate, trichloroacetic acid, and trichloroethanol leading to induction of lipid peroxidation via a
23 free radical mechanism. Drug Metab. Disp. 24(l):81-90.
24 Nomiyama, K; Nomiyama, H. (1971) Metabolism of trichloroethylene in human. Sex difference
25 in urinary excretion of trichloroacetic acid and trichloroethanol. Int Arch Arbeitsmed
26 28(l):37-48.
27 Nomiyama, H., and Nomiyama, K. 1979. Pathway and rate of metabolism of TCE in rats and
28 rabbits. Ind. Health. 17, 29 37.
6/22/2009
122
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 NRC (National Research Council) Committee on Human Health Risks of Trichloroethylene.
2 (2006) Assessing the Human Health Risks of Trichloroethylene: key scientific issues.
3 Washington, DC: National Academies Press, http://www.nap.edu/catalog/11707.html
4 Ogata, M; Takatsuka, Y; Tomokuni, K. (1971) Excretion of organic chlorine compounds in the
5 urine of persons exposed to vapours of trichloroethylene and tetrachloroethylene. Br J Ind Med
6 28(4):386-391.
7 Opdam, JJ. (1989) Intra and interindividual variability in the kinetics of a poorly and highly
8 metabolizing solvent. Br J Ind Med 46(12):831-845.
9 Park, SB; Osterich, JD; Vamvakas, S; Hashmi, M; Anders, MW; Cashman, JR. (1992) Flavin-
10 containing monooxygenase-dependent stereoselective S-oxygenation and cytotoxicity of cysteine
11 S-conjugates and mercapturates. Chem Res Toxicol 5(2): 193-201.
12 Parkinson, A; Mudra, DR; Johnson, C; Dwyer, A; Carroll, KM. (2004) The effects of gender,
13 age, ethnicity, and liver cirrhosis on cytochrome P450 enzyme activity in human liver
14 microsomes and inducibility in cultured human hepatocytes. Toxicol Appl Pharm 199(3): 193—
15 209.
16 Pastino, GM; Yap, WY; Carroquino, M. (2000) Human variability and susceptibility to
17 trichloroethylene. Environ Health Persp 108(Suppl 2):201-214.
18 Pellizzari ED, Hartwell TD, Harris BS, et al. 1982. Purgeable organic compounds in mother's
19 milk. Bull Environ Contam Toxicol 28:322-328
20 Perbellini, L; Olivato, D; Zedde, A; Miglioranzi, R. (1991) Acute trichloroethylene poisoning by
21 ingestion: clinical and pharmacokinetic aspects. Intens Care Med 17:234-235.
22 Perry, S.J., M.A. Schofield, M. Macfarlane et al. 1993. Isolation and expression of a cDNA
23 coding for rat kidney cytosolic cysteine conjugate P-lyase. Mol Pharmacol. 43:660-665.
24 Peter, R., R. Bocker, P.H. Beaune, M. Iwasaki, F.P. Guengerich and C.S. Yang. 1990.
25 Hydroxylation of chlorzoxazone as a specific probe for human liver cytochrome P-450TTE1
26 Chem. Res. Toxicol. 3:566-573.
27 Pfaffenberger CD, Peoples AJ, Enos HF. 1980. Distribution of volatile halogenated organic
28 compounds between rat blood serum and adipose tissue. Int J Environ Anal Chem 8:55-65.
6/22/2009
123
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Poet, T. S., Corley, R. A., Thrall, K. D., Edwards, J. A., Tanojo, H., Weitz, K. K., Hui, X.,
2 Maibach, H. I., and Wester, R. C. (2000). Assessment of the percutaneous absorption of
3 trichloroethylene in rats and humans using MS/MS real-time breath analysis and physiologically
4 based pharmacokinetic modeling. Toxicol Sci 56, 61-72.
5 Powell, J.F. 1945. Trichl or ethylene: Absorption, elimination and metabolism. Nuffield Dept. of
6 Clinical Medicine, University of Oxford, p. 142-145.
7 Prasanna, H. R., Hart, R. W., and Magee, P. N. (1989). Effect of food restriction on the
8 metabolism of dimethylnitrosamine (NDMA) in rats. Anticancer Res 9, 93-96.
9 Pravecek, T.L., S.R. Channel, W.J. Schmidt and J.K. Kidney. 1996. Cytotoxicity and
10 metabolism of dichloroacetic and trichloroacetic acid in B6C3F1 mouse liver tissue. In Vitro
11 Toxicol. 9(3):261-269.
12 Prout, MS; Provan, WM; Green, T. (1985) Species differences in response to trichloroethylene.
13 I. Pharmacokinetics in rats and mice. Toxicol Appl Pharm 79(3):389-400.
14 Ripp, S.L., L.H. Overby, R.M. Philpot and A.A. Elfarra. 1997. Oxidation of cysteine S-
15 conjugates by rabbit liver microsomes and cDNA-expressed flavin-containing mono-oxygenases:
16 Studies with »S'-(1,2-dichlorovinyl)-L-cysteine, »S'-(1,2,2-trichlorovinyl)-L-cysteine, .S-allyl-L-
17 cysteine, and ^-benzyl-L-cysteine. Mol Pharmacol. 51:507-515.
18 Ripp, SL; Itagaki, K; Philpot, RM; Elfarra, AA. (1999) Species and sex differences in expression
19 of flavin-containing monooxygenase form 3 in liver and kidney microsomes. Drug Metab
20 Dispos 27(l):46-52.
21 Ristoff, E., and Larsson, A. (2007). Inborn errors in the metabolism of glutathione. Orphanet J
22 Rare Dis. 2:16.
23
24 Rodriguez, CE; Mahle, DA; Gearhart, JM; et al. (2007) Predicting age-appropriate
25 pharmacokinetics of six volatile organic compounds in the rat utilizing physiologically based
26 pharmacokinetic modeling. Toxicol Sci 98(1) :43-56.
27 Saghir, SA; Schultz, IR. (2002) Low-dose pharmacokinetics and oral bioavailability of
28 dichloroacetate in naive and GST^-depleted rats. Environ Health Persp 110(8):757-763.
29 Sato, A and Nakajima, T. 1978. Differences following skin or inhalation exposure in the
30 absorption and excretion kinetics of trichloroethylene and toluene. Brit J Ind Med 35:43-49.
6/22/2009
124
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Sato, A; Nakajima, T. (1979) A vial-equilibration method to evaluate the drug-metabolizing
2 enzyme activity for volatile hydrocarbons. Toxicol Appl Pharm 47(l):41-46.
3 Sato A, Nakajima T, Fujiwara Y and MarayamaN. 1977. A pharmacokinetic model to study
4 the excretion of trichloroethylene and its metabolites after an inhalation exposure. Brit J Ind
5 Med 34:56-63.
6 Sausen PJ, Elfarra AA. (1990). Cysteine conjugate S-oxidase. Characterization of a novel
7 enzymatic activity in rat hepatic and renal microsomes. J Biol Chem. 265(11):6139-45.
8
9 Sausen, P.J. and Elfarra, A.A. 1991. Reactivity of cysteine s-conjugate sulfoxides: formation of
10 »S'-[1 -chloro-2-CS'-glutathionyl)vinyl]-L-cysteine sulfoxide by the reaction of
11 ^-(l^-dichloroviny^-L-cycsteine sulfoxide with glutathione. Chem. Res. Toxicol. 4:655-660.
12 Savolainen, H., Pfaffli, P., Tengen, M., and Vainio, H. (1977). Trichloroethylene and 1,1,1-
13 trichloroethane: effects on brain and liver after five days intermittent inhalation. Arch Toxicol 38,
14 229-237.
15 Schultz, IR; Merdink, JL; Gonzalez-Leon, A; Bull, RJ. (1999) Comparative toxicokinetics of
16 chlorinated and brominated haloacetates in F3444 rats. Toxicol Appl Pharm 158(2): 103-114.
17 Schultz, IR; Merdink, JL; Gonzalez-Leon, A; Bull, RJ. (2002) Dichloroacetate toxicokinetics and
18 disruption of tyrosine catabolism in B6C3F1 mice: dose-response relationships and age as a
19 modifying factor. Toxicology 173:229-247.
20 Sellers, EM; Lang, M; Koch-Weser, J; LeBlanc, E; Kalant, H. (1972) Interaction of chloral
21 hydrate and ethanol in man. I: metabolism. Clin Pharmacol Ther 13:37-49.
22 Sellers, E.M., B.S. Lang-Sellers, and J. Koch-Weser. 1978. Comparative metabolism of chloral
23 hydrate and triclofos. From the Division of Clinical Pharmacology, Clinical Institute, Addiction
24 Research Foundation and Toronto Western Hospital and Departments of Pharmacology and
25 Medicine, University of Toronto, Toronto, and the Centre de Recherche Merrell International,
26 Strasbourg, France.
27 Sheffels, P; Schroeder, JL; Altuntas, TG; Liggitt, HD; Kharasch, ED. (2004) Role of cytochrome
28 P4503 A in cysteine S-conjugates sulfoxidation and the nephrotoxicity of the sevoflurane
29 degradation product fluoromethyl-2,2-difluoro-l-(trifluoromethyl)vinyl ether (compound A) in
30 rats. Chem Res Toxicol 17(9): 1177-1189.
6/22/2009
125
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Shimada, T., H. Yamazaki, M. Mimura, Y, Inui, and F.P. Guengerich. 1994. Interindividual
2 variations in human liver cytochrome p-450 enzymes involved in the oxidation of drugs,
3 carcinogens and toxic chemicals: studies with liver microsomes of 30 japanese and 30
4 Caucasians. J. Pharmacol. Exp. Ther. 270:414-423.
5 Shroads, AL; Guo, X; Dixit, V; Liu, H-P; James, MO; Stacpoole, PW. (2008) Age-dependent
6 kinetics and metabolism of dichloroacetate: possible relevance to toxicity. J Pharmacol Exp
7 Ther 324(3): 1163-1171.
8 Shultz, J. and H. Weiner. 1979. Alteration of the enzymology of chloral hydrate reduction in the
9 presence of ethanol. Biochem. Pharmacol. 28:3379-3384.
10 Simmons, JE; Boyes, WK; Bushnell, PJ; Raymer, JH; Limsakun, T; McDonald, A; Sey, YM;
11 Evans, MV. (2002) A physiologically based pharmacokinetic model for trichloroethylene in the
12 male long-evans rat. Toxicol Sci 69(1):3—15.
13 Skender, LJ; Karacic, V; Prpic-Majic, D. (1991) A comparative study of human levels of
14 trichloroethylene and tetrachloroethylene after occupational exposure. Arch Environ Health
15 46(3): 174-178.
16 Soucek, B; Vlachova, D. (1959) Excretion of trichloroethylene metabolites in human urine. Brit
17 J Ind Med 17:60-66.
18 Soucek, B; Vlachova, D. (1960) Excretion of trichloroethylene metabolites in human urine. Brit
19 J Ind Med 17:60-64.
20 Stenner, RD; Merdink, JL; Stevens, DK; et al. (1997) Enterohepatic recirculation of
21 trichloroethanol glucuronide as a significant source of trichloroacetic acid. Metabolites of
22 trichloroethylene. Drug Metab Dispos 25(5): 529-535.
23 Stevens, J. 1985a. Cysteine conjugate P-lyase activities in rat kidney cortex: subcellular
24 localization and relationship to the hepatic enzyme. Biochem. Biophys. Res. Commun.
25 129(2):499-504.
26 Stevens, J. 1985b. Isolation of characterization of a rat liver enzyme with both cysteine
27 conjugate P-lyase and kynureninase activity. J. Biol. Chem. 260(13):7945-7950.
28 Stevens, J., and W.B. Jakoby. 1983. Cysteine conjugate P-lyase. Mol Pharmacol. 23:761-765.
6/22/2009
126
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Stevens, J.L., J.D. Robbins, and R. A. Byrd. 1986. A purified cycsteine conjugate P-lyase from
2 rat kidney cytosol. J. Biol. Chem. 261 (33): 15529-15537.
3 Stevens, J.L., N. Ayoubi, and J.D. Robbins. 1988. The role of mitochondrial matrix enzymes in
4 the metabolism and toxicity of cysteine conjugates. J. Biol. Chem. 263(7):3395-3401.
5 Stevens, J.L., P.B. Hatzinger, and P.J. Hayden. 1989. Quantitation of multiple pathways for the
6 metabolism of Mephrotoxic cycsteine conjugates using selective inhibitors of l-oc-hydroxy acid
7 oxidase (L-amino acid oxidase) and cysteine conjugate P-lyase. Drug Metab. Dispos.
8 17(3):297-303.
9 Stewart, RD; Dodd, HC; Gay, HH; et al. (1970) Experimental human exposure to
10 trichloroethylene. Arch Environ Health 20(1):64-71.
11 Tateishi, M., S. Suzuki, and H. Shimizu. 1978. Cysteine conjugate P-lyase in rat liver. J.Biol.
12 Chem. 253(24):8854-8859.
13 Templin, MV; Parker, JC; Bull, RJ. (1993) Relative formation of dichloroacetate and
14 trichloroacetate from trichloroethylene in male B6C3F1 mice. Toxicol Appl Pharm 123(1): 1-8.
15 Templin, MV; Stevens, DK; Stenner, RD; et al. (1995a). Factors affecting species differences in
16 the kinetics of metabolites of trichloroethylene. J Toxicol Env Health 44(4): 43 5-447.
17 Templin MV, Parker JC, Bull RJ.(1995b). Relative formation of dichloroacetate and
18 trichloroacetate from trichloroethylene in male B6C3F1 mice. Toxicol Appl Pharmacol. 123:1-8.
19 Tomisawa, H., S. Suzuki, S. Ichihara, H. Fukazawa, and M. Tateishi. 1984. Purification and
20 characterization of C-S lyase from fusobacterium varium. J. Biol. Chem. 259(4):2588-2593.
21 Tomisawa, H., S. Ichihara, H. Fukazawa, N. Ichimoto, M. Tateishi, and I. Yamamoto. 1986.
22 Purification and characterization of human hepatic cysteine-conjugate P-lyase. Biochem. J.
23 235:569-575.
24 Tong, Z; Board, PG; Anders, MW. (1998) Glutathione transferase zeta catalyses the oxygenation
25 of the carcinogen dichloroacetic acid to glyoxylic acid. Biochem J 331 ( Pt 2):371-374.
26 U.S. EPA. (1985) Assessment of trichloroethylene as a potentially toxic air pollutant; proposed
27 rule. Federal Register; Part IV 50(246):52422-52425.
6/22/2009
127
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 U.S. EPA. (2001) Trichloroethylene health risk assessment: synthesis and characterization
2 (external review draft). Office of Research and Development, National Center for
3 Environmental Assessment, Washington, DC, EPA/600/P-01/002A.
4 USAF (U.S. Air Force). (2004) Determination of partition coefficients for a mixture of volatile
5 organic compounds in rats and humans at different life stages. Air Force Research Laboratory,
6 Wright-Patterson Air Force Base, OH. AFRL-HE-WP-TR-2005-0012.
7 van Bree, L., Commandeur, J., Lamberts, B., Cornelissen, M., van Roon, M., Laterveer, H., and
8 de Vries, J. (1990). Induction of drug metabolism enzymes by dihalogenated biphenyls. J
9 Biochem Toxicol 5, 57-63.
10 Verma, Y; Rana, SV. (2003) Gender differences in the metabolism of benzene, toluene and
11 trichloroethylene in rat with special reference to certain biochemical parameters. J Environ Biol
12 24(2): 135 -140.
13 Vesterberg, O; Astrand, I. (1976) Exposure to trichloroethylene monitored by analysis of
14 metabolites in blood and urine. J Occup Med 18(4):224-226.
15 Vesterberg, O; Gorczak, J; Krasts, M. (1976) Exposure to trichloroethylene II. Metabolites in
16 blood and urine. Scand J Work Env Hea 4:212-219.
17 Wallace, LA; Pellizzari, E; Hartwell, T; Rosenzweig, M; Erickson, M; Sparacino, C; Zelon, H.
18 (1984) Personal exposure to volatile organic compounds. I. Direct measurements in breathing-
19 zone air, drinking water, food, and exhaled breath. Environ Res 35:293-319.
20 Wang, G.-Z., T. Takano, K. Tomita, K. Nakata, and K. Nakamura. 1993. Interaction of
21 trichloroethylene and isopropyl alcohol in the perfused rat liver. Jpn. J. Hyg. 48:1000-1005.
22 Werner, M; Birner, G; Dekant, W. (1995a) The role of cytochrome P4503A1/2 in the sex-
23 specific sulfoxidation of the hexachlorobutadiene metabolite, N-acetyl-S-
24 (pentachlorobutadienyl)-L-cysteine in rats. Drug Metab Dispos 23(8):861-868.
25 Werner, M; Guo, Z; Birner, G; Dekant, W; Guengerich, FP. (1995b) The sulfoxidation of the
26 hexachlorobutadiene metabolite N-acetyl-S-(l,2,3,4,4-pentachlorobutadienyl)-L-cysteine is
27 catalyzed by human cytochrome P450 3A enzymes. Chem Res Toxicol 8(7):917-923.
28 Werner, M; Birner, G; Dekant, W. (1996) Sulfoxidation of mercapturic acids derived from tri-
29 and tetrachloroethene by cytochromes P450 3 A: a bioactivation reaction in addition to
6/22/2009
128
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 deacetylation and cysteine conjugate beta-lyase mediated cleavage. Chem Res Toxicol
2 9(l):41-49.
3 Withey, JR; Karpinski, K. (1985) The fetal distribution of some aliphatic chlorinated
4 hydrocarbons in the rat after vapor phase exposure. Biol Res Preg Perin 6(2):79-88.
5 Withey, JR; Collins, BT; Collins, PG. (1983) Effect of vehicle on the pharmacokinetics and
6 uptake of four halogenated hydrocarbons from the gastrointestinal tract of the rat. J Appl
7 Toxicol 3(5):249-253.
8 Wolfgang, GHI; Gandolfi, AJ; Stevens, JL; Brendel, K. (1989) A-acetyl »Y-(1,2-dichlorovinyl)-L-
9 cysteine produces a similar toxicity to »Y-(1,2-dichlorovinyl)-L-cysteine in rabbit renal slices:
10 differential transport and metabolism. Toxicol Appl Pharm 101:205-219.
11 Yoshida, M., Fukabori, S., Hara, K., Yuasa, H., Nakaaki, K., Yamamura, Y., and Yoshida, K.
12 (1996). Concentrations of trichloroethylene and its metabolites in blood and urine after acute
13 poisoning by ingestion. Hum Exp Toxicol 15, 254-258.
14 Yoshioka T, Krauser JA, Guengerich FP. 2002. Microsomal oxidation of tribromoethylene and
15 reactions of tribromoethylene oxide. Chem Res Toxicol. 15(11): 1414-20.
16 Yu, KO; Barton, HA; Mahle, DA; Frazier, JM. (2000) In vivo kinetics of tricholoracetate in male
17 Fischer 344 rats. Toxicol Sci 54:302-311.
18 Yu, S., Rao, S., and Reddy, J. K. (2003). Peroxisome proliferator-activated receptors, fatty acid
19 oxidation, steatohepatitis and hepatocarcinogenesis. CurrMolMed 3, 561-572.
20 Zhang, G; Stevens, JL. (1989) Transport and activation of »Y-(1,2-dichlorovinyl)-L-cysteine and
21 A-acetyUY-(l,2-dichlrovinyl)-L-cysteine in rat kidney proximal tubules. Toxicol Appl Pharm
22 100:51-61.
23
6/22/2009
129
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.5 PBPK Modeling of TCE and Its Metabolites
3.5.1 Introduction
Physiologically based pharmacokinetic (PBPK) models are extremely useful tools for
quantifying the relationship between external measures of exposure and internal measures of
toxicologically relevant dose. In particular, for the purposes of this assessment, PBPK models
are evaluated for the following: (i) providing additional quantitative insights into the absorption,
distribution, metabolism, and excretion (ADME) of TCE and metabolites described in the
sections above; (ii) cross-species pharmacokinetic extrapolation of rodent studies of both cancer
and noncancer effects, (iii) exposure-route extrapolation; and (iv) characterization of human
pharmacokinetic variability. The following sections first describe and evaluate previous and
current TCE PBPK modeling efforts, then discuss the insights into ADME (i, above), and finally
present conclusions as to the utility of the model to predict internal doses for use in
dose-response assessment (ii—iv, above).
3.5.2 Previous PBPK Modeling of TCE for Risk Assessment Application
TCE has an extensive number of both in vivo pharmacokinetic and PBPK modeling
studies (see Chiu et al. 2006, supplementary material, for a review). Models previously
developed for occupational or industrial hygiene applications are not discussed here but are
reviewed briefly in Clewell et al. (2000). Models designed for risk assessment applications have
focused on descriptions of TCE and its major oxidative metabolites TCA, TCOH, and TCOG.
Most of these models were extensions of the "first generation" of models developed by Fisher
and coworkers (Allen and Fisher 1993; Fisher et al. 1991) in rats, mice, and humans. These
models, in turn, are based on a Ramsey and Andersen (1984) structure with flow-limited tissue
compartments and equilibrium gas exchange, saturable Michaelis-Menten kinetics for oxidative
metabolism, and lumped volumes for the major circulating oxidative metabolites TCA and
TCOH. Fisher and coworkers updated their models with new in vivo and in vitro experiments
performed in mice (Abbas and Fisher 1997; Greenberg et al. 1999) and human volunteers
(Fisher et al. 1998) and summarized their findings in Fisher (2000). Clewell et al. (2000) added
enterohepatic recirculation of TCOG and pathways for local oxidative metabolism in the lung
and GST metabolism in the liver. While Clewell et al. (2000) does not include the updated
Fisher data, they have used a wider set of in vivo and in vitro mouse, rat, and human data than
previous models. Finally, Bois (2000a, 2000b) performed re-estimations of PBPK model
6/22/2009
130
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
parameters for the Fisher and Clewell models using a Bayesian population approach (Gelman et
al. 1996, and discussed further below).
As discussed in Rhomberg (2000), the choice as to whether to use the Fisher, Clewell,
and Bois models for cross-species extrapolation of rodent cancer bioassays led to quantitative
results that differed by as much as an order of magnitude. There are a number of differences in
modeling approaches that can explain their differing results. First, the Clewell et al. (2000)
model differed structurally in its use of single-compartment volume-of-distribution models for
metabolites as opposed to the Fisher (2000) models' use of multiple physiologic compartments.
Also, the Clewell et al. (2000) model, but not the Fisher models, includes enterohepatic
recirculation of TCOH/TCOG (although reabsorption was set to zero in some cases). In addition
to structural differences in the models, the input parameter values for these various models were
calibrated using different subsets of the overall in vivo database (see Chiu et al. 2006,
supplementary material, for a review). The Clewell et al. (2000) model is based primarily on a
variety of data published before 1995; the Fisher (2000) models were based primarily on new
studies conducted by Fisher and coworkers (after 1997); and the Bois (2000a, 2000b) re-
estimations of the parameters for the Clewell et al. (2000) and Fisher (2000) models used slightly
different datasets than the original authors. The Bois (2000a, 2000b) re-analyses also led to
somewhat different parameter estimates than the original authors, both because of the different
data sets used as well as because the methodology used by Bois allowed many more parameters
to be estimated simultaneously than were estimated in the original analyses.
Given all these methodological differences, it is not altogether surprising that the
different models led to different quantitative results. Even among the Fisher models themselves,
Fisher (2000) noted some inconsistencies, including differing estimates for metabolic parameters
between mouse gavage and inhalation experiments. These authors included possible
explanations for these inconsistencies: the impact of corn oil vehicle use during gavage
(Staats et al. 1991) and the impact of a decrease in ventilation rate in mice due to sensory
irritation during the inhalation of solvents (e.g., Stadler and Kennedy 1996).
As discussed in NRC (2006), several additional PBPK models relevant to TCE
pharmacokinetics have been published since 2000 and are reviewed briefly here. Poet et al.
(2000) incorporated dermal exposure to TCE in PBPK models in rats and humans, and published
in vivo data in both species from dermal exposure (Thrall et al. 2000; Poet et al. 2000). Albanese
et al. (2002) published a series of models with more complex descriptions of TCE distribution in
adipose tissue but did not show comparisons with experimental data. Simmons et al. (2002)
developed a PBPK model for TCE in the Long-Evans rat that focused on neurotoxicity endpoints
and compared model predictions with experimentally determined TCE concentrations in several
tissues—including the brain. Keys et al. (2003) investigated the lumping and unlumping of
6/22/2009
131
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
various tissue compartments in a series of PBPK models in the rat and compared model
predictions with TCE tissue concentrations in a multitude of tissues. Although none of these
TCE models included metabolite descriptions, the experimental data was available for either
model or evaluation. Finally, Keys et al. (2004) developed a model for DC A in the mouse that
included a description of suicide inhibition of GST-zeta, but this model was not been linked to
TCE.
3.5.3 Development and Evaluation of an Interim "Harmonized" TCE PBPK Model
Throughout 2004, U.S. EPA and the U.S. Air Force jointly sponsored an integration of
the Fisher, Clewell, and Bois modeling efforts (Hack et al. 2006). In brief, a single interim
PBPK model structure combining features from both the Fisher and Clewell models was
developed and used for all 3 species of interest (mice, rats, and humans). An effort was made to
combine structures in as simple a manner as possible; the evaluation of most alternative
structures was left for future work. The one level of increased complexity introduced was
inclusion of species- and dose-dependent TCA plasma binding, although only a single in vitro
study of Lumpkin et al. (2003) was used as parameter inputs. As part of this joint effort, a
hierarchical Bayesian population analysis using Markov chain Monte Carlo (MCMC) sampling
(similar to the Bois 2000a, 2000b analyses) was performed on the revised model with a
cross-section of the combined database of kinetic data to provide estimates of parameter
uncertainty and variability (Hack et al. 2006). Particular attention was given to using data from
each of the different efforts, but owing to time and resource constraints, a combined analysis of
all data was not performed. The results from this effort suggested that a single model structure
could provide reasonable fits to a variety of data evaluated for TCE and its major oxidative
metabolites TCA, TCOH, and TCOG. However, in many cases, different parameter values—
particularly for metabolism—were required for different studies, indicating significant
interindividual or interexperimental variability. In addition, these authors concluded that
dosimetry of DCA, conjugative metabolites, and metabolism in the lung remained highly
uncertain (Hack et al. 2006).
Subsequently, EPA conducted a detailed evaluation of the Hack et al. (2006) model that
included (i) additional model runs to improve convergence; (ii) evaluation of posterior
distributions for population parameters; and (iii) comparison of model predictions both with the
data used in the Hack et al. (2006) analysis as well as with additional datasets identified in the
literature. Appendix A provides the details and conclusions of this evaluation, briefly
summarized in Table 3.5.1, along with their pharmacokinetic implications.
6/22/2009
132
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.5.1. Conclusions from evaluation of Hack et al. (2006), and implications for PBPK model development.
Conclusion from evaluation of Hack et al. (2006) model
Implications for PBPK model parameters, structure, or data
For some model parameters, posterior distributions were somewhat inconsistent with
the prior distributions.
• For parameters with strongly informative priors (e.g., tissue volumes and
flows), this may indicate errors in the model.
• For many parameters, the prior distributions were based on visual fits to the
same data. If the posteriors are inconsistent, then that means they priors were
"inappropriately" informative, and, thus, the same data was used twice.
Re-evaluation of all prior distributions
• Update priors for parameters with independent data (physiological
parameters, partition coefficients, in vitro metabolism), looking across all
available data sets.
• For priors without independent data (e.g., many metabolism parameters), use
less informative priors (e.g., log-uniform distributions with wide bounds) so
as prevent bias.
Evaluate modifications to the model structure, as discussed below.
A number of datasets involve TCE (ia, portal vein), TCA (oral, iv), and TCOH (oral, iv)
dosing routes that are not currently in the model, but could be useful for calibration.
• Additional dosing routes can be added easily.
TCE concentrations in blood, air, and tissues well-predicted only in rats, not in mice
and humans. Specifically:
• In mice, the oral uptake model could not account for the time-course of
several datasets. Blood TCE concentrations after inhalation consistently over-
predicted.
• In rats, tissue concentrations measured in data not used for calibration were
accurately predicted.
• In humans, blood and air TCE concentrations were consistently over-
predicted in the majority of (but not all) datasets.
• In mice, uptake from the stomach compartment (currently zero), but
previously included in Abbas and Fisher 1997, may improve the model fit.
• In mice and humans, additional extrahepatic metabolism, either presy stemic
(e.g., in the lung) or postsystemic (e.g., in the kidney) and/or a wash-
in/wash-out effect may improve the model fit.
Total metabolism appears well-predicted in rats and mice based on closed chamber
data, but required significantly different Vmax values between dose groups. Total
recovery in humans (60-70%) is less than the model would predict. In all three species,
the ultimate disposition of metabolism is uncertain. In particular, there are uncertainties
in attributing the "missing" metabolism to
• GSH pathway (e.g., urinary mercapturates may only capture a fraction of the
total flux; moreover, in Bernauer et al. 1996, excretion was still on-going at
end of collection period; model does not accurately depict time-course of
mercapturate excretion).
• Other hepatic oxidation (currently attributed to DCA).
• Extra-hepatic systemic metabolism (e.g., kidney).
• Pre-systemic metabolism in the lung.
• Calibration of GSH pathway may be improved by utilizing in vitro data on
liver and kidney GSH metabolism, adding a DCVG compartment to improve
the prediction of the time-course for mercapturate excretion, and/or using
the Lash et al. (1999b) blood DCVG in humans (necessitating the addition
of a DCVG compartment).
• Pre-systemic lung metabolism can only be evaluated if added to the model
(in vitro data exists to estimate the VMax for such metabolism). In addition,
a wash-in/wash-out effect (e.g., suggested by Greenberg et al. 1999) can be
evaluated using a continuous breathing model that separately tracks inhaled
and exhaled air, with adsorption/desporption in the respiratory tract.
• Additional elimination pathways for TCOH and TCA can be added for
evaluation.
6/22/2009
133
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Conclusion from evaluation of Hack et al. (2006) model
Implications for PBPK model parameters, structure, or data
• Additional metabolism of TCOH or TCA (see below).
TCA blood/plasma concentrations well predicted following TCE exposures in all
species. However, there may be inaccuracies in the total flux of TCA production, as
well as its disposition.
• In TCA dosing studies, the majority (>50%), but substantially <100%, was
recovered in urine, suggesting significant metabolism of TCA. Although
urinary TCA was well predicted in mice and humans (but not in rats), if TCA
metabolism is significant, then this means that the current model
underestimates the flux of TCE metabolism to TCA.
• An improved TCOH/TCOG model may also provide better estimates of TCA
kinetics (see below).
TCOH/TCOG concentrations and excretion were inconsistently predicted, particularly
after TCOH dosing.
• In mice and rats, first-order clearance for TCOH glucuronidation was
predicted to be greater than hepatic blood flow, which is consistent with a first
pass effect that is not currently accounted for.
• In humans, the estimated clearance rate for TCOH glucuronidation was
substantially smaller than hepatic blood flow. However, the presence of
substantial TCOG in blood (as opposed to free TCOH) in the Chiu et al.
(2007) data is consistent with greater glucuronidation than predicted by the
model.
• In TCOH dosing studies, substantially <100%) was recovered in urine as
TCOG and TCA, suggesting another metabolism or elimination pathway.
• Additional elimination pathways for TCOH and TCA can be added for
evaluation.
• The addition of a liver compartment for TCOH and TCOG would permit
hepatic first-pass effects to be accounted for, as appears necessary for mice
and rats.
1
6/22/2009
134
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.5.4 PBPK Model for TCE and Metabolites Used for this Assessment
3.5.4.1 Introduction
Based on the recommendations of the NRC (2006) as well as additional analysis and
evaluation of the Hack et al. (2006) PBPK model, an updated PBPK model for TCE and
metabolites was developed for use in this risk assessment. This updated model included
modification of some of aspects of the Hack et al. (2006) PBPK model structure, incorporation
of additional in vitro and in vivo data for estimating model parameters, and an updated
hierarchical Bayesian population analysis of PBPK model uncertainty and variability. The sub-
sections below, the updated PBPK model, and baseline parameter values are described, and the
approach and results of the analysis of PBPK model uncertainty and variability. Appendix A
provies more detailed descriptions of the model and parameters, including background on
hierarchical Bayesian analyses, model equations, statistical distributions for parameter
uncertainty and variability, data sources for these parameter values, and the PBPK model code.
Additional computer codes containing input files to the MCSim program and scripts for data
analysis are available electronically.
3.5.4.2 Updated PBPK Model Structure
The updated TCE PBPK model is illustrated in Figure 3.5.1, with the major changes from
the Hack et al. (2006) model described here. The TCE submodel was augmented by the addition
of kidney and venous blood compartments, and an updated respiratory tract model that included
both metabolism and the possibility of local storage in the respiratory tissue. In particular, in the
updated lung, separate processes describing inhalation and exhalation allowed for adsorption and
desorption from tracheobronchial epithelium (wash-in/wash-out), with the possibility of local
metabolism as well. In addition, conjugative metabolism in the kidney was added, motivated by
the in vitro data on TCE conjugation described in section 3.3.3.2-3.3.3.3. With respect to
oxidation, a portion of the lung metabolism was assumed to produce systemically available
oxidative metabolites, including TCOH and TCA, with the remaining fraction assumed to be
locally cleared. This is clearly a lumping of a multistep process, but the lack of data precludes
the development of a more sequential model. TCE oxidation in the kidney was not included
because it was not likely to constitute a substantial flux of total TCE oxidation given the much
lower P450 activity in the kidney relative to the liver (Cummings et al., 1999; Cummings et al.,
2000) and the greater tissue mass of the liver. In addition, liver compartments were added to the
TCOH and TCOG submodels to account properly for first-pass hepatic metabolism, which is
important for consistency across routes of exposure. Furthermore, metabolism of TCOH and
TCA was added to their respective submodels as additional clearance pathways. With respect to
6/22/2009
135
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 TCE conjugation, in humans, an additional DCVG compartment was added between TCE
2 conjugation and production of DCVC.
3
4
6/22/2009
136
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(^Jnhaled air^)
~
TCE
Exhaled air
Respiratory
Tract Lumen
dnhalationl
I
Respiratory
Tract Tissue
_Oxidation _ (
(UeacJ space)
Respiratory
Tract Lumen
CExhalationl
I.
.P.a.® E_xchanc[_e_
Venous
Blood
Rapidly
Perfused
Slowly
Perfused
Fat
Gut
I
Liver
(jOraT)
~
Stomach
Duodenum
Kidney
>^-0"xi
+T. i^or
Oxidation & f
Conjugation ;
«. £°njugation_ I
Oxidative Metabolism
I Lung
t Oxidation i
I Local |
v _C[earance_ ^
Total
Systemic
.N . TCOH J
Liver
Oxidation
¦/
J TCA 1
_ Oxjdation_
l Other J
V _ ^XlUdllUn _ J
Conjugative Metabolism
(rat and human only)
, Liver , ^
DCVG
^ Conjugation t ^
(human only)
' ^ Y;
1 ®'°" 1
activation •
' Kidney ^
\ Conjugation f r
DCVC
Urine ,
v lNAcpCVC]_ f
TCOH
Blood
TCOG
TCOG
Other
TCOG
Blood
TCOH
TCOH
TCA
Plasma
Body
Liver
I TCE
I TCOH
Urine J
l Other J
Legend
o Input (exposure/dose)
"Dynamic" Compartment (solved by ODEs)
1~ [ "Static" Compartment (at local steady-state)
^ Transformation or Excretion
Figure 3.5.1. Overall structure of PBPK model for TCE and metabolites used in this assessment.
Boxes with underlined labels are additions or modifications of the Hack et al. (2006) model,
which are discussed in Table 3.5.2.
6/22/2009
137
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.5.2. Discussion of changes to the Hack et al. (2006) PBPK model implemented for this
2 assessment
Change to Hack et al.
(2006) PBPK Model
Discussion
TCE respiratory tract
compartments and
metabolism
In vitro data indicate that the lung (at least in the mouse) has a significant capacity for
oxidizing TCE. However, in the Hack et al. (2006) model, respiratory metabolism was
blood flow-limited. The model structure used was inconsistent with other PBPK
models in which the same mechanism for respiratory metabolism is assumed (e.g.,
styrene, Sarangapani et al. 2003). In these models, the main source of exposure in the
respiratory tract tissue is from the respiratory lumen—not from the tracheobronchial
blood flow. In addition, a wash-in/wash-out effect has also been postulated. The
current structure, which invokes a "continuous breathing" model with separate
"inhaled" and "exhaled" respiratory lumens, can accommodate both respiratory
metabolism due to exposure from the respiratory lumen as well as a wash-in/wash-out
effect in which there is temporary storage in the respiratory tract tissue.
Moreover, preliminary analyses indicated that these changes to the model structure
allowed for a substantially better fit to mouse closed chamber data under the
requirement that all the dose levels are modeled using the same set of parameters.
TCE kidney
compartment
In vitro data indicate that the kidney has a significant capacity for conjugating TCE
with GSH.
TCE venous blood
compartment
Many PBPK models have used a separate blood compartment. It was believed to be
potentially important and feasible to implement here because (i) TCE blood
concentrations were often not well predicted by the Hack et al. (2006) model; (ii) the
TCA sub-model has a plasma compartment, which is a fraction of the blood volume
based on the blood volume; (iii) adequate independent information on blood volume is
available; and (iv) the updated model was to include the intravenous route of exposure.
TCOH and TCOG liver
compartments
In mice and rats, the Hack et al. (2006) model estimated a rate of TCOH
glucuronidation that exceeded hepatic blood flow (all glucuronidation is assumed to
occur in the liver), indicated a significant first-pass effect. Therefore, a separate liver
compartment is necessary to account properly for hepatic first-pass.
TCOH and TCA "other"
elimination pathways
Mass-balance studies with TCOH and TCA dosing indicated that, although the
majority of TCOH and TCA are excreted in urine, the amount is still substantially less
than 100%. Therefore, additional elimination of TCOH and TCA must exist and
should be accounted for.
DCVG compartment
(human model only)
Blood DCVG data in humans exist as part of the Fisher et al. (1998) experiments,
reported in Lash et al. (1999b), and a DCVG compartment is necessary in order to
utilize those data.
3
4
6/22/2009
138
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.5.4.3 Specification of PBPK model parameter prior distributions
Point estimates for PBPK model parameters ("baseline values"), used as central estimates
in the prior distributions for population mean parameters in the hierarchical Bayesian statistical
model (see Appendix A), were developed using standard methodologies and were a refinement
of those used in Hack et al. (2006). Because the Bayesian parameter estimation methodology
utilizes the majority of the useable in vivo data on TCE pharmacokinetics, all baseline parameter
estimates were based solely on measurements independent of the in vivo data. This avoids using
the same data in both the prior and the likelihood. These parameters were, in turn, given
truncated normal or lognormal distributions for the uncertainty in the population mean. If no
independent data were available, as is the case for many "downstream" metabolism parameters,
then no baseline value was specified, and a noninformative prior was used. Section 3.5.5.4,
below, discusses the updating of these noninformative priors using interspecies scaling.
In keeping with standard practice, many of the PBPK model parameters were "scaled" by
body or organ weights, cardiac output, or allometrically by an assumed (fixed) power of body
weight. Metabolic capacity and cardiac output were scaled by the 3/4 power of body weight and
rate coefficients were scaled by the - Vi power of body weight, in keeping with general
expectations as to the relationship between metabolise rates and body size (USEPA, 1992; West
et al., 2002) So as to ensure a consistent model structure across species as well as improve the
performance of the Markov chain Monte Carlo (MCMC) algorithm, parameters were further
scaled to the baseline point-estimates where available, as was done by Hack et al. (2006). For
example, to obtain the actual liver volume in liters, a point estimate is first obtained by
multiplying the fixed, species-specific baseline point estimate for the fractional liver volume by a
fixed body weight (measured or species-specific default) with density of 1 kg per liter assumed
to convert from kg to liters. Then, any deviation from this point estimate is represented by
multiplying by a separate "scaled" parameter VLivC that has a value of 1 if there is no deviation
from the point estimate. These "scaled" parameters are those estimated by the MCMC
algorithm, and for which population means and variances are estimated.
Baseline physiological parameters were re-estimated based on the updated tissue lumping
(e.g., separate blood and kidney compartments) using the standard references ICRP (2002) and
Brown et al. (1997). For a few of these parameters, such as hematocrit and respiratory tract
volumes in rodents, additional published sources were used as available, but no attempt was
made to compile a comprehensive review of available measurements. In addition, a few
parameters, such as the slowly perfused volume, were calculated rather than sampled in order to
preserve total mass or flow balances.
For chemical-specific distribution and metabolism parameters, in vitro data from various
sources were used. Where multiple measurements had been made, as was the case for many
6/22/2009
139
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
partition coefficients, TCA plasma protein binding parameters, and TCE metabolism, different
results were pooled together, with their uncertainty reflected appropriately in the prior
distribution. Such in vitro measurements were available for most chemical partition coefficients,
except for those for TCOG (TCOH used as a proxy) and DCVG. There were also such data to
develop baseline values for the oxidative metabolism of TCE in the liver (Vmax and KM), the
relative split in TCE oxidation between formation of TCA and TCOH, and the Vmax for TCE
oxidation in the lung. All other metabolism parameters were not given baseline values and
needed to be estimated from the in vivo data.
3.5.4.4 Dose Metric Predictions
The purpose of this PBPK model is to make predictions of internal dose in rodents used
in toxicity studies or in humans in the general population, and not in the groups or individuals for
which pharmacokinetic data exist. Therefore, to evaluate its predictive utility for risk
assessment, a number of dose metrics were selected for simulation in a "generic" mouse, rat, or
human, summarized in Table 3.5.3. The parent dose metric was area-under-the-curve (AUC) in
blood. TCE metabolism dose metrics (i.e., related to the amount metabolized) included both
total metabolism, metabolism splits between oxidation versus conjugation, oxidation in the liver
versus the lung, the amount of oxidation in the liver to products other than TCOH and TCA, and
the amount of TCA produced. These metabolism rate dose metrics are scaled by body weight in
the case of TCA produced, by the metabolizing tissue volume in the case of the lung and "other"
oxidation in the liver, and by body weight to the 3/4 power in other cases. With respect to the
oxidative metabolites, liver concentrations of TCA and blood concentrations of free TCOH were
used. With respect to conjugative metabolites, the primary dose metric (in addition to total GSH
metabolism) was the amount of DCVC bioactivated (rather than excreted in urine) per unit
kidney mass.
All dose metrics are converted to daily or weekly averages based on simulations lasting
10 weeks for rats and mice and 100 weeks for humans. These simulation times were the shortest
for which additional simulation length did not add substantially to the average (i.e., less than a
few percent change with a doubling of simulation time).
6/22/2009
140
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.5.3. PBPK Model-Based Dose Metrics
Abbreviation
Description
ABioactDCVCKid
Amount of DCVC bioactivated in the kidney (mg) per unit kidney mass (kg)
AMetGSHBW34
Amount of TCE conjugated with GSH (mg) per unit body weight*7* (kg*7*)
AMetLivlBW34
Amount of TCE oxidized in the liver per unit body weight7* (kg*7*)
AMetLivOtherLiv
Amount of TCE oxidized to metabolites other than TCA and TCOH in the liver (mg) per unit liver mass (kg)
AMetLngResp
Amount of TCE oxidized in the respiratory tract (mg) per unit respiratory tract tissue mass (kg)
AUCCBld
Area under the curve of the venous blood concentration of TCE (mg-h/1)
AUCCTCOH
Area under the curve of the blood concentration of TCOH (mg-h/1)
AUCLivTCA
Area under the curve of the liver concentration of TCA (mg-h/1)
TotMetabBW34
Total amount of TCE metabolized (mg) per unit body weight*7* (kg*7*)
TotOxMetabBW34
Total amount of TCE oxidized (mg) per unit body weight7* (kg*7*)
TotTCAInBW
Total amount of TCA produced (mg) per unit body weight (kg)
3.5.5 Bayesian estimation of PBPK model parameters, and their uncertainty and
variability
3.5.5.1 Updated Pharmacokinetic Database
An extensive search was made for data not previously considered in the PBPK modeling
of TCE and metabolites, with a few studies identified or published subsequent to the review by
Chiu et al. (2006). The studies considered for analysis are listed in Tables 3.5.4-3.5.5, along
with an indication of whether and how they were used.
The least amount of data was available for mice, so an effort was made to include as
many studies as feasible for use in calibrating the PBPK model parameters. Exceptions include
mouse studies with CH or DCA dosing, since those metabolites are not included in the PBPK
model. In addition, the Birner et al. (1993) data only reported urine concentrations, not the
amount excreted in urine. Because there is uncertainty as to total volume of urine excreted, and
over what time period, these data were not used. Moreover, many other studies had urinary
excretion data, so this exclusion should have minimal impact. Several data sets not included by
Hack et al. (2006) were used here. Of particular importance was the inclusion of TCA and
TCOH dosing data from Abbas et al. (1997), Green and Prout (1985), Larson and Bull (1992a),
and Templin et al. (1993).
A substantial amount of data are available in rats, so some data that appeared to be
redundant was excluded from the calibration set and saved for comparison with posterior
predictions (a "validation" set). In particular, those used for "validation" are one closed-chamber
experiment (Andersen et al. 1987), several data sets with only TCE blood data (D'Souza et al.
1985, Jakobson et al. 1986, Lee et al. 1996, and selected time courses from Fisher et al. 1991 and
6/22/2009
141
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Lee et al. 2000a, b), and one unpublished data set (Bruckner et al., unpublished). The Andersen
et al. (1987) data was selected randomly from the available closed chamber data, while the other
datasets were selected because they unpublished or because they more limited in scope (e.g.,
TCE blood only) and so were not as efficient for use in the computationally-intensive calibration
stage. As with the mouse analyses, TCA and TCOH dosing data were incorporated to better
calibrate those pathways.
The human pharmacokinetic database of controlled exposure studies is extensive but also
more complicated. For the majority of the studies, only grouped or aggregated data were
available, and most of those data were saved for "validation" since there remained a large
number of studies for which individual data were available. However, some data that may be
uniquely informative are only available in grouped form, in particular DCVG blood
concentrations, NAcDCVC urinary excretion, and data from TCA and TCOH dosing. In
addition, several human data sets, while having individual data, involved sparse collection at
only one or a few time points per exposure (Bartonicek 1962, Bloemen et al. 2001) and were
subsequently excluded to conserve computational resources. Lapare et al. (1995), which
involved multiple, complex exposure patterns over the course of a month and was missing the
individual urine data, was also excluded due to the relatively low amount of data given the large
computational effort required to simulate it. Finally, data involving exercise during exposure
were excluded, since the model does not include changes in cardiac output, ventilation, and
regional blood flow associated with increased activity. Even with these exclusions, data on a
total of 42 individuals, some involving multiple exposures, were included in the calibration.
6/22/2009
142
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.5.4. Rodent studies with pharmacokinetic data considered for analysis.
Reference Species (strain) Sex TCE exposures Other exposures Calibration Validation Not Comments
used
Mouse studies
Abbas et al. 1996
Mouse (B6C3F1)
M
-
CH iv
V
CH not in model
Abbas and Fisher 1997
Mouse (B6C3F1)
M
Oral (corn oil)
-
V1
Abbas et al. 1997
Mouse (B6C3F1)
M
-
TCOH, TCA iv
V
Barton et al.1999
Mouse (B6C3F1)
M
-
DCA iv and oral
V
DCA not in model
(aqueous)
Birner et al. 1993
Mouse (NMRI)
M+F
Gavage
-
V
Only urine concentrations available,
not amount.
Fisher and Allen 1993
Mouse (B6C3F1)
M+F
Gavage (corn oil)
-
V
Fisher et al. 1991
Mouse (B6C3F1)
M+F
Inhalation
-
V1
Green and Prout 1985
Mouse (B6C3F1)
M
Gavage (corn oil)
TCA iv
V
Greenberg et al. 1999
Mouse (B6C3F1)
M
Inhalation
-
V1
Larson and Bull 1992a
Mouse (B6C3F1)
M
-
DCA, TCA oral
V
Only data on TCA dosing was used,
(aqueous)
since DCA is not in the model
Larson and Bull 1992b
Mouse (B6C3F1)
M
Oral (aqueous)
-
V
Merdink et al.1998
Mouse (B6C3F1)
M
iv
CH iv
V
Only data on TCE dosing was used,
since CH is not in the model.
Prout et al.1985
Mouse (B6C3F1,
M
Gavage (corn oil)
-
V1
Swiss)
Templin et al.1993
Mouse (B6C3F1)
M
Oral (aqueous)
TCA oral
V1
Rat studies
Andersen et al.1987
Rat (F344)
M
Inhalation
-
v1
Barton et al.1995
Rat (SD)
M
Inhalation
-
V
Initial chamber concentrations
unavailable, so not used.
Bernauer et al.1996
Rat (Wistar)
M
Inhalation
-
V1
Birner et al.1993
Rat (Wistar,
M+F
Gavage (ns)
-
V
Only urine concentrations available,
F344)
not amount.
Bruckner et al. unpublished
Rat (SD)
M
Inhalation
-
V
Not published, so not used for
calibration. Similar to Keys et al.
(2003) data.
Dallas et al.1991
Rat (SD)
M
Inhalation
-
V
D'Souza et al.1985
Rat (SD)
M
iv, oral (aqueous)
—
V
Only TCE blood measurements, and
1 Part or all of the data in the study was used for calibration in Hack et al. (2006).
6/22/2009
143
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Species (strain) Sex TCE exposures Other exposures Calibration Validation Not Comments
used
>10-fold greater than other similar
studies.
Fisher et al. 1989
Rat (F344)
F
Inhalation
-
V
Fisher etal.1991
Rat (F344)
M+F
Inhalation
-
v1
V
Experiment with blood only data not
used for calibration.
Green and Prout 1985
Rat (Osborne-
M
Gavage (corn oil)
TCA gavage
V
Mendel)
(aqueous)
Hissink et al.2002
Rat (Wistar)
M
Gavage (corn oil), iv
-
V
Jakobson et al.1986
Rat (SD)
F
Inhalation
Various pretreatments
V
Pre-treatments not included. Only
(oral)
blood TCE data available.
Kaneko et al.1994
Rat (Wistar)
M
Inhalation
Ethanol pretreatment
V
Pre-treatments not included
(oral)
Keys et al.2003
Rat (SD)
M
Inhalation,
-
V
oral (aqueous), ia
Kimmerle and Eben 1973a
Rat (Wistar)
M
Inhalation
-
V
Larson and Bull 1992a
Rat (F344)
M
-
DCA, TCA oral
V
Only TCA dosing data used, since
(aqueous)
DCA is not in the model.
Larson and Bull 1992b
Rat (SD)
M
Oral (aqueous)
-
V1
Lash et al.2006
Rat (F344)
M+F
Gavage (corn oil)
-
V
Highly inconsistent with other studies
Lee et al. 1996
Rat (SD)
M
Arterial, venous,
-
V
Only blood TCE data available
portal, stomach
injections
Lee et al.2000a,b
Rat (SD)
M
Stomach injection, iv,
p-nitrophenol
V
V
Pre-treatments not included. Only
pv
pretreatment (ia)
experiments with blood and liver data
used for calibration.
Merdink et al.1999
Rat (F344)
M
-
CH, TCOH iv
V
TCOH dosing used; CH not in model.
Poet et al.2000
Rat (F344)
M
Dermal
-
V
Dermal exposure not in model.
Prout et al.1985
Rat (Osborne-
M
Gavage (corn oil)
-
v1
Mendel, Wstar)
Saghir et al.2002
Rat (F344)
M
-
DCA iv, oral
V
DCA not in model
(aqueous)
Simmons et al.2002
Rat (Long-
M
Inhalation
-
V
Evans)
Stenner et al.1997
Rat (F344)
M
intraduodenal
TCOH, TCA iv
V
Templin et al.1995
Rat (F344)
M
Oral (aqueous)
-
V1
Thrall et al.2000
Rat (F344)
M
iv, ip
with tolune
V
Only exhaled breath data available
from iv study, ip dosing not in model.
6/22/2009
144
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Species (strain) Sex TCE exposures Other exposures Calibration Validation Not Comments
used
Yu et a 1.2000 Rat (F344) M ~ TCA iv 1
6/22/2009
145
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.5.5. Human studies with pharmacokinetic data considered for analysis.
Reference Species (number Sex TCE Other exposures Calibration Validation Not Comments
of individuals) exposures Used
Bartonicek 1962
Human (n=8)
M+F
Inhalation
-
V
Sparse data, so not included for calibration to
conserve computational resources.
Bernauer et al.1996
Human
M
Inhalation
V2
Grouped data, but unique in that includes
NAcDCVC urine data.
Bloemen et al.2001
Human (n=4)
M
Inhalation
-
V
Sparse data, so not included for calibration to
conserve computational resources.
Chiu et al. 2007
Human (n=6)
M
Inhalation
V
Ertle etal.1972
Human
M
Inhalation
CH oral
V
Very similar to Muller data.
Fernandez et al.1977
Human
M
Inhalation
-
V
Fisher etal.1998
Human (n=17)
M+F
Inhalation
V2
Kimmerle and Eben 1973b
Human (n=12)
M+F
Inhalation
V
Lapare et al.1995
Human (n=4)
M+F
Inhalation
-
V3
Complex exposure patterns, and only grouped
data available for urine, so used for validation.
Lash et al.1999b
Human
M+F
Inhalation
V
Grouped only, but unique in that DCVG blood
data available (same individuals as Fisher et
al. (1998)),
Monster et al.1976
Human (n=4)
M
Inhalation
V3
Experiments with exercise not included.
Monster et al.1979
Human
M
Inhalation
-
v2
Grouped data only.
Muller etal.1972
Human
ns
Inhalation
-
V
Same data also included in Muller et al.
(1975).
Muller etal.1974
Human
M
Inhalation
CH, TCA, TCOH oral V
V2
TCA and TCOH dosing data used for
calibration, since it is rare to have metabolite
dosing data. TCE dosing data used for
validation, since only grouped data available.
CH not in model.
Muller etal.1975
Human
M
Inhalation
Ethanol oral
V2
Grouped data only.
Paycok et al.1945
Human (n=3)
ns
-
TCA iv V
Poet et al.2000
Human
M+F
Dermal
-
Dermal exposure not in model.
Sato et al.1977
Human
M
Inhalation
-
V
Stewart et al.1970
Human
ns
Inhalation
-
V2
2 Part or all of the data in the study was used for calibration in Hack et al. (2006).
3 Grouped data from this study was used for calibration in Hack et al. (2006), but individual data was used here.
6/22/2009
146
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Treibig et al. 1976 Human ns Inhalation
Vesterberg and Astrand Human M Inhalation
1976
6/22/2009
147
All experiments included exercise, so were not
included.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.5.5.2 Updated Hierarchical Population Statistical Model
Generally, only aggregated pharmacokinetic data (arithmetic mean and standard
deviation or standard error) are available from rodent studies. In the Hack et al. (2006) model,
each simulation was treated as a separate observational unit, so different dosing levels within the
same study were treated separately and assigned different PBPK model parameters. However,
the dose-response data are generally also only separated by sex and strain, and otherwise
aggregated, so the variability that is of interest is interstudy (e.g., lot-to-lot), interstrain, and
intersex variability, rather than interindividual variability. In addition, any particular lot of
animals within a study, which are generally inbred and kept under similarly controlled
conditions, are likely to be relatively homogeneous. Therefore, in the revised model, for rodents,
different animals of the same sex and strain in the same study (or series of studies conducted
simultaneously) were treated as identical, and grouped together. Thus, the predictions from the
population model in rodents simulate "average" pharmacokinetics for a particular "lot" of
rodents of a particular species, strain, and sex.
In humans, however, interindividual variability is of interest, and , furthermore,
substantial individual data are available in humans. However, in some studies, the same
individual was exposed more than once, and, so, those data should be grouped together (in the
Hack et al. 2006 model, they were be treated as different "individuals"). Because the primary
interest here is chronic exposure, and because it would add substantially to the computational
burden, interoccasion variability - changes in pharmacokinetic parameters in a single individual
over time - is not addressed. Thus, the predictions from the population model in humans are the
"average" across different occasions for a particular individual (adult).
Figure A. 1 in Appendix A illustrations the hierarchical structure. Informative prior
distributions reflecting the uncertainty in the population mean and variance, detailed in Appendix
A, were updated from those used in Hack et al. (2006) based on an extensive analysis of the
available literature. Section 3.5.5.3, next, discusses specification of prior distributions in the
case where no data independent of the calibration data exist.
3.5.5.3 Use of interspecies scaling to update prior distributions in the absence of other data
For many metabolic parameters, little or no in vitro or other prior information is available
to develop prior distributions. Initially, for such parameters, noninformative priors in the form of
log-uniform distributions with a range spanning at least 104 were specified. However, in the
time available for analysis (up to about 100,000 iterations), only for the mouse did all these
parameters achieve adequate convergence. Additional preliminary runs indicated replacing the
log-uniform priors with lognormal priors and/or requiring more consistency between species
6/22/2009
148
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
could lead to adequate convergence. However, an objective method of "centering" the
lognormal distributions that did not rely on the in vivo data (e.g., via visual fitting or limited
optimization) being calibrated against was necessary in order to minimize potential bias.
Therefore, the approach taken was to consider three species sequentially, from mouse to
rat to human, and to use inter-species scaling to update the prior distributions across species.
This sequence was chosen because the models are essentially "nested" in this order, the rat
model adds to the mouse model the "downstream" GSH conjugation pathways, and the human
model adds to the rat model the intermediary DCVG compartment. Therefore, for those
parameters with little or no independent data only, the mouse posteriors were used to update the
rat priors, and both the mouse and rat posteriors were used to update the human priors. Table
3.5.6 contains a list of the parameters for which this scaling was used to update prior
distributions. The scaling relationship is defined by the "scaled parameters" listed in Appendix
A (Section A.4.1, Tables A.4a-A.4.g), and generally follows standard practice. For instance,
VMax and clearance rates scale by body weight to the % power, whereas KM values are assumed
to not scale, and rate constants (inverse time units) scale by body weight to the -Vi power.
The scaling model is given explicitly as follows. If 9, are the "scaled" parameters
(usually also natural-log-transformed) that are actually estimated, and A is the "universal"
(species-independent) parameter, then 9, = A + e,, where 8, is the species-specific "departure"
from the scaling relationship, assumed to be normally distributed with variance oe2. Therefore,
the mouse model gives an initial estimate of "A," which is used to update the prior distribution
for 9r = A + sr in the rat. The rat and mouse together then give a "better" estimate of A, which is
used to update the prior distribution for Qh = A + eh in the human, with the assumed distribution
for 8h- The mathematical details are given in Appendix A, but two key points in this model are
worth noting here:
— It is known that inter-species scaling is not an exact relationship, and that, therefore, in any
particular case it may either an over- or underestimate. Therefore, the variance in the new
priors reflect a combination of (i) the uncertainty in the "previous" species' posteriors as well
as (ii) a "prediction error" that is lognormally distribution with geometric standard deviation
(GSD) of 3.16-fold, so that the 95% confidence range about the central estimate spans 100-
fold. This choice was dictated partially by practicality, as larger values of the GSD used in
preliminary runs did not lead to adequate convergence within the time available for analysis.
— The rat posterior is a product of its prior (which is based on the mouse posterior) and its
likelihood. Therefore, using the rat and mouse posteriors together to update the human priors
would use the mouse posterior "twice." Therefore, the rat posterior is disaggregated into its
prior and its likelihood using a lognormal approximation (since the prior is lognormal), and
6/22/2009
149
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 only the (approximate) likelihood is used along with the mouse posterior to develop the
2 human prior.
3 With this methodology for updating the prior distributions, adequate convergence was
4 achieved for the rat and human after 110,000-140,000 iterations (discussed further below).
6/22/2009
150
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
2 Table 3.5.6. Parameters for which scaling from mouse to rat, or from mouse and rat to human, was used to update the prior
3 distributions.
Parameter with no or highly uncertain a priori data Mouse RatMouse+ Comments
-> Rat Human Rat ->
Human
Respiratory lumen->tissue diffusion flow rate
V
V
No a priori information
TCOG body/blood partition coefficient
V
V
Prior centered on TCOH data, but highly uncertain
TCOG liver/body partition coefficient
V
V
Prior centered on TCOH data, but highly uncertain
Fraction of hepatic TCE oxidation not to TCA+TCOH
V
V
No a priori information
VMax for hepatic TCE GSH conjugation
V
Rat data on at 1 and 2 mM. Human data at more
KM for hepatic TCE GSH conjugation
V
concentrations, so VMax and KM can be estimated.
VMax for renal TCE GSH conjugation
V
Rat data on at 1 and 2 mM. Human data at more
KM for renal TCE GSH conjugation
V
concentrations, so VMax and KM can be estimated.
VMax for Tracheo-bronchial TCE oxidation
V
V
Prior based on activity at a single concentration
KM for Tracheo-bronchial TCE oxidation
V
V
No a priori information
Fraction of respiratory oxidation entering systemic circulation
V
V
No a priori information
VMax for hepatic TCOH^TCA
V
V
No a priori information
KM for hepatic TCOH^TCA
V
V
No a priori information
VMax for hepatic TCOH^TCOG
V
V
No a priori information
KM for hepatic TCOH^TCOG
V
V
No a priori information
Rate constant for hepatic TCOH->other
V
V
No a priori information
Rate constant for TCA plasma->urine
V
V
Prior centered at GFR, but highly uncertain
Rate constant for hepatic TCA->other
V
V
No a priori information
Rate constant for TCOG liver->bile
V
V
No a priori information
Lumped rate constant for TCOG bile->TCOH liver
V
V
No a priori information
Rate constant for TCOG->urine
V
V
Prior centered at GFR, but highly uncertain
Lumped rate constant for DCVC->Urinary NAcDCVC
V
Not included in mouse model
Rate constant for DCVC bioactivation
V
Not included in mouse model
4 See Appendix A, Table A4a-g for scaling relationships.
5
6/22/2009
151
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.5.5.4 Implementation
The PBPK model was coded in for use in the MCSim software (version 5.0.0), which was
developed particularly for implementing MCMC simulations. As a QC check, results were
checked against the original Hack et al. (2006) model, with the original structures restored and
parameter values made equivalent, and the results were within the error tolerances of the ODE
solver after correcting an error in the Hack et al. (2006) model for calculating the TCA liver
plasma flow. In addition, the model was translated to MatLab (version 7.2.0.232) with
simulation results checked and found to be within the error tolerances of the ODE solver
(odel5s). Mass balances were also checked using the baseline parameters, as well as parameters
from preliminary MCMC simulations, and found to be within the error tolerances of the ODE
solver. Appendix A contains the MCSim model code
3.5.6 Evaluation of Updated PBPK model
3.5.6.1 Convergence
As in previous similar analyses (Gelman et al. 1996; Bois 2000a; 2000b; Hack et al.
2006; David et al. 2006), the potential scale reduction factor "R" is used to determine whether
different independent MCMC chains have converged to a common distribution. The R
diagnostic is calculated for each parameter in the model, and represents the factor by which the
standard deviation or other measure of scale of the posterior distribution (such as a confidence
interval) may be potentially be reduced with additional samples (Gelman et al. 2004). This
convergence diagnostic declines to 1 as the number of simulation iterations approaches infinity,
so values close to 1 indicate approximate convergence, with values of 1.1 and below commonly
considered adequate (Gelman et al. 2004). However, as an additional diagnostic, the
convergence of model dose metric predictions was also assessed. Specifically, dose metrics for a
number of generic exposure scenarios similar to those used in long-term bioassays were
generated, and their natural log (due to their approximate lognormal posterior distributions) was
assessed for convergence using the potential scale reduction factor "R." This is akin to the idea
of utilizing sensitivity analysis so that effort is concentrated on calibrating the most sensitive
parameters for the purpose of interest. In addition, predictions of interest which do not
adequately converge can be flagged as such, so that the statistical uncertainty associated with the
limited sample size can be considered.
The mouse model had the most rapid reduction in potential scale reduction factors.
Initially, four chains of 42,500 iterations each were run, with the first 12,500 discarded as
"burn-in" iterations. At this point, evaluating the 30,000 remaining iterations, all the population
6/22/2009
152
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
parameters except for the VMax for DCVG formation had R < 1.2, with only the first-order
clearance rate for DCVG formation and the VMax and KM for TCOH glucuronidation having R >
1.1. Each chain was then restarted and run for an additional 68,700-71,400 iterations (chains
were terminated at the same time, so the number of iterations per chains was slightly different).
For these iterations, all values of R were < 1.03. Dose metric predictions calculated for exposure
scenarios 10-600 ppm either continuously or 7 hr/d, 5 d/wk and 10-3000 mg/kg-d either
continuously or by gavage 5 d/wk. These predictions were all adequately converged, with all
values of R < 1.03.
As discussed above, for parameters with little or no a priori information, the posterior
distributions from the mouse model were used to update prior distributions for the rat model,
accounting for both the uncertainty reflected in the mouse posteriors as well as the uncertainty in
interspecies extrapolation. Four chains were run to 111,960-128,000 iterations each (chains
were terminated at the same time and run on computers with slightly different processing speeds,
so the number of iterations per chains was slightly different). The first 64,000 iterations were
discarded as "burn-in" iterations, and the remaining iterations were used for inferences. For
these remaining iterations, the diagnostic R was < 1.1 for all population parameters except the
fraction of oxidation not producing TCA or TCOH (R = 1.44 for population mean, R = 1.35 for
population variance), the KM for TCOH -> TCA (R = 1.19 for population mean), the Vmax and
Km for TCOH glucuronidation (R=1.23 and 1.12, respectively for population mean, and R=1.13
for both population variances), and the rate of "other" metabolism of TCOH (R = 1.29 for
population mean and R = 1.18 for population variance). Due to resource constraints, chains
needed to be stopped at this point. However, these are similar to the degree of convergence
reported in Hack et al. (2006). Dose metric predictions calculated for exposure scenarios 10-600
ppm either continuously or 7 hr/d, 5 d/wk, 10-3000 mg/kg-d either continuously or by gavage
5 d/wk.
All dose metric predictions had R < 1.04, except for the amount of "other" oxidative
metabolism (i.e., not producing TCA or TCOH), which had R = 1.12-1.16, depending on the
exposure scenario. The poorer convergence of this dose metric is expected given that a key
determining parameter, the fraction of oxidation not producting TCA or TCOH, had the poorest
convergence among the population parameters.
For the human model, a set of four chains was run for 74,160-84,690 iterations using
"preliminary" updated prior distributions based on the mouse posteriors and preliminary runs of
the rat model. Once the rat chains were completed, final updated prior distributions were
calculated and the last iteration of the preliminary runs were used as starting points for the final
runs. The center of the final updated priors shifted by less than 25% of the standard deviation of
either the preliminary or revised priors, so that the revised median was between the 40th
6/22/2009
153
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
percentile and 60th percentile quantiles of the preliminary median, and vice versa. The standard
deviations themselves changed by less than 5%. Therefore, the use of the preliminary chains as a
starting point should introduce no bias, as long as an appropriate burn-in period is used for the
final runs.
The final chains were run for an additional 59,140-61,780 iterations, at which point, due
to resource constraints, chains needed to be stopped. The first 20,000 iterations were discarded
as "burn-in" iterations, and for the remaining -40,000 iterations, all population mean parameters
had R<1.1 except for the respiratory tract diffusion constant (R = 1.20), the livenblood partition
coefficient for TCOG (R = 1.23), the rate of TCE clearance in the kidney producing DCVG
(R=1.20), and the rate of elimination of TCOG in bile (R = 1.46). All population variances also
had R<1.1 except for the variance for the fraction of oxidation not producing TCOH or TCA
(R=1.10). Dose metric predictions assessed for continuous exposure scenarios at 1-60 ppm in
air or 1-300 mg/kg-d orally. These predictions were all adequately converged with all values of
R < 1.02.
3.5.6.2 Evaluation of posterior parameter distributions
Posterior distributions of the population parameters need to be checked as to whether
they appear reasonable given the prior distributions. Inconsistency between the prior and
posterior distributions may indicate insufficiently broad (i.e., due to overconfidence) or
otherwise incorrectly specified priors, a misspecification of the model structure (e.g., leading to
pathological parameter estimates), or an error in the data. As was done with the evaluation of
Hack et al. (2006) in Appendix A, parameters were flagged if the interquartile regions of their
prior and posterior distributions did not overlap.
Appendix A contains detailed tables of the "sampled" parameters, and their prior and
posterior distributions. Because these parameters are generally scaled one or more times to
obtain a physically meaningful parameter, they are difficult to interpret. Therefore, in Tables
3.5.7-3.5.11, the prior and posterior distributions for the PBPK model parameters obtained after
scaling are summarized. Note that because these model parameters are at the individual (for
humans) or sex/species/study unit (for rodents) level, they were generated using the uncertainty
distributions for the population mean and variance, and hence the distributions reflect both
uncertainty in the population characteristics as well as variability in the population.
Furthermore, they account for correlations among the population-level parameters.
The prior and posterior distributions for most physiological parameters were similar
(Table 3.5.7). Only in the case of the diffusion rate from the respiratory lumen to the respiratory
tissue were the posterior distribution substantially narrower (i.e., less uncertainty) than the prior
6/22/2009
154
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
distribution, which also was to be expected given the very wide, noninformative prior for that
parameter.
For distribution parameters (Table 3.5.8), there were only relatively minor changes
between prior and posterior distributions for TCE and TCOH partition coefficients. The
posterior distributions for several TCA partition coefficients and plasma binding parameters
were substantially narrower than their corresponding priors, but the central estimates were
similar, meaning that values at the high and low extremes were not likely. For TCOG as well,
partition coefficient posterior distributions were substantially narrower, which was expected
given the greater uncertainty in the prior distributions (TCOH partition coefficients were used as
a proxy). Again, posterior distributions indicated that the high and low extremes were not likely.
Finally, posterior distribution for the distribution volume for DCVG was substantially narrower
than the prior distribution, which only provided a lower bound given by the blood volume. In
this case, the upper bounds were substantially lower in the posterior, particularly for humans in
which there are measurements of DCVG in blood.
Posterior distributions for oral absorption parameters (Table 3.5.9) in mice and rats (there
were no oral studies in humans) were also informed by the data, as reflected in their being
substantially more narrow than the corresponding priors. Finally, with a few exceptions, TCE
and metabolite kinetic parameters (Tables 3.5.10-3.5.11) showed substantially narrower posterior
distributions than prior distributions, indicating that they were fairly well specified by the in vivo
data. The exceptions were the VMax for hepatic oxidation in humans (for which there was
substantial in vitro data) and the VMax for respiratory metabolism in mice and rats (although the
posterior distribution for the KM for this pathway was substantially narrower than the
corresponding prior).
In terms of general consistency between prior and posterior distributions, in only a few
cases did the interquartile regions of the prior and posterior distributions not overlap. In most of
these cases, including the diffusion rate from respiratory lumen to tissue, the KMs for renal TCE
GSH conjugation and respiratory TCE oxidation, and several metabolite kinetic parameters, the
prior distributions themselves were non-informative. However, for a noninformative prior, the
lack of overlap would only be an issue if the posterior distributions were affected by the
truncation limit, which was not the case here. The only other parameter for which there was a
lack of interquartile overlap between the prior and posterior distribution was the KM for hepatic
TCE oxidation in mice and in rats, though the prior and posterior 95% confidence intervals did
overlap within each species. As discussed section 3.3, there is some uncertainty in the
extrapolation of in vitro KM values to in vivo values (within the same species). In addition, in
mice, it has been known for some time that KM values appear to be discordant among different
studies (Abbas and Fisher 1997, Greenberg et al. 1999, Fisher et al. 1991).
6/22/2009
155
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 In sum, the Bayesian analysis of the updated PBPK model and data exhibited no major
2 inconsistencies in prior and posterior parameter distributions. The most significant issue was the
3 KM for hepatic oxidative metabolism, for which the posterior estimates were low compared to,
4 albeit somewhat uncertain, in vitro estimates, and it could be argued that a wider prior
5 distribution would have been better. However, the central estimates were not at or near the
6 truncation boundary, so it is unlikely that wider priors would change the results substantially.
7 Therefore, there were no indications based on this evaluation of prior and posterior distributions
8 either that prior distributions were overly restrictive or that model specification errors led to
9 pathological parameter estimates.
10
6/22/2009
156
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.5.7 Physiological Parameters
Mouse
Rat
Human
Prior
Posterior
Prior
Posterior
Prior
Posterior
Median ( 2.5%,
Parameter Description
PBPK Parameter
Median ( 2.5% , 97.5% )
Median (2.5% ,97.5% )
Median ( 2.5% , 97.5% )
Median ( 2.5% , 97.5% )
Median (2.5% ,97.5% )
97.5% )
Cardiac output (L/hr)
QC
0.84(0.49, 1.4)
1 (0.46 ,1.7)
5.4 (3.7 , 7.9)
6.4 (3.5, 9.1 )
390 ( 230 , 670 )
340 (190,720)
Alveolar ventilation (L/hr)
QP
2.1 (0.99 , 4.4)
2.1 (0.84 , 4.5)
10(4.3,25)
7.6(3.4, 19)
370 (170,780)
440 (170, 1100)
Scaled fat blood flow
QFatC
0.07(0.012,0.13)
0.073 (0.015 , 0.13)
0.07 (0.012 0.13)
0.081 (0.023,0.13)
0.05 (0.0082 , 0.092)
0.044 (0.0076,0.09)
Scaled gut blood flow
QGutC
0.14(0.098,0.18)
0.16 (0.11 , 0.19)
0.15 (0.11 , 0.2)
0.17 (0.12 , 0.2)
0.19 (0.13 , 0.25)
0.16(0.12,0.22)
Scaled liver blood flow
QLivC
0.02 (0.014 , 0.026)
0.021 (0.014 , 0.026)
0.021 (0.015 , 0.027)
0.022 (0.015 , 0.027)
0.064 (0.012 , 0.12)
0.039 ( 0.0087 , 0.091 )
Scaled slowly perfused blood
flow
QSIwC
0.22 (0.1 ,0.33)
0.21 (0.1 , 0.33)
0.34 (0.15 , 0.52)
0.31 (0.15,0.5)
0.22 (0.094 0.35)
0.17 (0.085,0.3)
Scaled rapidly perfused
blood flow
QRapC
0.46 (0.31 , 0.61 )
0.44 (0.3 , 0.59)
0.28 (0.073 0.49)
0.28 (0.074,0.45)
0.28 (0.11 , 0.46)
0.39(0.23 , 0.51 )
Scaled kidney blood flow
QKidC
0.091 (0.038 0.14)
0.09 (0.038, 0.14)
0.14 (0.11 , 0.17)
0.14(0.11 ,0.17)
0.19 (0.15 , 0.23)
0.19(0.15,0.23)
Respiratory lumen:tissue
diffusive clearance rate (L/hr)
DResp
0.02(0.000027, 16)
2.5 (0.8 , 7.2)
10(0.4, 100)
21 (6.6, 74)
570 ( 35 , 3900 )
270 ( 63 , 930 )
Fat fractional compartment
volume
VFatC
0.07(0.014,0.13)
0.089 (0.029 , 0.13)
0.07 (0.013 0.13)
0.068 (0.016,0.12)
0.2 (0.038 , 0.36)
0.16 (0.036 , 0.31 )
Gut fractional compartment
volume
VGutC
0.049 (0.037,0.06)
0.048 (0.037 , 0.06)
0.032 (0.024 , 0.04)
0.031 (0.025,0.039)
0.02 (0.017 0.023)
0.02 (0.017 , 0.023)
Liver fractional compartment
volume
VLivC
0.055 (0.031 0.079)
0.046 (0.03,0.073)
0.034 (0.023 , 0.045)
0.033 (0.023 , 0.044)
0.025 (0.015 , 0.035)
0.026 (0.016,0.035)
Rapidly perfused fractional
compartment volume
VRapC
0.1 (0.082 , 0.12)
0.1 (0.082 , 0.12)
0.088 (0.069 , 0.11
0.088 (0.07 , 0.11 )
0.088 (0.075 , 0.1 )
0.088 (0.076,0.099)
Fractional volume of
respiratory lumen
VRespLumC
0.0047 (0.0037 , 0.0056)
0.0047 (0.0038 , 0.0056)
0.0047 (0.0031 , 0.0062)
0.0047 (0.0033, 0.0061 )
0.0024 (0.0015,0.0033)
0.0024 (0.0016 , 0.0032)
Fractional volume of
respiratory tissue
VRespEffC
0.0007 (0.00056 , 0.00084)
0.0007 (0.00056 , 0.00084)
0.0005 ( 0.00034 , 0.00066 ) 0.0005 ( 0.00035 , 0.00066 )
0.00018 (0.00011 ,
0.00025)
0.00018 (0.00012 , 0.00024
)
Kidney fractional
compartment volume
VKidC
0.017 (0.014 0.02)
0.017 (0.014 , 0.02)
0.007 (0.0051 , 0.0089)
0.007 (0.0052 , 0.0088)
0.0043 (0.003,0.0056)
0.0043(0.0031 , 0.0055)
Blood fractional
compartment volume
VBIdC
0.049 (0.038 0.06)
0.049 (0.039 , 0.059)
0.074 (0.058 , 0.09)
0.074 (0.059,0.09)
0.077 (0.06 0.094)
0.078 (0.062 , 0.092)
Slowly perfused fractional
compartment volume
VSIwC
0.55 (0.48 , 0.62)
0.54 (0.48 , 0.61 )
0.59 (0.53 , 0.66)
0.6 (0.54 , 0.66)
0.44 (0.28 , 0.61 )
0.48(0.32 , 0.61 )
Plasma fractional
compartment volume
VPIasC
0.025 (0.012 0.041 )
0.022 (0.012 , 0.036)
0.039 (0.019 , 0.062)
0.04(0.023,0.059)
0.043 (0.033 , 0.055)
0.044 (0.035 , 0.054)
TCA Body fractional
compartment volume [not
VBodC
0.79(0.76 , 0.81 )
0.79 (0.77 , 0.81 )
0.79 (0.77 , 0.81 )
0.79 (0.77 , 0.81 )
0.75 (0.73 , 0.77)
0.75(0.74,0.77)
incl. blood+liver]
TCOH/G Body fractional
compartment volume [not
VBodTCOHC
0.83(0.81 , 0.86)
0.84 (0.82 , 0.86)
0.87 (0.85 , 0.88)
0.87 (0.86 , 0.88)
0.83 (0.82 , 0.84)
0.83(0.82,0.84)
incl. liver]
6/22/2009
157
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.5.8 Distribution Parameters
Mouse
Rat
Human
Prior
Posterior
Prior
Posterior
Prior
Posterior
Median ( 2.5%,
Parameter Description
PBPK Parameter
Median ( 2.5% , 97.5% )
Median (2.5% ,97.5% )
Median ( 2.5% , 97.5% )
Median ( 2.5% , 97.5% )
Median (2.5% ,97.5% )
97.5% )
TCE Blood/air partition
coefficient
PB
15(8.2,27)
14 (7.5 , 29)
22(12 , 41 )
19 (11 , 34)
9.6 (5.9 , 16)
9.3(6.2, 14)
TCE Fat/Blood partition
coefficient
PFat
36(17,74)
35(18 , 71 )
27 (13,56)
31 (17 , 57)
67 (41 , 110)
57 ( 41 , 87 )
TCE Gut/Blood partition
coefficient
PGut
1.9 (0.72 , 5.1 )
1.5 (0.71 , 3.8)
1.4 (0.53 , 3.7)
1.2 (0.55 , 2.7)
2.6 (0.99 , 6.8)
2.8 (1.2 , 6.1 )
TCE Liver/Blood partition
coefficient
PLiv
1.7(0.65,4.5)
2.2 (0.82 , 4.7)
1.5(1 ,2.2)
1.5 (1.1 , 2.1 )
4.1 (1.5, 11 )
4.1 (2 , 8.3)
TCE Rapidly perfused/Blood
partition coefficient
PRap
1.9 (0.72 , 5)
1.8 (0.77 , 4.5)
1.3 (0.5 , 3.4)
1.3 (0.56 , 3)
2.6 (0.99 , 6.8)
2.4(1 ,6.2)
TCE Respiratory tissue:air
partition coefficient
PResp
2.6 (0.98 , 6.8)
2.5(1.1 ,6.2)
1 (0.38,2.6)
1 (0.45 , 2.3)
1.3 (0.5 , 3.5)
1.3 (0.64 , 2.7)
TCE Kidney/Blood partition
coefficient
PKid
2.1 (0.8, 5.6)
2.7 (0.9 , 6.1 )
1.3 (0.63 , 2.7)
1.2 (0.66 , 2.3)
1.6 (0.98 , 2.6)
1.6(1.1 ,2.3)
TCE Slowly perfused/Blood
partition coefficient
PSIw
2.4(0.92,6.4)
2.2 (0.96 , 5.6)
0.58 (0.28 ,1.2)
0.72(0.37, 1.3)
2.1 (1 , 4.4)
2.4 (0.96 , 4.9)
TCA blood/plasma
concentration ratio
TCAPIas
0.8 (0.35 , 19)
1.1 (0.65 , 2.6)
0.79 (0.53 , 1.1 )
0.78(0.61 0.97)
0.78 (0.53 , 18)
0.64 (0.54 , 2.7)
Free TCA Body/blood
plasma partition coefficient
PBodTCA
0.82(0.21 , 19)
0.89 (0.4 , 2.5)
0.7 (0.12 , 3.9)
0.77(0.24,2.7)
0.5 (0.15 , 10)
0.43 (0.2 , 1.7)
Free TCA Liver/blood
plasma partition coefficient
PLivTCA
1.1 (0.3, 25)
1.1 (0.48 , 3.1 )
0.92 (0.16 , 5.1 )
1.2 (0.31 , 4)
0.63 (0.2 , 13)
0.54 (0.26 , 2.3)
Protein/TCA dissociation
constant (umole/L)
kDissoc
110 (5.8 , 2000)
130 (11 ,1600)
280 (62 , 1200)
270 ( 76 , 860 )
180 (160,210)
180 (160, 200)
Maximum binding
concentration (umole/L)
BMax
95(4.1 , 2200)
140 (9.3 , 2200)
330 (50 , 2100)
320 (68 , 1400)
840 (530 ,1300)
740 (520 , 1100)
TCOH body/blood partition
coefficient
PBodTCOH
1.1 (0.49 , 2.5)
0.89 (0.48 , 1.9)
1.1 (0.2 , 5.9)
1 (0.26 , 3.8)
0.9 (0.4 , 2)
1.5 (0.76 , 2.4)
TCOH liver/body partition
coefficient
PLivTCOH
1.3 (0.58 , 2.9)
1.9 (0.74 , 3.4)
1.3 (0.24 , 7.1 )
1.2 (0.28 , 5.6)
0.6 (0.26 , 1.3)
0.64 (0.34 , 1.1 )
TCOG body/blood partition
coefficient
PBodTCOG
1.1 (0.015 , 84)
0.47 (0.13 , 1.6)
0.47 (0.021 , 15)
1.9 (0.09 , 19)
0.75 (0.03 , 18)
0.69 (0.014,44)
TCOG liver/body partition
coefficient
PLivTCOG
1.3 (0.017 , 100)
1.3 (0.36 , 4.6)
1.3 (0.052 , 33)
9.7(2.4,47)
1.7 (0.092 , 29)
3.1 (0.074 , 43)
DCVG effective volume of
distribution
VDCVG
-
-
-
-
64 (4.8 , 37000)
6.1 (4.8, 7.8)
6/22/2009
158
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.5.9 Absorption Parameters
Mouse
Rat
Human
Prior
Posterior
Prior Posterior
Prior Posterior
Parameter Description
PBPK Parameter
Median ( 2.5% , 97.5% )
Median ( 2.5% , 97.5% )
Median ( 2.5% , 97.5% ) Median ( 2.5% , 97.5% )
Median (2.5% , 97.5% ) Me(Jian ( 2.S% ,
97.5% )
TCE Stomach absorption
coefficient (/hr)
TCE Stomach-duodenum
transfer coefficient (/hr)
TCE Duodenum absorption
coefficient (/hr)
TCA Stomach absorption
coefficient (/hr)
TCOH Stomach absorption
coefficient (/hr)
kAS
kTSD
kAD
kASTCA
kASTCOH
1.6 (0.0022 , 890)
1.3 (0.019 , 99)
0.78(0.0012,460)
0.7 (0.0011 , 450)
0.79(0.0012,490)
1.8 (0.052 , 75)
5.2 (0.05 , 98)
0.26 (0.0078 ,15)
3.9 (0.016 , 300)
0.83 (0.0028 ,160)
1.3 (0.0022 , 890)
1.5 (0.019 , 100)
0.71 (0.0011 , 490)
0.77 (0.0012 , 470)
0.64 (0.0012 , 470)
2.4 (0.014 , 310)
3 (0.047,94)
0.19(0.0057,5.3)
1.4 (0.032 , 84)
0.72(0.0064, 110)
0.69 (0.0012 , 480)
0.82 (0.0012 , 490)
4.4 (0.011 , 490)
7.7 (0.022 , 460)
Table 3.5.10 TCE Metabolism Parameters
Mouse
Rat
Human
Prior
Posterior
Prior
Posterior
Prior Posterior
Median ( 2.5%,
Parameter Description
PBPK Parameter
Median ( 2.5% , 97.5% )
Median (2.5% ,97.5% )
Median ( 2.5% , 97.5% )
Median ( 2.5% , 97.5% )
Median (2.5% ,97.5% )
97.5% )
VMax for hepatic TCE
oxidation (mg/hr)
KM for hepatic TCE
oxidation (mg/L)
Fraction of hepatic TCE
oxidation not to TCA+TCOH
Fraction of hepatic TCE
oxidation to TCA
VMax for hepatic TCE GSH
conjugation (mg/hr)
KM for hepatic TCE GSH
conjugation (mg/L)
VMax for renal TCE GSH
conjugation (mg/hr)
KM for renal TCE GSH
conjugation (mg/L)
VMax for Tracheo-bronchial
TCE oxidation (mg/hr)
KM for Tracheo-bronchial
TCE oxidation (mg/L)
Fraction of respiratory
metabolism to systemic circ.
VMax
KM
FracOther
FracTCA
VMaxDCVG
KMDCVG
VMaxKidDCVG
KMKidDCVG
VMaxClara
KMCIara
FracLungSys
4.3 (0.72 , 27)
35(2.3,520)
0.47(0.0015, 1 )
0.07(0.00021 , 0.66)
4.8 (0.0072 , 3300)
220 (0.0043 , 8200000)
0.3 (0.00046 , 200)
180 (0.0043 , 7600000)
0.3 (0.016 , 6)
1.1 (0.0014 , 670)
0.51 (0.0014, 1 )
2.4 (0.7 , 10)
2.7 (0.69 , 23)
0.023 (0.0025 , 0.19)
0.13 (0.052 , 0.31 )
0.65 (0.0084 , 640)
2500 (0.11 , 3700000)
0.029 (0.0011 , 22)
220 (0.11 , 430000)
0.45(0.012, 6.1 )
0.011 (0.0017 , 0.18)
0.79 (0.15 , 1 )
6 (1 , 36 )
21 (0.81 , 610)
0.026 (0.0014 , 0.54)
0.22 (0.024 , 0.74)
2.3 (0.012 , 1500)
1700 (1 ,4000000)
0.038 (0.00024, 13)
480 (0.34 , 760000)
0.19 (0.005 , 4.1 )
0.015 (0.0013 , 0.67)
0.81 (0.036 , 1 )
5.4(1.8, 17)
0.72(0.35,4)
0.28(0.017,0.87)
0.047 (0.0072,0.14)
6.5 (0.15 , 330)
6700 ( 87 , 780000)
0.0025 (0.00042 , 0.02)
0.27 (0.02 , 3.6)
0.2 (0.0056 , 2.3)
0.025 (0.0034 , 0.84)
0.75(0.049,0.99)
430 ( 72 , 2500 )
3.8 (0.11 , 140)
0.12 (0.0058 , 0.77)
0.18 (0.011 , 0.78)
96 (0.0066 , 1200000)
2.9 (0.17 , 50)
170 (0.018 , 1800000)
2.6 (0.15 , 48)
25 (0.84 , 490)
0.022 (0.0016,0.6)
0.75 (0.042 , 0.99)
180 (59 , 930)
0.16(0.017,3.8)
0.1 (0.0064, 0.67)
0.034 (0.0081 , 0.21 )
320 (8.5 , 12000)
3.4(0.16,77)
2.1 (0.035 , 120)
0.78(0.22,7)
17 (0.74, 160)
0.27 (0.0029 , 65)
0.96 (0.81 , 0.99)
6/22/2009
159
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.5.11 Metabolite Metabolism Parameters
Mouse
Rat
Human
Prior
Posterior
Prior
Posterior
Prior
Posterior
Median ( 2.5%,
Parameter Description
PBPK Parameter
Median ( 2.5% , 97.5% )
Median (2.5% ,97.5% )
Median ( 2.5% , 97.5% )
Median ( 2.5% , 97.5% )
Median (2.5% ,97.5% )
97.5% )
VMax for hepatic TCOH-
>TCA (mg/hr)
VMaxTCOH
0.066 (0.000012,450)
0.12 (0.03 , 0.52)
0.67 (0.023 , 21 )
0.71 (0.14 3.8)
42 (0.61 , 3300)
9 (0.83 , 110)
KM for hepatic TCOH->TCA
(mg/L)
KMTCOH
0.85(0.00017,6000)
0.92 (0.2 , 4.1 )
0.94 (0.029 , 33)
19(1.8, 130)
4.8 (0.23 , 100)
2.2 (0.29 , 21 )
VMax for hepatic TCOH-
>TCOG (mg/hr)
VMaxGluc
0.085 (0.000012,430)
4.8(1.4,25)
27(0.8,910)
11 (1.3,120)
820 (11 , 56000 )
890 ( 89 , 5800 )
KM for hepatic TCOH-
>TCOG (mg/L)
KMGluc
1.1 (0.0015 , 670)
34 (2.7 , 200)
28 (0.73,580)
6.1 (0.25 , 54)
11 (0.46 , 250)
130 (20 , 490)
Rate constant for hepatic
TCOH->other (/hr)
kMetTCOH
0.27(0.000038, 1500)
8.7(1.3,36)
4.5 (0.14 , 160)
2.5 (0.25 , 31 )
0.79 (0.036 , 18)
0.26 (0.0046 , 6.9)
Rate constant for TCA
plasma->urine (/hr)
kUrnTCA
25 (0.3 , 2000)
3.1 (0.59 , 15)
1.9 (0.16 , 54)
0.98(0.29,3.5)
0.26 (0.031 , 4.9)
0.12 (0.032,0.45)
Rate constant for hepatic
TCA->other (/hr)
kMetTCA
0.26(0.00036, 160)
1.5 (0.45 , 5)
0.82 (0.026 , 24)
0.47(0.11 1.7)
0.16 (0.0079 , 3.2)
0.1 (0.011 , 0.67)
Rate constant for TCOG
liver->bile (/hr)
kBile
0.25(0.00035, 160)
2.4 (0.5 , 13)
1.3 (0.04 , 44)
12(1.7,64)
1.1 (0.053 , 20)
2.6 (0.55 , 11 )
Lumped rate constant for
TCOG bile->TCOH liver (/hr)
kEHR
0.23(0.00034, 160)
0.036 (0.0024 , 0.16)
0.016 (0.00045,0.69)
1.8 (0.12 , 11 )
0.076 (0.0031 ,1.8)
0.054 (0.016,0.19)
Rate constant for TCOG-
>urine (/hr)
kUrnTCOG
0.67(0.000089,4800)
12 (0.62 , 420)
10 (0.078, 1200)
9.1 (0.27 , 540)
2.6 (0.027 , 230)
2.2 (0.0067 , 640)
Rate constant for hepatic
DCVG->DCVC (/hr)
kDCVG
-
-
-
-
0.034 (0.000053 , 22)
2.5(1.1 ,5.9)
Lumped rate constant for
DCVC->Urinary NAcDCVC
(/hr)
kNAT
-
-
0.13 (0.00021 , 92)
0.003 (0.00048,0.022)
0.00085 (0.00005,0.034)
0.00011 (0.000038 ,
0.00099)
Rate constant for DCVC
bioactivation (/hr)
kKidBioact
-
-
0.14 (0.00021 , 90)
0.0087 (0.00091 , 0.057)
0.0021 (0.000072 , 0.09)
0.023 (0.0036,0.095)
6/22/2009
160
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.5.6.3 Comparison of model predictions with data
As with the Hack et al. (2006) model, initially the sampled group- or individual-specific
parameters were used to generate predictions for comparison to the calibration data (see Figure
3.5.2). Thus, the predictions for a particular dataset are conditioned on the posterior parameter
distributions for same dataset. Because these parameters were "optimized" for each experiment,
these group- or individual-specific predictions should be accurate by design—and, on the whole,
were so. In addition, the "residual error" estimate for each measurement provides some
quantitative measure of the degree to which there were deviations due to intrastudy variability
and model misspecification, including any difficulties fitting multiple dose levels in the same
study using the same model parameters.
Next, only samples of the population parameters (means and variances) were used, and
"new" groups or individuals were sampled from appropriate distribution using these population
means and variances (see Figure 3.5.2). That is, the predictions were only conditioned on the
population-level parameters distributions, representing an "average" over all the datasets, and not
on the specific predictions for that dataset. These "new" groups or individuals then represent the
predicted population distribution, incorporating variability in the population as well as
uncertainty in the population means and variances. Because of the limited amount of mouse
data, all available data for that species was utilized for calibration, and there was no data
available for "out-of-sample" evaluation (often referred to as "validation data," but this term is
not used here due to ambiguities as to its definition). In rats, several studies that contained
primarily blood TCE data, which were abundant, were used for out-of-sample evaluation. In
humans, there were substantial individual and aggregated (group mean) data that was available
for out-of-sample evaluation, as computational intensity limited the number of individuals that
could be used in the MCMC-based calibration.
6/22/2009
161
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
MCMC outputs
Posterior
Posterior population
Posterior I2
Posterior population
prediction/\
PBPK
model,
Posterior group-
specific A
—Posterior group-specific
nrorli^+inn /\
Experiment j
Group/
Individual i
Figure 3.5.2. Schematic of how posterior predictions were generated for comparison with
experimental data. Two sets of posterior predictions were generated: population predictions
(diagonal hashing) and group-specific predictions (vertical hashing). (Same as Figure A.2 in
Appendix A)
6/22/2009
162
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.5.12. Estimates of the residual error
Measurement
Abbreviation
Measurement Description
GSD for "residual" error
(median estimate)
Mouse Rat Human
RetDose
CAIvPPM
ClnhPPM
CMixExh
CArt
CVen
CBIdMix
CFat
CGut
CKid
CLiv
CMus
AExhpost
CTCOH
CLivTCOH
CPIasTCA
CBIdTCA
CLivTCA
AUrnTCA
AUrnTCA_collect
ABileTCOG
CTCOG
CTCOGTCOH
CLivTCOGTCOH
AUrnTCOGTCOH
AUrnTCOGTCOH_
CDCVGmol
AUrnNDCVC
AUrnTCTotMole
TotCTCOH
Notes: GSD =
Retained TCE dose (mg)
TCE concentration in alveolar air (ppm)
TCE concentration in closed chamber (ppm)
TCE concentration in mixed exhaled air (mg/l)
TCE concentration in arterial blood (mg/l)
TCE concentration in venous blood (mg/l)
TCE concentration in mixed arterial and venous blood
(mg/l)
TCE concentration in fat (mg/l)
TCE concentration in gut (mg/l)
TCE concentration in kidney (mg/l)
TCE concentration in liver (mg/l)
TCE concentration in muscle (mg/l)
Amount of TCE exhaled post-exposure (mg)
Free TCOH concentration in blood (mg/l)
Free TCOH concentration in liver (mg/l)
TCA concentration in plasma (mg/l)
TCA concentration in blood (mg/l)
TCA concentration in liver (mg/l)
Cumulative amount of TCA excreted in urine (mg)
Cumulative amount of TCA collected in urine (non-
continuous sampling) (mg)
Cumulative amount of bound TCOH excreted in bile (mg)
Bound TCOH concentration in blood
Bound TCOH concentration in blood in free TCOH
equivalents
Bound TCOH concentration in liver in free TCOH
equivalents (mg/l)
Cumulative amount of total TCOH excreted in urine (mg)
collect Cumulative amount of total TCOH collected in urine
(non-continuous sampling) (mg)
DCVG concentration in blood (mmol/l)
Cumulative amount of NAcDCVC excreted in urine (mg)
Cumulative amount of TCA+total TCOH excreted in urine
(mmol)
Total TCOH concentration in blood (mg/l)
Geometric Standard Deviation. Values higher than
1.1£
2.68
1.61
2.49
2.23
1.71
1.23
1.54
1.59
1.40
1.49
1.34
1.34
1.49
1.63
1.26
1.13
1.44-1.83
1.11-1.12
1.5
1.17-1.52
1.22-4.46 1.62-2.95
1.5
1.85-2.66
1.86
1.47
1.67-1.78
1.65
1.12-1.17
1.14-1.64 1.14-2.1
1.13-1.21 1.12-1.17
1.13-1.59 1.12-1.49
1.67
1.18-1.95 1.11-1.54
2-2.79
2.13
2.76
1.12-2.27 1.11-1.13
1.3-1.63
1.17
1.12-1.54
1.53
1.17
1.85 1.49 1.2-1.69
2-fold are in bold.
3.5.6.3.1 Mouse model and data
Table 3.5.12 provies an evaluation of the predictions of the mouse model for each data
set, with figures showing data and predictions in Appendix A. With exception of the remaining
over-prediction of TCE in blood following inhalation exposure, the parent PBPK model (for
TCE) appears to now be robust in mice. Most of the problems previously encountered with the
Abbas and Fisher (1997) gavage data were solved by allowing absorption from both of the
6/22/2009
163
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
stomach and duodenal compartments. Notably, the addition of possible wash-in/wash-out,
respiratory metabolism, and extrahepatic metabolism (i.e., kidney GSH conjugation) was
insufficient to remove the long-standing discrepancy of PBPK models over-predicting TCE
blood levels, suggesting another source of model or experimental error is the cause. However,
the availability of tissue concentration levels of TCE somewhat ameliorates this limitation.
In terms of TCA and TCOH, the overall mass balance and metabolic disposition to these
metabolites also appeared to be robust, as urinary excretion following dosing with TCE, TCOH,
as well as TCA could be modeled accurately. This improvement over the Hack et al. (2006)
model was likely due in part to the addition of nonurinary clearance ("untracked" metabolism) of
TCA and TCOH. Also, the addition of a liver compartment for TCOH and TCOG, so that first-
pass metabolism could be properly accounted for, was essential for accurate simulation of the
metabolite pharmacokinetics both from iv dosing of TCOH and from exposure to TCE.
These conclusions are corroborated by the estimated "residual" errors, which include
intrastudy variability, interindividual variability, and measurement and model errors. The
implied geometric standard deviation (GSD) for this error in each in vivo measurement is
presented in Table 3.5.12. As expected, the venous blood TCE concentration had the largest
residual error, with a GSD of 2.7, reflecting largely the difficulty in fitting TCE blood levels
following inhalation exposure. In addition, the fat and kidney TCE concentrations also are
somewhat uncertain, with a GSD for the residual error of 2.5 and 2.2, respectively, while other
residual errors had GSD of less than 2-fold. These tissues were only measured in two studies,
Abbas and Fisher (1997) and Greenberg et al. (1999), and the residual error reflects the
difficulties in simultaneously fitting model to the different dose levels with the same parameters.
In terms of total metabolism, closed-chamber data were fit accurately with the updated
model. While the previous analyses of Hack et al. (2006) allowed for each chamber experiment
to be fit with different parameters, the current analysis made the more restrictive assumption that
all experiments in a single study utilize the same parameters. Furthermore, the accuracy of
closed chamber predictions did not require the very high values for cardiac output that were used
by Fisher et al. (1991), confirming the suggestion (discussed in Appendix A) that additional
respiratory metabolism would resolve this discrepancy. The accurate model means that
uncertainty with respect to possible wash-in/wash-out, respiratory metabolism, and extrahepatic
metabolism could be well characterized. For instance, the absence of in vivo data on GSH
metabolism in mice means that this pathway remains relatively uncertain; however, the current
model should be reliable for estimating lower and upper bounds on the GSH pathway flux.
6/22/2009
164
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.5.13. Summary comparison of updated PBPK model predictions and in vivo data in
mice
Study
Exposure(s)
Discussion
Abbas and Fisher 1997
TCE gavage (corn oil)
Abbas etal. 1997
TCOH, TCA iv
Fisher and Allen 1993
TCE gavage (corn oil)
Fisher et al. 1991
TCE inhalation
Green and Prout 1985
TCE gavage (corn oil)
Greenberg et al. 1999
TCE inhalation
Larson and Bull 1992a
TCE gavage (aqueous)
Generally, model predictions were quite good, especially with
respect to tissue concentrations of TCE, TCA, and TCOH. There
were some discrepancies in TCA and TCOG urinary excretion and
TCA and TCOG concentrations in blood due to the requirement
(unlike in Hack et al. 2006) that all experiments in the same study
utilize the same parameters. Thus, for instance, TCOG urinary
excretion was accurately predicted at 300 mg/kg, underpredicted at
600 mg/kg, over-predicted at 1200 mg/kg, and underpredicted again
at 2000 mg/kg, suggesting significant intra-experimental variability
(not addressed in the model).
Population predictions were quite good, with the almost all of the
data within the 95% confidence interval of the predictions, and most
within the inter-quartile region.
Both group-specific and population predictions were quite good.
Urinary excretion, which was over-predicted by the Hack et al. (2006)
model, was accurately predicted due to the allowance of additional
"untracked" clearance. In the case of population predictions, almost
all of the data were within the 95% confidence interval of the
predictions, and most within the inter-quartile region.
Both group-specific and population predictions were quite good.
Some discrepancies in the time-course of TCE blood concentrations
were evidence across doses in the group-specific predictions, but not
in the population predictions, suggesting significant intra-group
variability (not addressed in the model).
Blood TCE levels during and following inhalation exposures
were still over-predicted at the higher doses. However, there was the
stringent requirement (absent in Hack et al. 2006) that the model
utilize the same parameters for all doses and in both the closed and
open chamber experiments. Moreover, the Hack et al. (2006) model
required significant differences in the parameters for the different
closed chamber experiments, while predictions here were accurate
utilizing the same parameters across different initial concentrations.
These conclusions were the same for group-specific and population
predictions (e.g., TCE blood levels remained over-predicted in the
later case).
Both group-specific and population predictions were adequate,
though the data collection was sparse. In the case of population
predictions, almost all of the data were within the 95% confidence
interval of the predictions, and about half within the inter-quartile
region.
Model predictions were quite good across a wide variety of
measures that included tissue concentrations of TCE, TCA, and
TCOH. However, as with the Hack et al. (2006) predictions, TCE
blood levels were over-predicted by up to 2-fold. Population
predictions were quite good, with the exception of TCE blood levels.
Almost all of the other data was within the 95% confidence interval of
the predictions, and most within the inter-quartile region.
Both group-specific and population predictions were quite good,
though the data collection was somewhat sparse. In the case of
population predictions, all of the data were within the 95% confidence
6/22/2009
165
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
interval of the predictions,
Larson and Bull 1992b
TCA gavage (aqueous)
Both group-specific and population predictions were quite good.
In the case of population predictions, most of the data were within the
inter-quartile region.
Merdink et al.1998
TCE iv
Both group-specific and population predictions were quite good,
though the data collection was somewhat sparse. In the case of
population predictions, all of the data were within the 95% confidence
interval of the predictions,
Prout et al.1985
TCE gavage (corn oil)
Both group-specific and population predictions were adequate,
though there was substantial scatter in the data due to the use of
single animals at each data point.
Templin et al.1993
TCE gavage (aqueous)
Both group-specific and population predictions were quite good.
With respect to population predictions, almost all of the other data
was within the 95% confidence interval of the predictions, and most
within the inter-quartile region.
3.5.6.3.2 Rat model and data
A summary evaluation of the predictions of the rat model as compared to the data is
provided in Tables 3.5.14 and 3.5.15, with figures showing data and predictions in Appendix A.
Similar to previous analyses (Hack et al. 2006), the TCE submodel for the rat appears to be
robust, with blood and tissue concentrations accurately predicted. Unlike in the mouse, some
data consisting of TCE blood and tissue concentrations were used for "out-of-sample evaluation"
(sometimes loosely termed "validation"). These data were generally well simulated; most of the
data within the 95% confidence interval of posterior predictions. This provides additional
confidence in the predictions for the parent compound.
In terms of TCA and TCOH, as with the mouse, the overall mass balance and metabolic
disposition to these metabolites also appeared to be robust: urinary excretion following dosing
with TCE, TCOH, as well as TCA, could be modeled accurately, and, secondly, the residual
errors did not indicate substantial mis-fit (GSD < 1.25). This improvement over the Hack et al.
(2006) model was likely due in part to the addition of non-urinary clearance ("untracked"
metabolism) of TCA and TCOH. In addition, the addition of a liver compartment for TCOH and
TCOG, so that first-pass metabolism could be properly accounted for, was essential for accurate
simulation of the metabolite pharmacokinetics both from iv dosing of TCOH and from TCE
exposure. Blood and plasma concentrations of TCA and TCOH were fairly well simulated, with
GSD for the residual error of 1.2-1.3, but a bit more discrepancy was evidence with TCA liver
concentrations. However, TCA liver concentrations were only available in one study (Yu et al.
2000), and the data show a change in the ratio of liver to blood concentrations at the last time
point, which may be the source of the added residual error.
In terms of total metabolism, as with the mouse, closed-chamber data were fit accurately
with the updated model (residual error GSD of about 1.11). In addition, the data on NAcDCVC
urinary excretion was well predicted (residual error GSD of 1.18), in particular the fact that
6/22/2009
166
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
excretion was still ongoing at the end of the experiment. Thus, there is greater confidence in the
estimate of the flux through the GSH pathway than there was from the Hack et al. (2006) model.
However, the overall flux is still estimated indirectly, and there remains some ambiguity as to the
relative contributions respiratory wash-in/wash-out, respiratory metabolism, extrahepatic
metabolism, DCVC bioactivation versus A-acetylation, and oxidation in the liver producing
something other than TCOH or TCA. Therefore there remain a large range of possible values for
the flux through the GSH conjugation and other indirectly estimated pathways that are
nonetheless consistent with all the available in vivo data. The use of noninformative priors for
the metabolism parameters for which there were no in vitro data means that a fuller
characterization of the uncertainty in these various metabolic pathways could be achieved. Thus,
the model should be reliable for estimating lower and upper bounds on several of these
pathways.
Table 3.5.14. Summary comparison of updated PBPK model predictions and in vivo data used
for "calibration" in rats
Study
Exposure(s)
Discussion
Bernauer et al.1996
TCE Inhalation
Dallas et al. 1991
TCE Inhalation
Fisher et al. 1989
TCE Inhalation
Fisher etal.1991
Green and Prout 1985
TCE Inhalation
TCE gavage (corn oil)
TCA iv
TCA gavage (aqueous)
Posterior fits to these data were adequate, especially with the
requirement that all predictions for dose levels utilize the same PBPK
model parameters. Predictions of TCOG and TCA urinary excretion
was relatively accurate, though the time-course of TCA excretion
seemed to proceed more slowly with increasing dose, an aspect not
captured in by model. Importantly, unlike the Hack et al. (2006) results,
the time-course of NAcDCVC excretion was quite well simulated, with
the excretion rate remaining non-negligible at the last time point (48 hr).
It is likely that the addition of the DCVG sub-model between TCE and
DCVC, along with prior distributions that accurately reflected the lack of
reliable independent (e.g., in vitro) data on bioactivation, allowed for the
better fit.
These data, consisting of arterial blood and exhaled breath
concentrations of TCE, were accurately predicted by the model using
both group-specific and population sampled parameters. In the case of
population predictions, most of the data were within the 95% confidence
interval of the predictions.
These data, consisting of closed chamber TCE concentrations,
were accurately simulated by the model using both group-specific and
population sampled parameters. In the case of population predictions,
most of the data were within the 95% confidence interval of the
predictions.
These data, consisting of TCE blood, and TCA blood and urine
time-courses, were accurately simulated by the model using both group-
specific and population sampled parameters. In the case of population
predictions, most of the data were within the 95% confidence interval of
the predictions.
For TCE treatment, these data, consisting of one time point each in
urine for TCA, TCA +TCOG, and TCOG, were accurately simulated by
both group-specific and population predictions.
6/22/2009
167
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Hissink et al.2002
Kaneko et al.1994
TCE Gavage (corn oil)
TCE iv
TCE Inhalation
Keys et al.2003
TCE Inhalation, gavage
(aqueous), ia
Kimmerle and Eben 1973a TCE Inhalation
Larson and Bull 1992a TCA gavage (aqueous)
Larson and Bull 1992b TCE gavage (aqueous)
Lee et al.2000a
Merdink et al. 1999
Prout et al. 1985
TCE iv, pv
TCOH iv
TCE Gavage (corn oil)
For TCA iv treatment, the single datum of urinary TCA+TCOG at
24 hr was at the lower 95% confidence interval in the group-specific
simulations, but accurately predicted with the population sampled
parameters, suggesting intra-study variability is adequately accounted
for by population variability.
For TCA gavage treatment, the single datum of urinary
TCA+TCOG at 24 hr was accurately simulated by both group-specific
and population predictions.
These data, consisting of TCE blood, and TCA+TCOG urinary
excretion time-courses, were accurately simulated by the model using
group-specific parameters. In the case of population predictions,
TCA+TCOH urinary excretion appeared to be somewhat under-
predicted.
These data, consisting of TCE blood and TCA and TCOG urinary
excretion time-courses, were accurately predicted by the model using
both group-specific and population sampled parameters. In the case of
population predictions, TCA+TCOH urinary excretion appeared to be
somewhat underpredicted, However all of the data were within the 95%
confidence interval of the predictions.
These data, consisting of TCE blood, gut, kidney, liver, muscle and
fat concentration time-courses, were accurately predicted by the model
using both group-specific and population sampled parameters. In the
case of population predictions, most of the data were within the 95%
confidence interval of the predictions.
Some inaccuracies were noted in group-specific predictions,
particularly with TCA and TCOG urinary excretion, TCE exhalation post-
exposure, and TCE venous blood concentrations. In the case of TCA
excretion, the rate was underpredicted at the lowest dose (49 mg/kg)
and over-predicted at 330 ppm. In terms of TCOG urinary excretion,
the rate was over-predicted at 175 ppm and underpredicted at 330 ppm.
Similarly for TCE exhaled post-exposure, there was some
overprediction at 175 ppm and some underprediction at 300 ppm.
Finally, venous blood concentrations were over-predicted at 3000 ppm.
However, for population predictions, most of the data were within with
95% confidence region.
These data, consisting of TCA plasma time-courses, were
accurately predicted by the model using both group-specific and
population sampled parameters. In the case of population predictions,
all of the data were within the 95% confidence interval of the
predictions.
These data, consisting of TCE, TCA, and TCOH in blood, were
accurately predicted by the model using both group-specific and
population sampled parameters. In the case of population predictions,
all of the data were within the 95% confidence interval of the
predictions.
These data, consisting of TCE concentration time course in mixed
arterial and venous blood and liver, were predicted using both the group
specific and population predictions. In both cases, most of the data
were within the 95% confidence interval of the predictions.
TCOH blood concentrations were accurately predicted using
group-specific parameters. However, population-based parameters
seemed to lead to some under-prediction, though most of the data were
within the 95% confidence interval of the predictions.
Most of these data were accurately predicted using both group-
6/22/2009
168
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
specific and population-sampled parameters. However, at the highest
two doses (1000 and 2000 mg/kg), there were some discrepancies in
the (very sparsely collected) urinary excretion measurements. In
particular, using group-specific parameters, TCA+TCOH urinary
excretion was under-predicted at 1000 mg/kg and overpredicted at 2000
mg/kg. Using population-sampled parameters, this excretion was
underpredicted in both cases, though not entirely outside of the 95%
confidence interval.
Simmons et al.2002 TCE Inhalation Most of these data were accurately predicted using both group-
specific and population-sampled parameters. In the open chamber
experiments, there was some scatter in the data that did not seem to be
accounted for in the model. The closed chamber data were accurately
fit.
Stenner et al.1997 TCE Intraduodenal These data, consisting of TCA and TCOH in blood and TCA and
TCOH iv TCOG in urine, were generally accurately predicted by the model using
TCOH iv, bile- both group-specific and population sampled parameters. However,
cannulated using group-specific parameters, the amount of TCOG in urine was
overpredicted for 100 TCOH mg/kg iv dosing, though total TCOH in
blood was accurately simulated. In addition, in bile-cannulated rats, the
TCOG excretions at 5 and 20 mg/kg iv were underpredicted, while the
amount at 100 mg/kg was accurately predicted. On the other hand, in
the case of population predictions, all of the data were within the 95%
confidence interval of the predictions, and mostly within the inter-
quartile region, even for TCOG urinary excretion. This suggests that
intra-study variability may be a source of the poor fit in using the group-
specific parameters.
Templin et al.1995 TCE Oral (aqueous) These data, consisting of TCE, TCA, and TCOH in blood, were
accurately predicted by the model using both group-specific and
population sampled parameters. In the case of population predictions,
all of the data were within the 95% confidence interval of the
predictions.
Yu et al.2000 TCA iv These data, consisting of TCA in blood, liver, plasma, and urine,
were generally accurately predicted by the model using both group-
specific and population sampled parameters. The only notable
discrepancy was at the highest dose of 50 mg/kg, in which the rate of
urinary excretion from 0-6 hr appeared to more rapid than the model
predicted. However, all of the data were within the 95% confidence
interval of the predictions based on population-sampled parameters.
1
6/22/2009
169
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.5.15. Summary comparison of updated PBPK model predictions and in vivo data used
for "out-of-sample" evaluation in rats
Study
Exposure(s)
Discussion
Andersen et al.1987
Bruckner et al. unpub
Fisher et al. 1991
Jakobson et al. 1986
Lee et al. 1996
Lee et al.2000a,b
TCE Inhalation
TCE Inhalation
TCE Inhalation
TCE Inhalation
TCE ia, iv, pv, gavage
TCE gavage
These closed chamber data were well within the 95% confidence
interval of the predictions based on population-sampled parameters.
These data on TCE in blood, liver, kidney, fat, muscle, gut, and
venous blood, were generally accurately predicted based on population-
sampled parameters. The only notable exception was TCE in the
kidney during the exposure period at the 500 ppm level, which were
somewhat under-predicted (though levels post-exposure were
accurately predicted).
These data on TCE in blood were well within the 95% confidence
interval of the predictions based on population-sampled parameters.
These data on TCE in arterial blood were well within the 95%
confidence interval of the predictions based on population-sampled
parameters.
Except at some very early time-points (<0.5 hr), these data on TCE
in blood were well within the 95% confidence interval of the predictions
based on population-sampled parameters.
These data on TCE in blood were well within the 95% confidence
interval of the predictions based on population-sampled parameters.
3.5.6.3.3 Human model
Table 3.5.16-3.5.17 provide a summary evaluation of the predictions of the model as
compared to the human data, with figures showing data and predictions in Appendix A. With
respect to the TCE submodel, blood and exhaled air measurements appeared more robust than
previously found from the Hack et al. (2006) model. TCE blood concentrations from most
studies were well predicted. However, those from Chiu et al. (2007) were consistently
overpredicted, and a few of those from Fisher et al. (1998) were consistently underpredicted.
Alveolar or mixed exhaled breath concentrations of TCE from all studies except Fisher et al.
(1998) were well predicted, though the discrepancy appeared smaller than that originally
reported by Fisher et al. (1998) for their PBPK model. In addition, the majority of the "out-of-
sample" evaluation data consisted of TCE in blood or breath, and were generally well predicted,
lending confidence to the model predictions for the parent compound.
In terms of TCA and TCOH, as with the mouse and rat, the overall mass balance and
metabolic disposition to these metabolites also appeared to be robust, as urinary excretion
following TCE exposure could be modeled accurately. However, data from Chiu et al. (2007)
indicated substantial interoccasion variability, as the same individual exposed to the same
concentration on different occasions sometimes had substantial differences in urinary excretion.
Since Chiu et al. (2007) was the only calibration study for which this urine collection was
intermittent, this interoccasion variability was also reflected in the larger residual error (GSD of
6/22/2009
170
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1.55 and 1.59 for TCA and TCOH, respectively—Table 3.5.12) for intermittent urine collection
as compared to cumulative collection (respective residual error GSD of 1.36 and 1.11). Blood
and plasma concentrations of TCA and free TCOH were fairly well simulated, with GSD for the
residual error of 1.1-1.4, though total TCOH in blood had greater residual error with GSD of
about 1.6. This partially reflects the "sharper" peak concentrations of total TCOH in the Chiu et
al. (2007) data relative to the model predictions. In addition, TCA and TCOH blood and urine
data were available from several studies for "out-of-sample" evaluation and were generally well
predicted by the model, lending further confidence to the model predictions for these
metabolites.
In terms of total metabolism, no closed-chamber data exist in humans, but alveolar breath
concentrations were generally well simulated, suggesting that total metabolism may be fairly
robust. In addition, as with the rat, the data on NAcDCVC urinary excretion was well predicted
(residual error GSD of 1.12), in particular the fact that excretion was still ongoing at the end of
the experiment (48 hrs after the end of exposure). Thus, there is greater confidence in the
estimate of the flux through the GSH pathway than there was from the Hack et al. (2006) model,
in which excretion was completed within the first few hours after exposure. However, as with
the rat, the overall flux is still estimated indirectly, and there remains some ambiguity as to the
relative contributions respiratory wash-in/wash-out, respiratory metabolism, extrahepatic
metabolism, DCVC bioactivation versus A-acetylation, and oxidation in the liver producing
something other than TCOH or TCA. However, unlike in the rat, the blood DCVG data, while
highly variable, nonetheless provide substantial constraints (at least a strong lower bound) on the
flux of GSH conjugation. Importantly, the high residual error GSD for blood DCVG reflects the
fact that only grouped or unmatched individual data were available, so in this case, the residual
error includes interindividual variabilty, which is not included in the other residual error
estimates. For the other indirectly estimated pathways, there remain a large range of possible
values that are nonetheless consistent with all the available in vivo data. The use of
noninformative priors for the metabolism parameters for which there were no in vitro data means
that a fuller characterization of the uncertainty in these various metabolic pathways could be
achieved. Thus, as with the rat, the model should be reliable for estimating lower and upper
bounds on several of these pathways.
6/22/2009
171
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.5.16. Summary comparison of updated PBPK model predictions and in vivo data used
for "calibration" in humans
Reference Exposure(s) Discussion
Bernauer et al.1996 TCE Inhalation These data, consisting of TCA, TCOG and NAcDCVC excreted in
urine, were accurately predicted by the model using both individual-specific
and population sampled parameters. The posterior NAcDCVC predictions
were an important improvement over the predictions of Hack et al. (2006),
which predicted much more rapid excretion than observed. The fit
improvement is probably a result of the addition of the DCVG sub-model
between TCE and DCVC, along with the broader priors on DCVC excretion
and bioactivation. Interestingly, in terms of population predictions, the
NAcDCVC excretion data from this study were on the low end, though still
within the 95% CI.
Chiu et al. 2007 TCE Inhalation Overall, posterior predictions were quite accurate across most of
the individuals and exposure occasions. TCE alveolar breath concentrations
were well simulated for both individual-specific and population-generated
simulations, though there was substantial scatter (intra-occasion variability).
However, TCE blood concentrations were consistently over-predicted in most
of the experiments, both using individual-specific and population-generated
parameters. This was not unexpected, as Chiu et al. (2007) noted the TCE
blood measurements to be lower by about 2-fold relative to previously
published studies. As discussed in Chiu et al. (2007), wash-in/wash-out and
extra-hepatic (including respiratory) metabolism were not expected to be able
to account for the difference, and indeed all these processes were added to
the current model without substantially improving the discrepancy.
With respect to metabolite data, TCA and total TCOH in blood were
relatively accurately predicted. There was individual experimental variability
observed for both TCA and TCOH in blood at six hours (end of exposure).
The population-generated simulations overpredicted TCA in blood, while they
were accurate in predicting blood TCOH . Predictions of free TCOH in blood
also showed overprediction for individual experiments, with variability at the
end of exposure timepoint. However, TCOH fits were improved for the
population-generated simulations. TCA and TCOG urinary excretion was
generally well simulated, with simulations slightly under- or over-predicting the
individual experimental data in some cases.
Fisher et al.1998 TCE Inhalation The majority of the predictions for these data were quite accurate.
Interestingly, in contrast to the predictions for Chiu et al. (2007), TCE blood
levels were somewhat underpredicted in a few cases, both from using
individual-specific and population-generated predictions. These two results
together suggest some unaccounted-for study-to-study variance, though inter-
individual variability cannot be discounted as the data from Chiu et al. (2007)
were from individuals in the Netherlands and that from Fisher et al. (1998)
were from individuals in the United States. As reported by Fisher et al.
(1998), TCE in alveolar air was somewhat over-predicted in several cases,
however, the discrepancies seemed smaller than originally reported for the
Fisher et al. (1998) model.
With respect to metabolite data, TCOH and TCA in blood and
TCOG and TCA in urine were generally well predicted, though data for some
individuals appeared to exhibit inter- and/or intra-occasion variability. For
example, in one case in which the same individual (female) was exposed to
both 50 and 100 ppm, the TCOH blood data was overpredicted at the higher
one exposure. In addition, in one individual, initial individual-specific
simulations for TCA in urine were underpredicted but shifted to
6/22/2009
172
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
overpredictions towards the end of the simulations. The population-generated
results overpredicted TCA in urine for the same individual. Given the results
from Chiu et al. (2007), inter-occasion variability is likely to be the cause,
though some dose-related effect cannot be ruled out.
Finally, DCVG data was well predicted in light of the high variability
in the data and availability of only grouped data or data from multiple
individual who cannot be matched to the appropriate TCE and oxidative
metabolite dataset. In all cases, the basic shape (plateau and then sharp
decline) and order of magnitude of the time-course were well predicted,
Furthermore, the range of the data was well-captured by the 95% confidence
interval of the population-generated predictions.
Kimmerle and Eben 1973b TCE Inhalation These data were well fit by the model, using either individual-
specific or population-generated parameters.
Monster et al.1976 TCE Inhalation The data simulated in this case were exhaled alveolar TCE, TCE in
venous blood, TCA in blood, TCA in urine, and TCOG in urine. Both using
individual-specific and population-generated simulations, all fits are within the
95% CI. The one exception was the retained dose for a male exposed to 65
ppm, which was outside the 95% CI for the population-generated results.
Muller et al.1974 TCA, The data measured after oral TCA was timecourse TCA measured
in plasma and urine. Individual-specific predictions were accurate, but both
datasets were overpredicted in the population-generated simulations.
TCOH oral The data measured after oral TCOH was timecourse TCOH in
blood, TCOG in urine, TCA in plasma, and TCA in urine. Individual-specific
predictions were accurate, but the population-generated simulations
overpredicted TCOH in blood and TCOG in urine. The population-based TCA
predictions were accurate.
These results indicate that "unusual" parameter values were
necessary in the individual-specific simulations to give accurate predictions.
Paycok et al.1945 TCA iv These data were well fit by the model, using either individual-
specific or population-generated parameters.
Note: CI = confidence interval.
Table 3.5.17. Summary comparison of updated PBPK model predictions and in vivo data used
for "out-of-sample" evaluation in humans
Reference
Exposure(s)
Discussion
Bartonicek 1962
TCE
Inhalation
While these data were mostly within the 95% confidence interval of the
predictions, they tended to be at the high end for all the individuals in the
study.
Bloemen et al.2001
TCE
Inhalation
These data were all well within the 95% confidence interval of the predictions.
Fernandez et al.1977
TCE
Inhalation
These data were all well within the 95% confidence interval of the predictions.
Lapare et al.1995
TCE
Inhalation
These data were all well within the 95% confidence interval of the predictions.
Monster et al.1979
TCE
Inhalation
These data were all well within the 95% confidence interval of the predictions.
Muller et al.1974, 1975
TCE
Inhalation
Except for TCE in alveolar air, which was overpredicted during exposure,
these data were all well within the 95% confidence interval of the predictions.
Sato et al.1977
TCE
Inhalation
These data were all well within the 95% confidence interval of the predictions.
Stewart et al.1970
TCE
Inhalation
These data were all well within the 95% confidence interval of the predictions.
Treibig et al. 1976
TCE
Inhalation
Except for TCE in alveolar air, these data were all well within the 95%
confidence interval of the predictions.
6/22/2009
173
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.5.6.4 Summary Evaluation of Updated PBPK Model
Overall, the updated PBPK model, utilizing parameters consistent with the available
physiological and in vitro data from published literature, provides reasonable fits to an extremely
large database of in vivo pharmacokinetic data in mice, rats, and humans. Posterior parameter
distributions obtained were by MCMC sampling using a hierarchical Bayesian population
statistical model and a large fraction of this in vivo database. Convergence of the MCMC
samples for model parameters was good for mice, and adequate for rats and humans. In addition,
in rats and humans, the model did produce predications that are consistent with in vivo data from
many studies not used for calibration (insufficient studies were available in mice for such "out of
sample" evaluation).
3.5.7 PBPK Model Dose Metric Predictions
3.5.7.1 Characterization of Uncertainty and Variability
Since it is desirable to characterize the contributions from both uncertainty in population
parameters and variability within the population, so the following procedure is adopted. First
500 sets of population parameters (i.e., population mean and variance for each parameter) are
extracted from the posterior MCMC samples - these represent the uncertainty in the population
parameters. To minimize autocorrelation, they were obtained by "thinning" the chains to the
appropriate degree. From each of these sets of population parameters, 100 sets of "individual,"
or "study group" in the case of rodents, parameters were generated by Monte Carlo - each of
these represents the population variability, given a particular set of population parameters. Thus
a total of 50,000 individuals (or study groups, for rodents), representing 100 (variability) each for
500 different populations (uncertainty), were generated.
Each set was run for a variety of generic exposure scenarios. The combined distribution
of all 50,000 individuals reflects both uncertainty and variability - i.e., the case in which one is
trying to predict the dosimetry for a single random study (for rodents) or individual (for humans).
In addition, for each dose metric, the mean predicted internal dose was calculated from set of the
500 sets of 100 individuals, resulting in a distribution for the uncertainty in the population mean.
Comparing the combined uncertainty and variability distribution with the uncertainty distribution
in the population mean gives a sense of how much of the overall variation is due to uncertainty
versus variability.
Figures 3.5.3-3.5.11 show the results of these simulations for a number of representative
dose metrics across species continuously exposed via inhalation or orally. For display purposes,
dose metrics have been scaled by total intake (resulting in a prediced "fraction" metabolized) or
exposure level (resuling in an internal dose per ppm for inhalation or per mg/kg-d for oral
6/22/2009
174
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
exposures). In these figures, the thin error bars representing the 95% confidence interval for
overall uncertainty and variability, and the thick error bars representing the 95% confidence
internval for the uncertainty in the population mean. The interpretation of these figures is that if
the thick error bars are much smaller (or greater) than the think error bars, then variability (or
uncertainty) contributes the most to overall uncertainty and variability.
For application to human health risk assessment, the uncertainty in and variability among
rodent internal dose estimates both contribute to uncertainty in human risk estimates. Therefore,
it is appropriate to combine uncertainty and variability when applying rodent dose metric
predictions to quantitative risk assessment. The median and 95% confidence interval for each
dose metric at some representative exposures in rodents are given in Tables 3.5.18-3.5.19, and
the confidence interval in these tables includes both uncertainty in the population mean and
variance as well as variability in the population. On the other hand, for use in predicting human
risk, it is often necessary to separate, to the extent possible, interindividual variability from
uncertainty, and this disaggregation is summarized in Table 3.5.20.
6/22/2009
175
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
2
3
4
5
6
7
Figure 3.5.3. PBPK model predictions for the fraction of intake that is metabolized under
continuous inhalation (A) and oral (B) exposure conditions in mice (white), rats (diagonal
hashing), and humans (horizontal hashing). Bars and thin error bars represent the median
estimate and 95% confidence interval for a random rodent group or human individual, and reflect
combined uncertainty and variability. Circles and thick error bars represent the median estimate
and 95% confidence interval for the population mean, and reflect uncertainty only.
Fraction Metabolized
~
Mouse
0
Rat
B
Human
B
Fraction Metabolized
~
Mouse
m
Rat
B
Human
Continuous inhalation ( ppm )
Continuous oral ( mg/kg-d )
6/22/2009
176
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Figure 3.5.4. PBPK model predictions for the fraction of intake that is metabolized by oxidation
2 (in the liver and lung) under continuous inhalation (A) and oral (B) exposure conditions in mice
3 (white), rats (diagonal hashing), and humans (horizontal hashing). Bars and thin error bars
4 represent the median estimate and 95% confidence interval for a random rodent group or human
5 individual, and reflect combined uncertainty and variability. Circles and thick error bars
6 represent the median estimate and 95% confidence interval for the population mean, and reflect
7 uncertainty only.
A Fraction Oxidized B
~ Mouse
0 Rat
B Human
Fraction Oxidized
~
Mouse
m
Rat
B
Human
Continuous inhalation ( ppm )
Continuous oral ( mg/kg-d )
6/22/2009
177
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Figure 3.5.5. PBPK model predictions for the fraction of intake that is metabolized by GSH
2 conjugation (in the liver and kidney) under continuous inhalation (A) and oral (B) exposure
3 conditions in mice (dotted line), rats (dashed line), and humans (solid line). X-values are slightly
4 offset for clarity. Crosses and thin error bars represent the median estimate and 95% confidence
5 interval for a random rodent group or human individual, and reflect combined uncertainty and
6 variability. Circles and thick error bars represent the median estimate and 95% confidence
7 interval for the population mean, and reflect uncertainty only.
r
Fraction Conjugated
IVRH IVRH IVRH IVRH IVRH
T
T
T
1
10 1 10 10 10
Continuous inhalation ( ppm )
Fraction Conjugated
IVRH IVRH IVRH IVRH IVRH
T
T
T
10 1 10 10
Continuous oral ( mg/kg-d )
10
6/22/2009
178
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Figure 3.5.6. PBPK model predictions for the weekly rate of bioactivation of DCVC in the
2 kidney per kg tissue weight per unit exposure (ppm or mg/kg-d) under continuous inhalation (A)
3 and oral (B) exposure conditions in rats (dashed line) and humans (solid line). X-values are
4 slightly offset for clarity. Crosses and thin error bars represent the median estimate and 95%
5 confidence interval for a random rodent group or human individual, and reflect combined
6 uncertainty and variability. Circles and thick error bars represent the median estimate and 95%
7 confidence interval for the population mean, and reflect uncertainty only.
O
E
Q_
Q_
CD
Q_
O)
Bioactivation in kidney per kg tissue
A per ppm
o -
RH
RH
I
RH
I
*
RH
I
RH
10 1 10 10 10
Continuous inhalation (ppm)
O)
O)
E
CD
Q_
"B)
E
o —
o -=
O -s
Bioactivation in kidney per kg tissue
3 per mg/kg-d
o —
RH RH RH RH RH
I 1 1 1 1
1(T1 1 101 102 103
Continuous oral (mg/kg-d)
6/22/2009
179
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Figure 3.5.7. PBPK model predictions for the weekly rate of oxidation of TCE in the respiratory
2 tract per kg tissue weight per unit exposure (ppm or mg/kg-d) under continuous inhalation (A)
3 and oral (B) exposure conditions in mice (dotted line), rats (dashed line), and humans (solid
4 line). X-values are slightly offset for clarity. Crosses and thin error bars represent the median
5 estimate and 95% confidence interval for a random rodent group or human individual, and reflect
6 combined uncertainty and variability. Circles and thick error bars represent the median estimate
7 and 95% confidence interval for the population mean, and reflect uncertainty only.
O s
O -=
O -=
E
Q_ co
Q. O -J
CD
Q- cnj
^ O
O)
Lung oxidation per kg tissue
A per ppm
o -=
MRH IVRH IVRH IVRH IVRH
I 1 1 1 1
1(T1 1 101 102 103
Continuous inhalation (ppm)
O)
O)
E
CD
Q_
"B)
E
o -=
o -=
O -s
O -s
o -=
O -s
o -=
Lung oxidation per kg tissue
3 per mg/kg-d
10 1 10 10 10
Continuous oral (mg/kg-d)
6/22/2009
180
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Figure 3.5.8. PBPK model predictions for the weekly rate of "untracked" oxidation of TCE in
2 the liver per kg tissue weight per unit exposure (ppm or mg/kg-d) under continuous inhalation
3 (A) and oral (B) exposure conditions in mice (dotted line), rats (dashed line), and humans (solid
4 line) X-values are slightly offset for clarity. Crosses and thin error bars represent the median
5 estimate and 95% confidence interval for a random rodent group or human individual, and reflect
6 combined uncertainty and variability. Circles and thick error bars represent the median estimate
7 and 95% confidence interval for the population mean, and reflect uncertainty only.
O
E
Q_
Q_
CD
Q_
O)
'Other' liver oxidation per kg liver
A per ppm
o -
MRH IVRH IVRH IVRH IVRH
I 1 1 1 1
1(T1 1 101 102 103
Continuous inhalation (ppm)
o —
O)
O)
E
CD
Q_
"B)
E
'Other' liver oxidation per kg liver
3 per mg/kg-d
o —
IVRH IVRH IVRH MRH IVRH
I 1 1 1 1
1(T1 1 101 102 103
Continuous oral (mg/kg-d)
6/22/2009
181
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Figure 3.5.9. PBPK model predictions for the weekly area-under-the-curve (AUC) of TCE in
2 venous blood (mg-h/l-wk) per unit exposure (ppm or mg/kg-d) under continuous inhalation (A)
3 and oral (B) exposure conditions in mice (dotted line), rats (dashed line), and humans (solid
4 line). X-values are slightly offset for clarity. Crosses and thin error bars represent the median
5 estimate and 95% confidence interval for a random rodent group or human individual, and reflect
6 combined uncertainty and variability. Circles and thick error bars represent the median estimate
7 and 95% confidence interval for the population mean, and reflect uncertainty only.
8
AUC TCE in blood AUC TCE in blood
per mg/kg-d
o
o
T3
E
0)
Q.
O)
E
o
IVRH IVRH IVRH MRH IVRH
o
•1
1
,2
,3
per ppm
o
o
E
Q_
Q_
CD
Q_
O)
E
o
MRH IVRH IVRH IVRH IVRH
o
•1
1
,2
,3
Continuous inhalation (ppm) Continuous oral (mg/kg-d)
9
6/22/2009
182
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Figure 3.5.10. PBPK model predictions for the weekly area-under-the-curve (AUC) of TCOH in
2 blood (mg-h/l-wk) per unit exposure (ppm or mg/kg-d) under continuous inhalation (A) and oral
3 (B) exposure conditions in mice (dotted line), rats (dashed line), and humans (solid line). X-
4 values are slightly offset for clarity. Crosses and thin error bars represent the median estimate
5 and 95% confidence interval for a random rodent group or human individual, and reflect
6 combined uncertainty and variability. Circles and thick error bars represent the median estimate
7 and 95% confidence interval for the population mean, and reflect uncertainty only.
8
E
Q_
Q_
CD
Q_
O)
E
AUC TCOH in blood
per ppm
o -=
o -
o —1
i
H S S""
l5
E
a3
Q.
O)
E
o -=
o -=
o —
MRH IVRH IVRH IVRH IVRH
i r~
T
~T
1
o —
o —1
10 1 10 10 10
Continuous inhalation (ppm)
AUC TCOH in blood
per mg/kg-d
—
H S[»-
IVRH
I—
IVRH
—I—
IVRH
—I—
MRH
—I—
IVRH
10 1 10 10 10
Continuous oral (mg/kg-d)
6/22/2009
183
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Figure 3.5.11. PBPK model predictions for the weekly area-under-the-curve (AUC) of TCA in
2 the liver (mg-h/l-wk) per unit exposure (ppm or mg/kg-d) under continuous inhalation (A) and
3 oral (B) exposure conditions in mice (dotted line), rats (dashed line), and humans (solid line). X-
4 values are slightly offset for clarity. Crosses and thin error bars represent the median estimate
5 and 95% confidence interval for a random rodent group or human individual, and reflect
6 combined uncertainty and variability. Circles and thick error bars represent the median estimate
7 and 95% confidence interval for the population mean, and reflect uncertainty only.
O
E
Q_
Q_
CD
Q_
O)
O)
E
o -
AUC TCA in liver
per ppm
o —
O)
O)
E
CD
Q_
2
O)
E
o —
MRH IVRH IVRH IVRH IVRH
I 1 1 1 1
1(T1 1 101 102 103
Continuous inhalation (ppm)
AUC TCA in liver
per mg/kg-d
IVRH IVRH IVRH MRH IVRH
I 1 1 1 1
1(T1 1 101 102 103
Continuous oral (mg/kg-d)
9
10
6/22/2009
184
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.5.18. Posterior predictions for representative internal doses: Mouse
2
Posterior Predictions for Mouse Dose Metrics: Median (2.5%, 97.5%)
Dose Metric 100 ppm, 7 hr/d, 5 d/wk 600 ppm, 7 hr/d, 5 d/wk 300 mg/kg-d, 5 d/wk 1000 mg/kg-d, 5 d/wk Units
ABioactDCVCKid
42 ( 0.0799 , 2020 )
323 ( 0.741 , 6060 )
100(0.26 , 3030)
434 ( 1.08 , 6800 )
mg/wk-kg tissue
AMetGSHBW34
0.707 (0.0322 , 16.1 )
5.39 ( 0.38 , 43.7 )
1.79 (0.0794,22.9)
6.55 ( 0.527 , 49.5 )
mg/wk-kg3'4
AMetLivl BW34
173 (60.8,395)
893 (342, 1960)
398(133,608)
880 (248, 1960)
mg/wk-kg3'4
AMetLivOtherLiv
203 ( 18.4, 2020)
1070 (99.6, 10500)
451 ( 46.4 , 4050 )
1040 (90.6, 10700)
mg/wk-kg tissue
AMetLngResp
651000 ( 24900 , 2540000 )
922000 (34800,8170000)
141000 ( 11300,512000)
441000 ( 27100 , 1620000 ) mg/wk-kg tissue
AUCCBId
97.5 (47.2,215)
823 (367,2010)
111 ( 7.32 , 426 )
616(55.8 , 1970)
mg-hr/l-wk
AUCCTCOH
98.8 ( 9.8 , 602 )
543 ( 51.9 , 4260 )
148 ( 18.8 , 670 )
427 (44.2,2410)
mg-hr/l-wk
AUCLivTCA
1890 ( 453 , 7270 )
5190 ( 1250, 19400)
2270 ( 497 , 8900 )
4650 (951 , 18700)
mg-hr/l-wk
TotMetabBW34
383 ( 146 , 928 )
1280 ( 456 , 3570 )
468 ( 183 , 616 )
1100 ( 324 , 2020 )
mg/wk-kg3'4
TotOxMetabBW34
380 ( 144 , 927 )
1270 ( 442 , 3560 )
463 ( 178 , 615 )
1090 (313 , 2010)
mg/wk-kg3'4
TotTCAInBW
270 ( 86 , 725 )
737(252,2110 )
322 ( 102 , 889 )
676 ( 179 , 1930)
mg/wk-kg
Note: Mouse body weight is assumed to be 0.03 kg. Predictions are weekly averages over 10 weeks of the specified exposure protocol. Confidence interval
reflects both uncertainties in population parameters (mean, variance) as well as population variability.
Table 3.5.19. Posterior predictions for representative internal doses: Rat
Posterior Predictions for Rat Dose Metrics: Median (2.5%,97.5%)
Dose Metric
100 ppm, 7 hr/d, 5 d/wk
600 ppm, 7 hr/d, 5 d/wk
300 mg/kg-d, 5 d/wk
1000 mg/kg-d, 5 d/wk
Units
ABioactDCVCKid
67.8 (6.03,513)
450 ( 35.4 , 4350 )
420 ( 31.6 , 3890 )
1720 ( 134, 15800)
mg/wk-kg tissue
AMetGSHBW34
0.331 (0.0626 , 2.16)
2.27(0.315, 19.3)
2.13 (0.293, 16)
8.84 ( 1.35 , 69.3 )
mg/wk-kg3'4
AMetLivl BW34
176 (81.1 , 344)
623 (271 ,1270)
539(176,1060)
951 ( 273 , 2780 )
mg/wk-kg3'4
AMetLivOtherLiv
1870(92.1 ,8670)
6660 (313,31200)
5490 ( 280 , 27400 )
9900 ( 492 , 59600 )
mg/wk-kg tissue
AMetLngResp
41900 ( 1460,496000)
67900 ( 2350 , 677000 )
40800 ( 1500 , 325000 )
85700 ( 2660 , 877000 )
mg/wk-kg tissue
AUCCBId
86.7 ( 39.2 , 242 )
1160 (349,2450)
670 (47.8, 1850)
3340 ( 828 , 8430 )
mg-hr/l-wk
AUCCTCOH
83.6( 1.94,1560)
446(6 , 10900 )
304 ( 4.71 , 7590 )
685 (8.14,32500)
mg-hr/l-wk
AUCLivTCA
587 ( 53.7 , 4740 )
2030 ( 186 , 13400)
1730(124,11800)
3130 ( 200 , 21000 )
mg-hr/l-wk
TotMetabBW34
206 ( 103 , 414)
682 (288, 1430)
572(199,1080)
1030 (302 , 2920)
mg/wk-kg3'4
TotOxMetabBW34
206 ( 103 , 414)
677 (285, 1430)
568 ( 191 , 1080)
1010 (286 , 2910)
mg/wk-kg3'4
TotTCAInBW
31.7(3.92 ,174)
110( 13.8,490)
90.1 (10.4,417)
164( 17.3,800)
mg/wk-kg
7 Note: Rat body weight is assumed to be 0.3 kg. Predictions are weekly averages over 10 weeks of the specified exposure protocol. Confidence interval reflects
8 both uncertainties in population parameters (mean, variance) as well as population variability.
6/22/2009
185
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 3.5.20. Posterior predictions for representative internal doses: Human
2
Dose Metric Posterior Predictions for Human Dose Metrics:
2.5% population: median (2.5%, 97.5%)
50% population: median (2.5%, 97.5%)
97.5% population: median (2.5%, 97.5%)
Female
0.001 ppm continuous
Male
0.001 ppm continuous
Female
0.001 mg/kg-d continuous
Male
0.001 mg/kg-d continuous
ABioactDCVCKid
0.02 ( 0.00549 , 0.0709 )
0.16 (0.0671 , 0.324)
0.95 ( 0.56 , 1.45 )
0.0207 ( 0.00558 , 0.0743 )
0.163 (0.0679,0.342)
0.979 (0.563 , 1.51 )
0.0152(0.0048 , 0.0384)
0.207 ( 0.0957 , 0.43 )
1.68 ( 0.956 , 2.26 )
0.016 (0.00493,0.0407)
0.22 (0.102,0.459)
1.81 ( 1.03 , 2.43 )
AMetGSHBW34
0.000159 ( 4.38e-05 , 0.000539 )
0.00126 ( 0.000536 , 0.00253 )
0.00736 ( 0.00442 , 0.011 )
0.000157 ( 4.37e-05 , 0.00054)
0.00125 ( 0.000528 , 0.00254 )
0.00736 (0.00434,0.0112)
0.000121 ( 3.82e-05 , 0.000316 )
0.00161 (0.000748,0.00331 )
0.013(0.00725,0.0164)
0.000123 ( 3.82e-05 , 0.000323 )
0.00167 ( 0.000777 , 0.00343 )
0.0136 (0.00759 , 0.0171 )
AMetLivl BW34
0.00161 (0.000619,0.00303)
0.00637 ( 0.00501 , 0.00799 )
0.0157 (0.0118 , 0.0206)
0.00157 ( 0.000608 , 0.00292 )
0.00619(0.00484,0.00779)
0.0152 (0.0115,0.02)
0.00465 (0.00169,0.0107)
0.0172 (0.0153,0.0183)
0.0192 (0.019,0.0193)
0.00498 (0.00184,0.0112)
0.018(0.0161 , 0.0191 )
0.02 (0.0198 , 0.0201 )
AMetLivOtherLiv
0.000748 ( 0.000138 , 0.00335 )
0.0104(0.00225 , 0.0237)
0.0805 (0.00871 ,0.147)
0.00065 ( 0.000119 , 0.00288 )
0.00898 (0.00193,0.0203)
0.0691 (0.00751 , 0.127)
0.00214 ( 0.000354 , 0.00979 )
0.0253 ( 0.00564 , 0.0543 )
0.157 (0.0188 , 0.251 )
0.00197 (0.00033,0.00907)
0.0234 ( 0.00526 , 0.0503 )
0.146(0.0173 , 0.232)
AMetLngResp
0.0144(0.00116 , 0.155)
2.44(0.613 , 6.71 )
25.8 ( 12.4,42.3)
0.0146 (0.00118,0.157)
2.44(0.621 , 6.65)
25.3(12.2,41.2)
0.00015 ( 1.27e-05, 0.00153)
0.0313 (0.00725,0.0963)
0.813(0.216,2.13)
0.000134 ( 1.15e-05, 0.00137)
0.0279 ( 0.00644 , 0.086 )
0.716(0.189, 1.9)
AUCCBId
0.00151 (0.00122,0.00186)
0.00285 (0.00252,0.00315)
0.00444 ( 0.00404 , 0.00496 )
0.00158 (0.00127,0.00191 )
0.00295 ( 0.00262 , 0.00326 )
0.00456(0.00416,0.00507)
4.33e-05 ( 3.3e-05 , 6.23e-05 )
0.000229 ( 0.000122 , 0.000436 )
0.00167 ( 0.000766 , 0.00324 )
3.84e-05 ( 2.89e-05 , 5.61 e-05 )
0.000204 ( 0.000109 , 0.000391 )
0.00153 ( 0.000693 , 0.00303 )
AUCCTCOH
0.00313(0.00135,0.00547)
0.0181 (0.0135 , 0.0241 )
0.082 (0.0586 ,0.118)
0.00305(0.00134,0.00532)
0.0179 (0.0133,0.0238)
0.0812 (0.0585,0.117)
0.00584(0.00205,0.0122)
0.0333 ( 0.025 , 0.0423 )
0.115(0.0872 , 0.163)
0.00615 (0.00213,0.0127)
0.035 ( 0.0264 , 0.0445 )
0.122 (0.0919,0.172)
AUCLivTCA
0.0152 (0.00668 , 0.0284)
0.126 (0.0784 , 0.194)
0.754 ( 0.441 , 1.38 )
0.0137 (0.00598,0.0258)
0.114(0.0704,0.177)
0.699 ( 0.408 , 1.3 )
0.029(0.0116 , 0.0524)
0.227(0.138,0.343)
1.11 (0.661 ,1.87)
0.0279 (0.0114,0.0501 )
0.219(0.133,0.33)
1.09 (0.64,1.88)
TotMetabBW34
0.0049 ( 0.00383 , 0.00595 )
0.0107 (0.00893,0.0129)
0.0246 (0.0185 , 0.0326)
0.00482 ( 0.0038 , 0.00585 )
0.0105 (0.00877,0.0127)
0.0244(0.0183,0.0324)
0.0163 (0.0136 , 0.0181 )
0.0191 (0.0188,0.0193)
0.0194(0.0194, 0.0194)
0.0173 (0.0147,0.019)
0.0199 (0.0196 , 0.0201 )
0.0202 ( 0.0202 , 0.0202 )
6/22/2009
186
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Dose Metric Posterior Predictions for Human Dose Metrics:
2.5% population: median (2.5%, 97.5%)
50% population: median (2.5%, 97.5%)
97.5% population: median (2.5%, 97.5%)
TotOxMetabBW34 0.00273 ( 0.00143 , 0.00422 )
0.00871 (0.0069 , 0.0111 )
0.0224(0.0158,0.0309)
0.00269(0.00143,0.00415)
0.00857 ( 0.00675 , 0.011 )
0.0222 (0.0155 , 0.0308)
0.0049 (0.00183 , 0.0108)
0.0173 (0.0154,0.0183)
0.0192 (0.019,0.0193)
0.00516 (0.00194,0.0114)
0.018(0.0161 ,0.0191 )
0.02 (0.0198 , 0.0201 )
TotTCAInBW 0.000259 ( 0.000121 , 0.000422 )
0.00154(0.00114,0.00202)
0.00525 ( 0.00399 , 0.00745 )
0.000246 ( 0.000114 , 0.000397 )
0.00146(0.00109,0.00193)
0.00499 ( 0.0038 , 0.0071 )
0.000501 ( 0.000189 , 0.000882 )
0.00286 ( 0.00222 , 0.00357 )
0.00659 ( 0.00579 , 0.00724 )
0.000506 ( 0.000192 , 0.00089 )
0.00289 ( 0.00222 , 0.0036 )
0.00662 ( 0.00581 , 0.00726 )
Note: Human body weight is assumed to be 70 kg for males, 60 kg for females. Predictions are weekly averages over 100 weeks of continuous exposure (dose
metric units same as previous tables). Each row represents a different population percentile (2.5%, 50%, 97.5%), and the confidence interval in each entry reflects
uncertainty in population parameters (mean, variance).
6/22/2009
187
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 3.5.21. Degree of variance in dose metric predictions due to incomplete convergence (columns 2-4), combined uncertainty and
population variability (columns 5-7), uncertainty in particular human population percentiles (columns 8-10), model fits to in vivo data
(column 11). The GSD is the geometric standard deviation, which is a "fold-change" from the central tendency.
Convergence: R for Generic GSD for Combined GSD for Uncertainty in human
Uncertainty and Variability population percentiles
Dose Metric
Abbreviation
Scenarios
Mouse Rat
Human
Mouse
Rat
Human 1%~5% 25%~75% 95%~99%
Comments regarding model fits to in vivo
data
ABioactDCVCKid
-
<1.016
<1.015
-
<3.92
<3.77
<2.08
<1.64
<1.30
Good fits to urinary NAcDCVC, and blood
DCVG.
AMetGSHBW34
<1.011
<1.024
<1.015
<9.09
<3.28
<3.73
<2.08
<1.64
<1.29
Good fits to urinary NAcDCVC, and blood
DCVG.
AMetLiv1BW34
<1.000
<1.003
<1.004
<2.02
<1.84
<1.97
<1.82
<1.16
<1.16
Good fits to oxidative metabolites.
AMetLivOtherLiv
<1.004
<1.151
<1.012
<3.65
<3.36
<3.97
<2.63
<1.92
<2.05
No direct in vivo data.
AMetLngResp
<1.001
<1.003
<1.002
<4.65
<4.91
<10.4
<4.02
<2.34
<1.83
No direct in vivo data, but good fits to closed
chamber.
AUCCBId
<1.001
<1.004
<1.005
<3.04
<3.16
<3.32
<1.20
<1.43
<1.49
Generally good fits, but poor fit to a few
mouse and human studies
AUCCTCOH
<1.001
<1.029
<1.002
<3.35
<8.78
<5.84
<1.73
<1.20
<1.23
Good fits across all three species.
AUCLivTCA
<1.000
<1.005
<1.002
<2.29
<3.18
<2.90
<1.65
<1.30
<1.40
Good fits to rodent data.
TotMetabBW34
<1.001
<1.004
<1.004
<1.92
<1.82
<1.81
<1.13
<1.12
<1.18
Good fits to closed chamber.
TotOxMetabBW34
<1.001
<1.003
<1.004
<1.94
<1.85
<1.96
<1.77
<1.15
<1.20
Good fits to closed chamber and oxidative
metabolites.
TotTCAInBW
<1.002
<1.002
<1.001
<1.96
<2.69
<2.30
<1.68
<1.19
<1.19
Good fits to TCA data.
6/22/2009
188
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
3.5.7.2 Implications for the Population Pharmacokinetics of TCE
3.5.7.2.1 Results
The overall uncertainty and variability in key toxicokinetic predictions, as a function of
dose and species, is shown in Figures 3.5.3-2.5.11. As expected, TCE that is inhaled or ingested
is substantially metabolized in all species, predominantly by oxidation (Figures 3.5.3-3.5.4). At
higher exposures, metabolism becomes saturated and the fraction metabolized declines. Mice on
average have a greater capacity to oxidized TCE than rats or humans, and this is reflected in the
predictions at the two highest levels for each route. The uncertainty in the predictions for the
population means for total and oxidative metabolism is relatively modest, therefore the wide
confidence interval for combined uncertainty and variability largely reflects inter-group (for
rodents) or inter-individual (for humans) variability. Of particular note is the high variability in
oxidative metabolism at low doses in humans, with the 95% confidence interval spanning from
0.1-0.7 for inhalation and 0.2-1.0 for ingestion.
Predictions of GSH conjugation and renal bioactivation of DCVC are highly uncertain in
rodents, spanning more than 1000-fold in mice and 100-fold in rats (Figure 3.5.5-3.5.6). In both
mice and rats, the uncertainty in the population mean virtually overlaps with the combined
uncertainty and variability, reflecting the lack of GSH-conjugate specific data in mice (the
bounds are based on mass balance) and the availability of only urinary NAcDCVC excretion in
one study in rats. In humans, however, the blood concentrations of DCVG from Lash et al.
(1999b) combined with the urinary NAcDCVC data from Bernauer et al. (1996) were able to
better constrain GSH conjugation and bioactivation of DCVC, with 95% confidence intervals on
the population mean spanning only 3-fold or so. However, substantial variability is predicted
(reflecting variability in the measurements of Lash et al. 1999b), for the error bars for the
population mean are substantially smaller than that for overall uncertainty and variability. Of
particular interest is the prediction of one or two orders of magnitude more GSH conjugation and
DCVC bioactivation, on average, in humans than in rodents. Furthermore, although the 95%
confidence intervals for the overall uncertainty and variability overlap, the 95% confidence
intervals of the predicted population means between humans and rats do not overlap. Therefore,
the model predicts significantly greater GSH conjugation and DCVC bioactivation in humans
relative to rats, although the difference in predicted population means, based on the 95%
confidence bounds, may range from as little as 2-fold to as much as 1000-fold.
Predictions for respiratory tract oxidative metabolism were, as expected, greatest in mice,
followed by rats and then humans (Figure 3.5.7). In addition, due to the "pre-systemic" nature of
the respiratory tract metabolism model as well as the hepatic first-pass effect, substantially more
6/22/2009
189
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
metabolism was predicted from inhalation exposures as compared to oral exposures.
Interestingly, the population means appeared to be fairly well constrained despite the lack of
direct data, suggesting that overall mass balance is an important constraint for the pre-systemic
respiratory tract metabolism modeled here.
Some constraints were also placed on "other" hepatic oxidation - i.e., through a pathway
that does not result in chloral formation and subsequent formation of TCA and TCOH (Figure
3.5.8). The 95% confidence interval for overall uncertainty and variability spanned about 100-
fold, a large fraction of that due to uncertainty in the population mean. Interestingly, a higher
rate per kg tissue was predicted for rats than for mice or humans, although importantly, the 95%
confidence intervals for the population means overlap among all three species.
The area-under-the-curve (AUC) of TCE in blood (Figure 3.5.9) showed the expected
non-linear behavior with increasing dose, with the non-linearity was more pronounced with oral
exposure, as would be expected by hepatic first-pass. Interestingly, the AUC of TCOH in blood
(Figure 3.5.10) was relatively constant with dose, reflecting the parallel saturation of both TCE
oxidation and TCOH glucuronidation. In fact, in humans, the mean AUC for TCOH in blood
increases up to 100 ppm or 100 mg/kg-d, due to saturation of TCOH glucuronidation, before
decreasing at 1000 ppm or 1000 mg/kd-d, due to saturation of TCE oxidation.
The predictions for the AUC for TCA in the liver showed some interesting features
(Figure 3.5.11). The predictions for all three species with within an order of magnitude of each
other, with a relatively modest uncertainty in the population mean (reflecting the substantial
amount of data on TCA). The shape of the curves, however, differs substantially, with humans
showing saturation at much lower doses than rodents. In fact, the ratio between the liver TCA
AUC and the rate of TCA production, though it differs between species, is relatively constant as
a function of dose within species (not shown). Therefore, the shape of the curves largely reflect
saturation in the production of TCA from TCOH, not in the oxidation of TCE itself, for which
saturation is predicted at higher doses, particularly via the oral route (Figure 3.5.4). In addition,
while for the same exposure (ppm or mg/kg-d TCE) more TCA (on a mg/kg-d basis) is produced
in mice relative to rats and humans (not shown), humans and rats have longer TCA half-lives
even though plasma protein binding of TCA is on average greater.
3.5.7.2.2 Discussion
This analysis substantially informs four of the major areas of pharmacokinetic
uncertainty previously identified in numerous reports (reviewed in Chiu et al. 2006): GSH
conjugation pathway, respiratory tract metabolism, alternative pathways of TCE oxidation
including DCA formation, and the impact of plasma binding on TCA kinetics particularly in the
liver.
6/22/2009
190
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
With respect to the first, GSH conjugation and subsequent bioactivation of DCVC in
humans appears substantially greater than previously estimated based on urinary excretion data
alone (Bernauer et al. 1996; Birner et al. 1993). This result is supported by in vitro data, noted in
Chiu et al. (2006), reporting the formation rate of DCVG from TCE in freshly isolated
hepatocytes was similar in order of magnitude to that measured for oxidative metabolites
(Lipscomb et al., 1998; Lash et al., 1999a). Such in vitro data on GSH conjugation were used for
developing prior distributions for GSH conjugation rates in the PBPK model reported here, but
were not used in previous PBPK models for TCE. This conclusion is also a result of the
incorporation in the analysis of DCVG blood data reported by Lash et al. (1999b) after controlled
TCE inhalation exposures (which was not included in previous PBPK-based analyses) and
urinary NAcDCVC excretion data from Bernauer et al. (1996). Indeed, as discussed in Section
3.3, DCVG blood levels in the Lash et al. (1999b) study were comparable on a molar basis to
TCOH blood levels, suggesting substantial GSH conjugation in humans independent of any
PBPK model. In particular, the reported mean peak blood DCVG concentrations of 46 [xM in
males exposed to 100 ppm TCE for 4 hrs (Lash et al. 1999b), multiplied by a typical blood
volume of 5 1 (ICRP 2002), yields a peak amount of DCVG in blood of 0.23 mmoles. In
comparison, the retained dose from 100 ppm exposure for 4 hr is 4.4 mM, assuming retention of
about 50% (Monster et al. 1976) and minute-volume of 9 1/min (ICRP 2002). Thus, in these
subjects, about 5% of the retained dose is present in blood as DCVG at the time of peak blood
concentration. This is a strong lower bound on the total fraction of retained TCE undergoing
GSH conjugation because DCVG clearance is ongoing at the time of peak concentration, and
DCVG may be distributed to tissues other than blood. It should be reiterated that only grouped
DCVG blood data were available for PBPK model-based analysis; however, this should only
result in an underestimation of the degree of variation in GSH conjugation. Finally, replication
or corroboration of the findings of Lash et al. (1999b) in future studies would further increase
confidence in the predictions.
Several other aspects of the predictions related to GSH conjugation of TCE are worthy of
discussion. Predictions for rats and mice remain more uncertain due to their having less
toxicokinetic data, but are more highly constrained by total recovery studies. It is also notable
that the extent of total recovery in human studies (60-70%, as reviewed in Chiu et al. 2007) is
substantially less than in rodent studies (upwards of 90%), consistent with a greater role for GSH
conjugation in humans. In addition, it has been suggested that "saturation" of the oxidative
pathway for volatiles may lead to marked increases in flux through the GSH conjugation
pathway (Slikker et al., 2004a,b), but the PBPK model predicts only a modest, at most ~2-fold,
change in flux, because there is evidence that both pathways are saturable for this substrate at
6/22/2009
191
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
similar exposures. Therefore, any substantial non-linearities in toxicity cannot be explained
solely by metabolic saturation of the oxidative pathway.
With respect to the other areas of uncertainty, consistent with the qualitative suggestions
from in vitro data, the analysis here predicts that mice have greater rate of respiratory tract
oxidative metabolism as compared to rats and humans. However, the predicted difference of 50-
fold or so on average was not as great as the 600-fold suggested by previous reports (Green et al.
1997, Green 2000, NRC 2006). In addition, available data are consistent with a wide range of
variability in respiratory tract metabolism, particularly in humans, likely due inter-individual
variability observed in blood TCE levels after inhalation exposure. With respect to "untracked"
oxidative metabolism, this pathway appears to be a relatively small contribution to total
oxidative metabolism. While it is temping to use this pathway as a surrogate for DCA
production through from the TCE epoxide (Cai and Guengerich, 1999), one should be reminded
that DCA may be formed through multiple pathways (see Section 3.3). Therefore, this pathway
at best represents a lower bound on DCA production. Finally, with respect to TCA dosimetry,
this analysis predicts that inter-species differences in liver TCA AUC are modest, with a range of
10-fold or so across species, due to the combined effects of inter-species differences in the yield
of TCA from TCE, plasma protein binding, and elimination half-life.
In a situation such as TCE in which there is large database of studies coupled with
complex toxicokinetics, the Bayesian approach provides a systematic method of simultaneously
estimating model parameters and characterizing their uncertainty and variability. While such an
approach is not necessarily needed for all applications, such as route-to-route extrapolation (Chiu
and White 2006), as discussed in Barton et al. (2007), characterization of uncertainty and
variability is increasingly recognized as important for risk assessment while representing a
continuing challenge for both PBPK modelers and users. An endeavor such as that reported here
is clearly not trivial, as evidenced by the evolution of the methodology from Bois (2000a,
2000b), to Hack et al. (2006), to the present analysis.
Part of this evolution has been a more refined specification of the problem being
addressed, in particular the precise hierarchical population model for each species so that
relevant data can be selected for analysis (e.g., excluding most grouped human data in favor of
individual human data) and data can be appropriately grouped (e.g., in rodent data, grouping by
sex and strain within a particular study). Thus, the predictions from the population model in
rodents are the "average" for a particular "lot" of rodents of a particular species, strain, and sex.
This is in contrast to the Hack et al. (2006) model, in which each dose group was treated as a
separate "individual." As discussed above, this previous population model structure led to the
unlikely result that different dose groups within a closed chamber study had significantly
different Vmax values. In humans, however, inter-individual variability is of interest, and
6/22/2009
192
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
furthermore, substantial individual data are available in humans. Hack et al. (2006) mixed
individual- and group-level data, depending on the availability from the published study, but this
approach likely underestimates population variability due to group means being treated as
individuals. In addition, in some studies, the same individual was exposed more than once, and
in Hack et al. (2006), these were treated as different "individuals." In this case, actual inter-
individual variability may be either over- or underestimated, depending on the degree of inter-
occasion variability. While it is technically feasible to include inter-occasion variabilty, it would
have added substantially to the computational burden and reduced parameter identifiability. In
addition, the primary interest for risk assessment is chronic exposure, so the predictions from the
population model in humans are the "average" across different occasions for a particular
individual (adult).
The present analysis to be maximally objective and transparent in the sense that available
information, or the lack thereof, is formally codified and explicit either in prior distributions or in
the data used to generate posterior distributions (and not both). Specific innovations aimed at
minimizing subjectivity (and hence improving reproducibility) in parameter estimation include:
(i) clear separation between the in vitro or physiologic data used to develop prior distributions
and the in vivo data used to generate posterior distributions; (ii) use of non-informative
distributions, first updated using a probabilistic model of interspecies-scaling that allows for
prediction error, for parameters lacking in prior information; (iii) use of a more comprehensive
database of physiologic data, in vitro measurements, and in vivo data for parameter calibration or
for out-of-sample evaluation ("validation"). Because of these measures, we feel confident that
the approach employed also yields an accurate characterization of the uncertainty in metabolic
pathways for which available data was sparse or relatively indirect, such as GSH conjugation in
rodents and respiratory tract metabolism.
This analysis has a number of limitations and opportunities for refinement. One would
be the inclusion of a CH sub-model, so that pharmacokinetic data, such as that recently published
by Merdink et al. (2008), could be incorporated. In addition, our probabilistic analysis is still
dependent on a model structure substantially informed by deterministic analyses that test
alternative model structures (Evans et al., 2009), as probabilistic methods for discrimination or
selection among complex, non-linear models such as that for TCE have not yet been widely
accepted. Therefore, additional refinement of the respiratory tract model may be possible,
though the lack of more direct in vivo data would likely preclude one from strongly
discriminating between models. Furthermore, additional model changes that may be of utility to
risk assessment, such as development of models for different lifestages (including childhood and
pregnancy), would likely require additional in vivo or in vitro data to ensure model
identifiability. Finally, improvements are possible in the statistical and population models and
6/22/2009
193
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
analyses, such as incorporation of inter-occasion variability (Bernillon and Bois 2000),
application of more sophisticated "validation" methods (such as cross-validation), and more
rigorous treatment of grouped data (Chiu and Bois 2007).
3.5.7.3 Overall evaluation of PBPK model-based internal dose predictions
The utility of the PBPK model developed here for making predictions of internal dose
can be evaluated based on four different components: (i) the degree to which the simulations
have converged to the true posterior distribution; (ii) the degree of overall uncertainty and
variability; (iii) for humans, the degree of uncertainty in the population; and (iv) the degree to
which the model predictions are consistent with in vivo data that are informative to a particular
dose metric. Table 3.5.21 summarizes these considerations for each dose metric prediction.
Note that this evaluation does not consider in any way the extent to which a dose metric may be
the appropriate choice for a particular toxic endpoint.
Overall, the least uncertain dose metrics are the fluxes of total metabolism
(TotMetabBW34), total oxidative metabolism (TotOxMetabBW34), and hepatic oxidation
(AMetLivlBW34). These all have excellent posterior convergence (R diagnostic < 1.01),
relatively low uncertainty and variability (GSD < 2), and relatively low uncertainty in human
population variability (GSD for population percentiles < 2). In addition, the PBPK model
predictions compare well with the available in vivo pharmacokinetic data.
Predictions for TCE in blood (AUCCBld) are somewhat more uncertain. Although
convergence was excellent across species (R < 1.01), overall uncertainty and variability was
about 3-fold. In humans, the uncertainty in human population variability was relatively low
(GSD for population percentiles < 1.5). TCE blood level predictions were somewhat high in
comparison to the Chiu et al. (2006) study at 1 ppm, though the predictions were better for most
of the other studies at higher exposure levels. In mice, TCE blood levels were somewhat over-
predicted in open-chamber inhalation studies. In both mice and rats, there were some cases in
which fits were inconsistent across dose groups if the same parameters were used across dose
groups, indicating unaccounted-for dose-related effects or intra-study variability. However, in
both rats and humans, TCE blood (humans and rats) and tissue (rats only) concentrations from
studies not used for calibration (i.e., saved for "out-of-sample" evaluation/"validation") were
well simulated, adding confidence to the parent compound dose metric predictions.
For the TCA dose metric predictions (TotTCAInBW, AUCLivTCA) convergence in all
three species was excellent (R < 1.01). Overall uncertainty and variability was intermediate
between dose metrics for metabolism and that for TCE in blood, with GSD of about 2 to 3-fold.
Uncertainty in human population percentiles was relatively low (GSD of 1.2 to 1.7). While liver
TCA levels were generally well fit, the data was relatively sparse. Plasma and blood TCA levels
6/22/2009
194
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
were generally well fit, though in mice, there were again some cases in which fits were
inconsistent across dose groups if the same parameters were used across dose groups, indicating
unaccounted-for dose-related effects or intra-study variability. In humans, the accurate
predictions for TCA blood and urine concentrations from studies used for "out of sample"
evaluation lends further confidence to dose metrics involving TCA.
The evaluation of TCOH in blood followed a similar pattern. Convergence in all three
species was good, though the rat model had slightly worse convergence (R ~ 1.03) than the
mouse and humans (R < 1.01). In mice, overall uncertainty and variability was slightly more
than for TCE in blood. There much higher overall uncertainty and variability in the rat
predictions (GSD of almost 9) that likely reflects true inter-study variability. The
population-generated predictions for TCOH and TCOG in blood and urine were quite wide, with
some in vivo data both at the upper and lower ends of the range of predictions. In humans, the
overall uncertainty and variability was intermediate between mice and rats (GSD = 5.8). As with
the rats, this likely reflects true population heterogeneity, as the uncertainty in human population
percentiles was relatively low (GSD of around 1.2-1.7-fold). For all three species, fits to in vivo
data are generally good. In mice, however, there were again some cases in which fits were
inconsistent across dose groups if the same parameters were used across dose groups, indicating
unaccounted-for dose-related effects or intra-study variability. In humans, the accurate
predictions for TCOH blood and urine concentrations from studies used for "out of sample"
evaluation lends further confidence to those dose metrics involving TCOH.
GSH metabolism dose metrics (ABioactDCVCKid, AMetGSHBW34) had the greatest
overall uncertainty in mice but was fairly well characterized in rats and humans. In mice, there
was no in vivo data informing this pathway except for the indirect constraint of overall mass
balance. So although convergence was adequate (R < 1.02), the uncertainty/variability was very
large, with a GSD of 9-fold for the overall flux (the amount of bioactivation was not
characterized because there are no data constraining downstream GSH pathways). For rats, there
were additional constraints from (well-fit) urinary NAcDCVC data, which reduced the overall
uncertainty and variability substantially (GSD < 4-fold). In humans, in addition to urinary
NAcDCVC data, DCVG blood concentration data was available, though only at the group level.
However, these data, both of which were well fit, in addition to the greater amount of in vitro
metabolism data, allowed for the flux through the GSH pathway and the rate of DCVC
bioactivation to be fairly well constrained, with overall uncertainty and variability having GSD <
4-fold, and uncertainty in population percentiles no more than about 2-fold.
The final two dose metrics, respiratory metabolism (AMetLngResp) and "other"
oxidative metabolism (AMetLivOtherLiv), also lacked direct in vivo data and were predicted
largely on the basis of mass balance and physiological constraints. Respiratory metabolism had
6/22/2009
195
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
good convergence (R < 1.01), helped by the availability of closed chamber data in rodents. In
rats and mice, overall uncertainty and variability was rather uncertain (GSD of 4~5-fold), but the
overall uncertainty and variability was much greater in humans, with a GSD of about 10-fold.
This largely reflects the significant variability across individuals as well as substantial
uncertainty in the low population percentiles (GSD of 4-fold). However, the middle (i.e.,
"typical" individuals) and upper percentiles (i.e., the individuals at highest risk) are fairly well
constrained with a GSD of around 2-fold. For the "other" oxidative metabolism dose metric,
convergence was good in mice and humans (R < 1.02), but less than ideal in rats (R~l. 15). In
rodents, the overall uncertainty and variability were moderate, with a GSD around 3.5-fold,
slightly higher than that for TCE in blood. The overall uncertainty and variability in this metric
in humans had a GSD of about 4-fold, slightly higher than for GSH conjugation metrics.
However, uncertainty in the middle and upper population percentiles had GSDs of only about 2-
fold, similar to that for respiratory metabolism.
Overall, as shown in Table 3.5.21, the updated PBPK model appears to be most reliable
for the fluxes of total, oxidative, and hepatic oxidative metabolism. In addition, dose metrics
related to blood levels of TCE and oxidative metabolites TCOH and TCA had only modest
uncertainty. In the case of TCE in blood, for some datasets, model predictions overpredicted the
in vivo data, and, in the case of TCOH in rats, substantial interstudy variability was evident. For
GSH metabolism, dose-metric predictions for rats and humans had only slightly greater
uncertainty than the TCE and metabolism metrics. Predictions for mice were much more
uncertain, reflecting the lack of GSD-specific in vivo data. Finally, for "other" oxidative
metabolism and respiratory oxidative metabolism, predictions also had somewhat more
uncertainty than the TCE and metabolism metrics, though uncertainty in middle and upper
human population percentiles was modest.
6/22/2009
196
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
REFERENCES
Abbas RR, Seckel CS, Kidney JK, Fisher JW.(1996). Pharmacokinetic analysis of chloral
hydrate and its metabolism in B6C3F1 mice. DrugMetab Dispos. 24:1340-6.
Abbas R, Fisher JW. (1997). physiologically based pharmacokinetic model for trichloroethylene
and its metabolites, chloral hydrate, trichloroacetate, dichloroacetate, trichloroethanol,
and trichloroethanol glucuronide inB6C3Fl mice. Toxicol ApplPharmacol. 147:15-30.
Abbas R, Seckel CS, MacMahon KL, Fisher JW. (1997). Determination of kinetic rate constants
for chloral hydrate, trichloroethanol, trichloroacetic acid, and dichloroacetic acid—a
physiologically based modeling approach. Toxicologist 36:32-33.
Albanese RA, Banks HT, Evans MV, Potter LK. (2002). Physiologically based pharmacokinetic
models for the transport of trichloroethylene in adipose tissue. Bull Math Biol 64:97-131.
Allen BC, Fisher JW. (1993). Pharmacokinetic modeling of trichloroethylene and trichloroacetic
acid in humans. Risk Anal. 13(l):71-86.
Andersen, M. E., Gargas, M. L., Clewell, H. J., 3rd, and Severyn, K. M. (1987). Quantitative
evaluation of the metabolic interactions between trichloroethylene and 1,1-
dichloroethylene in vivo using gas uptake methods. Toxicol Appl Pharmacol 89, 149-157.
Barton, H. A., Creech, J. R., Godin, C. S., Randall, G. M., and Seckel, C. S. (1995).
Chloroethylene mixtures: pharmacokinetic modeling and in vitro metabolism of vinyl
chloride, trichloroethylene, and trans-1,2-dichloroethylene in rat. Toxicol Appl
Pharmacol 130, 237-247.
Barton HA, Bull R, Schultz I, Andersen ME. (1999). Dichloroacetate (DCA) dosimetry:
interpreting DCA-induced liver cancer dose response and the potential for DCA to
contribute to trichloroethylene-induced liver cancer. Toxicol Lett. 106:9-21.
Barton HA, Chiu WA, Woodrow Setzer R, Andersen ME, Bailer AJ, Bois FY, Dewoskin RS,
Hays S, Johanson G, Jones N, Loizou G, Macphail RC, Portier CJ, Spendiff M, Tan YM.
(2007). Characterizing uncertainty and variability in physiologically based
pharmacokinetic models: state of the science and needs for research and implementation.
Toxicol Sci. 99:395-402.
Bartonicek, V. (1962). Metabolism and excretion of trichloroethylene after inhalation by human
subjects. Br JIndustMed. 19:134-141.
Bernillon P, Bois FY. (2000). Statistical issues in toxicokinetic modeling: a Bayesian
perspective. Environ. Health Perspect. 108(Suppl. 5): 883-893
Bernauer, U., Birner, G., Dekant, W., and Henschler, D. (1996). Biotransformation of
trichloroethene: dose-dependent excretion of 2,2,2-trichloro-metabolites and mercapturic
acids in rats and humans after inhalation. Arch Toxicol 70, 338-346.
6/22/2009
197
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Birner G, Vamvakas S, Dekant W, Henschler D. (1993). Nephrotoxic and genotoxic N-acetyl-S-
dichlorovinyl-L-cysteine is a urinary metabolite after occupational 1,1,2-trichloroethene
exposure in humans: implications for the risk of trichloroethene exposure. Environ
Health Perspect. 99:281-4.
Bloemen, L. J., Monster, A. C., Kezic, S., Commandeur, J. N., Veulemans, H., Vermeulen, N. P.,
and Wilmer, J. W. (2001). Study on the cytochrome P-450- and glutathione-dependent
biotransformation of trichloroethylene in humans. IntArch Occup Environ Health 74,
102-108.
Bois FY. (2000a). Statistical analysis of Fisher et al. PBPK model of trichloroethylene kinetics.
Environ Health Perspect. 108 Suppl 2:275-82.
Bois FY. (2000b). Statistical analysis of Clewell et al. PBPK model of trichloroethylene kinetics.
Environ Health Perspect. 108 Suppl 2:307-16.
Brown RP, Delp MD, Lindstedt SL, Rhomberg LR, Beliles RP. (1997). Physiological parameter
values for physiologically based pharmacokinetic models. ToxicolIndHealth. 13:407-84.
Cai H, Guengerich P. (1999). Mechanism of aqueous decomposition of trichloroethylene oxide. J
AmChemSoc. 121:11656-11663.
Chiu WA, White P. (2006). Steady-state solutions to PBPK models and their applications to risk
assessment I: Route-to-route extrapolation of volatile chemicals. Risk Anal. 26:769-80.
Chiu WA, Okino MS, Lipscomb JC, Evans MV. (2006). Issues in the pharmacokinetics of
trichloroethylene and its metabolites. Environ Health Perspect. 114:1450-6.
Chiu WA, Bois FY. (2007). An approximate method for population toxicokinetic analysis with
aggregated data. Journal of Agricultural, Biological, and Environmental Statistics.
12:346-363.
Chiu, W. A., Micallef, S., Monster, A. C., and Bois, F. Y. (2007). Toxicokinetics of inhaled
trichloroethylene and tetrachloroethylene in humans at 1 ppm: empirical results and
comparisons with previous studies. Toxicol Sci 95, 23-36.
Clewell HJ 3rd, Gentry PR, Covington TR, Gearhart JM. (2000). Development of a
physiologically based pharmacokinetic model of trichloroethylene and its metabolites for
use in risk assessment. Environ Health Perspect. 108 Suppl 2:283-305.
D'Souza RW, Bruckner JV, Feldman S. (1985). Oral and intravenous trichloroethylene
pharmacokinetics in the rat. J Toxicol Environ Health. 15:587-601.
Dallas, C. E., Gallo, J. M., Ramanathan, R., Muralidhara, S., and Bruckner, J. V. (1991).
Physiological pharmacokinetic modeling of inhaled trichloroethylene in rats. Toxicol
ApplPharmacol 110, 303-314.
6/22/2009
198
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
David RM, Clewell HJ, Gentry PR, Covington TR, Morgott DA, Marino DJ. (2006). Revised
assessment of cancer risk to dichloromethane II. Application of probabilistic methods to
cancer risk determinations. Regul Toxicol Pharmacol. 45:55-65.
Ertle T, Henschler D, Miiller G, Spassowski M. (1972). Metabolism of trichloroethylene in man.
I. The significance of trichloroethanol in long-term exposure conditions. Arch Toxikol.
29:171-88.
Evans, MV., Chiu, WA., Okino, MS., and Caldwell, JC. Development of an updated PBPK
model for trichloroethylene and metabolites in mice, and its application to discern the
role of oxidative metabolism in TCE-induced hepatomegaly, Toxicol. Appl. Pharmacol.,
236: 329-340.
Fernandez, J. G., Droz, P. O., Humbert, B. E., and Caperos, J. R. (1977). Trichloroethylene
exposure. Simulation of uptake, excretion, and metabolism using a mathematical model.
Br JIndMed 34, 43-55.
Fisher JW. (2000). Physiologically based pharmacokinetic models for trichloroethylene and its
oxidative metabolites. Environ Health Perspect. 108 Suppl 2:265-73.
Fisher JW, Allen BC. (1993). Evaluating the risk of liver cancer in humans exposed to
trichloroethylene using physiological models. Risk Anal. 13(l):87-95.
Fisher, J. W., Whittaker, T. A., Taylor, D. H., Clewell, H. J., 3rd, and Andersen, M. E. (1989).
Physiologically based pharmacokinetic modeling of the pregnant rat: a multiroute
exposure model for trichloroethylene and its metabolite, trichloroacetic acid. Toxicol
Appl Pharmacol 99, 395-414.
Fisher JW, Gargas ML, Allen BC, Andersen ME. (1991). Physiologically based pharmacokinetic
modeling with trichloroethylene and its metabolite, trichloroacetic acid, in the rat and
mouse. Toxicol Appl Pharmacol. 109:183-95.
Fisher, J. W., Mahle, D., and Abbas, R. (1998). A human physiologically based pharmacokinetic
model for trichloroethylene and its metabolites, trichloroacetic acid and free
trichloroethanol. Toxicol Appl Pharmacol 152, 339-359.
Gelman A, Bois F, Jiang J. (1996). Physiological pharmacokinetic analysis using population
modeling and infromatibe prior distributions. J Am Stat Assoc. 91:1400-1412.
Gelman A, Carlin JB, Stern HS, Rubin DB. (2004). Bayesian Data Analysis. New York:
Chapman & Hall/CRC.
Green, T., and Prout, M. S. (1985). Species differences in response to trichloroethylene. II.
Biotransformation in rats and mice. Toxicol Appl Pharmacol 79, 401-411.
Green, T., Mainwaring, G. W., and Foster, J. R. (1997). Trichloroethylene-induced mouse lung
tumors: studies of the mode of action and comparisons between species. Fundam Appl
Toxicol 37, 125-130.
6/22/2009
199
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Green T. (2000). Pulmonary toxicity and carcinogenicity of trichloroethylene: species
differences and modes of action. Environ Health Perspect. 108 Suppl 2:261-4.
Greenberg, M. S., Burton, G. A., and Fisher, J. W. (1999). Physiologically based
pharmacokinetic modeling of inhaled trichloroethylene and its oxidative metabolites in
B6C3F1 mice. Toxicol ApplPharmacol 154, 264-278.
Hack CE, Chiu WA, Jay Zhao Q, Clewell HJ.(2006). Bayesian population analysis of a
harmonized physiologically based pharmacokinetic model of trichloroethylene and its
metabolites. Regul Toxicol Pharmacol. 46:63-83.
Hissink, E. M., Bogaards, J. J. P., Freidg, A. P., Commandeur, J. N. M., Vermeulen, N. P. E.,
and van Bladeren, P. J. (2002). The use of in vitro metabolic parameters and
physiologically based pharmacokinetic (PBPK) modeling to explore the risk assessment
of trichloroethylene. Environmental Toxicology and Pharmacology 11, 259-271.
ICRP. International Commission on Radiological Protection. (2002). Basic anatomical and
physiological data for use in radiological protection: reference values. A report of age-
and gender-related differences in the anatomical and physiological characteristics of
reference individuals. ICRP Publication 89. Ann ICRP.32:5-265.
Jakobson, I., Holmberg, B., and Ekner, A. (1986). Venous blood levels of inhaled
trichloroethylene in female rats and changes induced by interacting agents. Acta
Pharmacol Toxicol (Copenh) 59, 135-143.
Kaneko, T., Wang, P. Y., and Sato, A. (1994). Enzymes induced by ethanol differently affect the
pharmacokinetics of trichloroethylene and 1,1,1-trichloroethane. Occup Environ Med 51,
113-119.
Keys, D. A., Bruckner, J. V., Muralidhara, S., and Fisher, J. W. (2003). Tissue dosimetry
expansion and cross-validation of rat and mouse physiologically based pharmacokinetic
models for trichloroethylene. Toxicol Sci 76, 35-50.
Keys DA, Schultz IR, Mahle DA, Fisher JW. (2004). A quantitative description of suicide
inhibition of dichloroacetic acid in rats and mice. Toxicol Sci 82:381-393.
Kimmerle, G., Eben, A. (1973a). Metabolism, excretion and toxicology of trichloroethylene after
inhalation. I. Experimental exposure on rats. Arch Toxikol 30:115-126.
Kimmerle, G., Eben, A. (1973b). Metabolism, excretion and toxicology of trichloroethylene
after inhalation. II. Experimental human exposure. Arch Toxikol 30:127-138.
Lapare S, Tardif R, Brodeur J.(1995). Effect of various exposure scenarios on the biological
monitoring of organicsolvents in alveolar air. II. 1,1,1-Trichloroethane and
trichloroethylene. Int Arch Occup Environ Health. 67:375-94.
6/22/2009 200
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Larson, J. L., and Bull, R. J. (1992a). Metabolism and lipoperoxidative activity of
trichloroacetate and dichloroacetate in rats and mice. Toxicol Appl Pharmacol 115, 268-
277.
Larson, J. L., and Bull, R. J. (1992b). Species differences in the metabolism of trichloroethylene
to the carcinogenic metabolites trichloroacetate and dichloroacetate. Toxicol Appl
Pharmacol 115, 278-285.
Lash, L. H., Visarius, T. M., Sail, J. M., Qian, W., and Tokarz, J. J. (1998). Cellular and
subcellular heterogeneity of glutathione metabolism and transport in rat kidney cells.
Toxicology 130, 1-15.
Lash LH, Lipscomb JC, PUtt DC, Parker JC. (1999a). Glutathione conjugation of
trichloroethylene in human liver and kidney: kinetics and individual variation. Drug
Metab Dispos. 27:351-359.
Lash LH, Putt DA, Brashear WT, Abbas R, Parker JC, Fisher JW. (1999b). Identification of S-
(l,2-dichlorovinyl)glutathione in the blood of human volunteers exposed to
trichloroethylene. J Toxicol Environ Health A. 56:1 -21.
Lash LH, Putt DA, Parker JC. (2006). Metabolism and tissue distribution of orally administered
trichloroethylene in male and female rats: identification of glutathione- and cytochrome
P-450-derived metabolites in liver, kidney, blood, and urine. J Toxicol Environ Health A.
69:1285-309.
Lee, K. M., Bruckner, J. V., Muralidhara, S., and Gallo, J. M. (1996). Characterization of
presystemic elimination of trichloroethylene and its nonlinear kinetics in rats. Toxicol
Appl Pharmacol 139, 262-271.
Lee, K. M., Muralidhara, S., Schnellmann, R. G., and Bruckner, J. V. (2000a). Contribution of
direct solvent injury to the dose-dependent kinetics of trichloroethylene: portal vein
administration to rats. Toxicol Appl Pharmacol 164, 46-54.
Lee, K. M., Muralidhara, S., White, C. A., and Bruckner, J. V. (2000b). Mechanisms of the dose-
dependent kinetics of trichloroethylene: oral bolus dosing of rats. Toxicol Appl
Pharmacol 164, 55-64.
Lipscomb, J. C., Fisher, J. W., Confer, P. D., and Byczkowski, J. Z. (1998). In vitro to in vivo
extrapolation for trichloroethylene metabolism in humans. Toxicol Appl Pharmacol 152,
376-387.
Lumpkin, M. H., Bruckner, J. V., Campbell, J. L., Dallas, C. E., White, C. A., and Fisher, J. W.
(2003). Plasma binding of trichloroacetic acid in mice, rats, and humans under cancer
bioassay and environmental exposure conditions. Drug Metab Dispos 31, 1203-1207.
6/22/2009
201
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Merdink, J. L., Gonzalez-Leon, A., Bull, R. J., and Schultz, I. R. (1998). The extent of
dichloroacetate formation from trichloroethylene, chloral hydrate, trichloroacetate, and
trichloroethanol in B6C3F1 mice. Toxicol Sci 45, 33-41.
Merdink, J. L., Stenner, R. D., Stevens, D. K., Parker, J. C., and Bull, R. J. (1999). Effect of
enterohepatic circulation on the pharmacokinetics of chloral hydrate and its metabolites
in F344 rats. J Toxicol Environ Health A 57, 357-368.
Merdink JL, Robison LM, Stevens DK, Hu M, Parker JC, Bull RJ. (2008). Kinetics of chloral
hydrate and its metabolites in male human volunteers. Toxicology. 245:130-40.
Monster, A. C., Boersma, G., and Duba, W. C. (1976). Pharmacokinetics of trichloroethylene in
volunteers, influence of workload and exposure concentration. Int Arch Occup Environ
Health 38, 87-102.
Monster, A. C., Boersma, G., and Duba, W. C. (1979). Kinetics of trichloroethylene in repeated
exposure of volunteers. Int Arch Occup Environ Health 42, 283-292.
Miiller G, Spassowski M, HenschlerD. (1972). Trichloroethylene exposure and trichloroethylene
metabolites in urine and blood. Arch. Toxikol. 29:335-340.
Muller, G., Spassovski, M., and Henschler, D. (1974). Metabolism of trichloroethylene in man.
II. Pharmacokinetics of metabolites. Arch Toxicol 32, 283-295.
Muller, G., Spassowski, M., and Henschler, D. (1975). Metabolism of trichloroethylene in man.
III. Interaction of trichloroethylene and ethanol. Arch Toxicol 33, 173-189.
National Research Council (NRC). (2006). Key Scientific Issues in the Health Risk Assessment of
Trichloroethylene. Washington: National Academy Press.
Paycok, Z.V., Powell, J.F. (1945). The excretion of sodium trichloroacetate. J Pharmacol Exper
Therapeut. 85: 289-293.
Poet TS, Corley RA, Thrall KD, Edwards JA, Tanojo H, Weitz KK, Hui X, Maibach HI, Wester
RC. (2000). Assessment of the percutaneous absorption of trichloroethylene in rats and
humans using MS/MS real-time breath analysis and physiologically based
pharmacokinetic modeling. Toxicol Sci. 56:61-72.
Prout, M. S., Provan, W. M., and Green, T. (1985). Species differences in response to
trichloroethylene. I. Pharmacokinetics in rats and mice. Toxicol Appl Pharmacol 79, 389-
400.
Ramsey JC, Andersen ME. (1984). A physiologically based description of the inhalation
pharmacokinetics of styrene in rats and humans. Toxicol Appl Pharmacol 72:159-175.
Rhomberg LR. (2000). Dose-response analyses of the carcinogenic effects of trichloroethylene in
experimental animals. Environ Health Perspect 108(suppl 2):343-358.
6/22/2009
202
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Saghir SA, Schultz IR. (2002). Low-dose pharmacokinetics and oral bioavailability of
dichloroacetate in naive and GST-zeta-depleted rats. Environ Health Perspect. 110:757-
63.
Sarangapani R, Gentry PR, Covington TR, Teeguarden JG, Clewell HJ 3rd. (2003). Evaluation
of the potential impact of age- and gender-specific lung morphology and ventilation rate
on the dosimetry of vapors. Inhal Toxicol. 15:987-1016.
Sato, A., Nakajima, T., Fujiwara, Y., and Murayama, N. (1977). A pharmacokinetic model to
study the excretion of trichloroethylene and its metabolites after an inhalation exposure.
Br JIndMedSA, 56-63.
Simmons, J. E., Boyes, W. K., Bushnell, P. J., Raymer, J. H., Limsakun, T., McDonald, A., Sey,
Y. M., and Evans, M. V. (2002). A physiologically based pharmacokinetic model for
trichloroethylene in the male long-evans rat. Toxicol Sci 69, 3-15.
Slikker W Jr, Andersen ME, Bogdanffy MS, Bus JS, Cohen SD, Conolly RB, David RM,
Doerrer NG, Dorman DC, Gaylor DW, Hattis D, Rogers JM, Woodrow Setzer R,
Swenberg JA, Wallace K.(2004a). Dose-dependent transitions in mechanisms of toxicity.
ToxicolApplPharmacol. 201:203-25.
Slikker W Jr, Andersen ME, Bogdanffy MS, Bus JS, Cohen SD, Conolly RB, David RM,
Doerrer NG, Dorman DC, Gaylor DW, Hattis D, Rogers JM, Setzer RW, Swenberg
JA,Wallace K. (2004b). Dose-dependent transitions in mechanisms of toxicity: case
studies. Toxicol Appl Pharmacol. 201:226-94.
Staats DA, Fisher JW, Connolly RB. (1991). Gastrointestinal absorption of xenobiotics in
physiologically based pharmacokinetic models. A two-compartment description. Drug
Metab Dispos 19(1): 144-148.
Stadler JC, Kennedy GL Jr. (1996). Evaluation of the sensory irritation potential of volatile
organic chemicals from carpets—alone and in combination. Food Chem Toxicol 34(11-
12): 1125—1130.
Stenner, R. D., Merdink, J. L., Stevens, D. K., Springer, D. L., and Bull, R. J. (1997).
Enterohepatic recirculation of trichloroethanol glucuronide as a significant source of
trichloroacetic acid. Metabolites of trichloroethylene. Drug Metab Dispos 25, 529-535.
Stewart, R. D., Dodd, H. C., Gay, H. H., and Erley, D. S. (1970). Experimental human exposure
to trichloroethylene. Arch Environ Health 20, 64-71.
Templin, M. V., Parker, J. C., and Bull, R. J. (1993). Relative formation of dichloroacetate and
trichloroacetate from trichloroethylene in male B6C3F1 mice. Toxicol Appl Pharmacol
123, 1-8.
6/22/2009
203
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Templin, M. V., Stevens, D. K., Stenner, R. D., Bonate, P. L., Tuman, D., and Bull, R. J. (1995).
Factors affecting species differences in the kinetics of metabolites of trichloroethylene. J
Toxicol Environ Health 44, 435-447.
Thrall KD, Poet TS. (2000). Determination of biokinetic interactions in chemical mixtures using
real-time breath analysis and physiologically based pharmacokinetic modeling. J Toxicol
Environ Health A. 59:653-70.
Triebig, G., Essing, H.-G., Schaller, K.-H., Valentin, H. (1976). Biochemical and psychological
examinations of test subjects exposed to trichloroethylene. Zbl. Bakt. Hyg., I. Abt. Orig.
B 163: 383-416.
USEPA (U.S. Environmental Protection Agency). (1992). Draft Report: A cross-species scaling
factor for carcinogen risk assessment based on equivalence of mg/kg3/4/day;
Notice.Federal Register 57(109)24152-24173.
Vesterberg O, Astrand I. (1976). Exposure to trichloroethylene monitored by analysis of
metabolites in blood and urine. JOccupMed. 18:224-226.
West GB, Woodruff WH, Brown JH (2002) .Allometric scaling of metabolic rate from molecules
and mitochondria to cells and mammals. Proc Natl Acad Sci USA 99: 2473-2478.
Yu, K. O., Barton, H. A., Mahle, D. A., and Frazier, J. M. (2000). In vivo kinetics of
trichloroacetate in male Fischer 344 rats. Toxicol Sci 54, 302-311.
6/22/2009
204
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4 Hazard Characterization
This chapter presents the hazard characterization of TCE health effects. Because of the
number of studies and their relevance to multiple endpoints, the evaluation of epidemiologic
studies of cancer and TCE is summarized in Section 4.0 (endpoint-specific results are presented
in subsequent sections). Genotoxicity data are discussed in Section 4.1. Due to the large number
of endpoints and studies in the toxicity database, subsequent sections (4.2-4.9) are organized by
tissue/organ system. Each section is further organized by noncancer and cancer endpoints,
discussing data from human epidemiologic and laboratory experimental studies. In cases where
there is adequate information, the role of metabolism in toxicity, comparisons of toxicity
between TCE and its metabolites, and carcinogenic MOA are also discussed. Finally, 4.10
summarizes the overall hazard characterization and the weight of evidence for noncancer and
carcinogenic effects.
4.0 Epidemiologic studies on cancer and TCE—summary evaluation
This brief overview of the epidemiologic studies on cancer and TCE below is meant to
provide background to the discussion contained in Sections 4.3-4.9. Over 50 epidemiologic
studies on cancer and TCE exposure (Tables 4.0.1-4.0.3) were examined according to 15
standards of study design (Table 4.0.4), conduct, and analysis in a systematic review. Full
details of the systematic review may be found in Appendix B. Overall, of the more than 50
studies reviewed, 18 studies in which there is a high likelihood of TCE exposure in individual
study subjects (e.g., based on job-exposure matrices or biomarker monitoring) were judged to
have approached, to a sufficient degree, the standards of epidemiologic design and analysis.
Individual studies fully meet all standards and study differences existed in both strengths and
deficiencies. Consideration of possible bias and alternative reasons is necessary to evaluate a
study's ability for identifying a cancer hazard. What follows here is a brief summary of the
results of the evaluation, organized by study type, and of the endpoints and studies analyzed
using meta-analysis.
The cohort studies (Wilcosky et al., 1984; Shindell and Ulrich, 1985; Garabrant et al.,
1988; Costa et al., 1989; Sinks et al., 1992; Axelson et al., 1994; Greenland et al., 1994; Anttila
et al., 1995; Henschler et al., 1995; Ritz, 1999; Blair et al., 1998; Morgan et al., 1998; Boice et
al., 1999, 2006; Hansen et al., 2001; Raaschou-Nielsen et al., 2003; Chang et al., 2003; ATSDR,
2004; Chang et al., 2005; Zhao et al., 2005; Krishnadasan et al., 2007; Sung et al., 2007, 2008;
Clapp and Hoffman, 2008; Radican et al., 2008), with data on the incidence or morality of site-
specific cancer in relation to trichloroethylene exposure, range in size (803 [Hansen et al., 2001]
6/22/2009
205
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
to 86,868 [Chang et al., 2003, 2005]), and were conducted in Denmark, Sweden, Finland,
Germany, Taiwan, and the United States (Table 4.0.1). Three case-control studies nested within
cohorts (Wilcosky et al., 1984; Greenland et al., 1994; Krishnadasan et al., 2007) are considered
as cohort studies because the summary risk estimate from a nested case-control study, the odds
ratio, was estimated from incidence density sampling and is considered an unbiased estimate of
the hazard ratio, similar to a relative risk estimate from a cohort study. Cohort and nested case-
control study designs are analytical epidemiologic studies and are generally relied on for
identifying a causal association between human exposure and adverse health effects (U.S. EPA,
2005a).
Ten of these studies were judged to approach, to a sufficient degree, the standards of
epidemiologic design and analysis: the cohorts of Blair et al. (1998) and its follow-up by Radican
et al. (2008); Morgan et al. (1998), Boice et al. (1999, 2006), Zhao et al. (2005), and
Krishnadasan et al. (2007) of aerospace workers or aircraft mechanics; and Axelson et al. (1994),
Anttila et al. (1995), Hansen et al. (2001), and Raaschou-Nielsen et al. (2003) of Nordic workers
in multiple industries with TCE exposure. Subjects or cases and controls in these studies are
considered to sufficiently represent the underlying population, and the bias associated with
selection of referent populations is considered minimal. The exposure-assessment approaches
included semi-quantitative or quantitative surrogate exposure metrics. The statistical analyses
methods were appropriate and well documented, the measured endpoint was an accurate
indicator of disease, and the follow-up was sufficient for cancer latency. These studies are
considered as high-quality studies for weight-of-evidence characterization. The nested case-
control study of Greenland et al. (1994) also approached many of these standards, however, to a
lesser degree than the 10 high-quality studies. Although TCE was one of several exposures
examined, the low exposure prevalence and ambiguous exposure assessment approach were
judged to lower this study's sensitivity. The remaining cohort studies less satisfactorily meet
identified criteria or standards of epidemiologic design and analysis, having deficiencies in
multiple criteria (Wilcosky et al., 1984; Shindell and Ulrich, 1985; Garabrant et al., 1988; Costa
et al., 1989; Sinks et al., 1992; Henschler et al., 1995; Ritz, 1999; Chang et al., 2003, 2005;
AT SDR, 2004; Sung et al., 2007, 2008; Clapp and Hoffman, 2008).
The case-control studies on TCE exposure are of several site-specific cancers, including
bladder (Siemiatycki, 1991; Siemiatycki et al., 1994; Pesch et al., 2000a); brain (Heineman et al.,
1994; DeRoos et al., 2001); childhood lymphoma or leukemia (Lowengart et al., 1987;
McKinney et al., 1991; Shu et al., 1999; 2004; Costas et al., 2002); colon cancer (Siemiatycki,
1991; Goldberg et al., 2001); esophageal cancer (Siemiatycki, 1991; Parent et al., 2000a); liver
cancer (Lee et al., 2003); lung (Siemiatycki, 1991); adult lymphoma or leukemia (Hardell et al.,
1994 [NHL, Hodgkin lymphoma]; leukemia (Siemiatycki, 1991; Fritschi and Siemiatycki,
6/22/2009
206
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1996a; Nordstrom et al., 1998 [hairy cell leukemia]; Persson and Fredriksson, 1999 [NHL];
Miligi et al., 2006 [NHL and chronic lymphocytic leukemia (CLL)]; Seidler et al., 2007 [NHL,
Hodgkin lymphoma]; Costantini et al., 2008 [leukemia types, CLL included with NHL in Miligi
et al., 2006]); melanoma (Siemiatycki, 1991; Fritchi and Siemiatycki, 1996b); rectal cancer
(Siemiatycki, 1991; Dumas et al., 2000); renal cell carcinoma, a form of kidney cancer
(Siemiatycki, 1991; Parent et al. (2000b); Vamvakas et al., 1998; Dosemeci et al., 1999; Pesch et
al., 2000b; Briining et al., 2003; Charbotel et al., 2006); pancreatic cancer (Siemiatyck, 1991);
and prostate cancer (Siemiatycki, 1991; Aronson et al., 1996). No case-control studies of
reproductive cancers (breast or cervix) and TCE exposure were found in the peer-reviewed
literature.
Seven of the case-control studies meet most evaluation criteria for standards of
epidemiologic design and analysis (Dosemeci et al., 1999; Pesch et al., 2000; Briining et al.,
2003; Miligi et al., 2006; Charbotel et al., 2006; Seidler et al., 2007; Costantini et al., 2008).
Cases and controls in these studies adequately represent underlying populations; bias associated
with selection of referent populations is considered minimal; exposure assessment approaches
included semi-quantitative or quantitative surrogate exposure metrics; face-to-face or telephone
interviews were conducted using a structured questionnaire; and analyses methods were
appropriate, well-documented, and included adjustment for potential confounding exposures.
These studies are considered as high quality for weight-of-evidence characterization of hazard.
Seven other case-control studies (Siemiatycki, 1991 [and related publications, Siemiatyki
et al., 1994; Aronson et al., 1996; Fritchi and Siemiatycki, 1996 a, b; Dumas et al., 2000; Parent
et al., 2000a, b; Goldberg et al., 2001]; Hardell et al., 1994; Nordstrom et al., 1998; Vamvakas et
al., 1998; Persson and Fredriksson, 1999; Shu et al., 1999, 2004; Costas et al., 2002) were judged
to have met many of the evaluation criteria but to a lesser degree. Potential for bias from low
exposure prevalence, self-reported information, or proxy respondents were considered more
likely in these studies compared to the above seven higher-quality case-control studies and may
explain observed findings. Three remaining case-control studies of childhood leukemia
(Lowengart et al., 1987; McKinney et al., 1991) or multiple cancer sites, including liver (Lee et
al., 2003) were judged as low quality for weight-of-evidence characterization of cancer hazard.
The geographic-based studies (Isacson et al., 1985; ADHS, 1990, 1995; Mallin, 1990;
Aicken et al., 1992, 2004; Vartianinen et al., 1993; Cohn et al., 1994, Morgan and Cassady,
2002; ATSDR, 2006, 2008) with data on cancer incidence are correlation studies to examine
cancer outcomes of residents in communities with TCE and other chemicals detected in
groundwater wells or in municipal drinking water supplies. These studies fall short in many of
the 15 criteria for standards of epidemiologic design and analysis. A major deficiency in all
studies is their low level of detail to individual subjects for TCE. One level of exposure to all
6/22/2009
207
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
subjects in a geographic area is assigned without consideration of water distribution networks,
which may influence TCE concentrations delivered to a home, or a subject's ingestion rate to
estimate TCE exposure to individual study subjects. Some inherent measurement error and
misclassification bias is likely in these studies because not all subjects are exposed uniformly.
These studies are of low sensitivity for weight-of evidence characterization of hazard compared
to high-quality cohort and case-control studies.
Examination of heterogeneity in observations for lymphoma, liver cancer, and kidney
cancer between studies was done using meta-analysis methods. Studies judged as high quality
for identifying a cancer hazard are examined and include the following: Axelson et al. (1994),
Anttila et al. (1995), Blair et al. (1998) and its follow-up by Radican et al. (2008), Morgan et al.
(1998), Dosemeci et al. (1999), Boice et al. (1999, 2006), Pesch et al. (2000), Hansen et al.
(2001), Briining et al. (2003), Raaschou-Nielsen et al. (2003), Zhao et al. (2005), Miligi et al.
(2006), Charbotel et al. (2006), and Seidler et al. (2007). Studies of Siemiatycki (1991),
Greenland et al. (1994), Hardell et al. (1994), Nordstrom et al. (1998), and Persson and
Fredriksson (1999) have a decreased sensitivity for identifying a cancer hazard, having met to a
lesser degree criteria compared to the 15 studies above. However, recognizing a predominance
of positive attributes in these studies, they are included in the meta-analysis.
6/22/2009
208
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.0.1: Description of Epidemiologic Cohort Studies Assessing Cancer and Trichloroethylene (TCE) Exposure
Reference Description
Study Group (N)
Comparison Group (N)
Exposure Assessment and Other Information
Aircraft and Aerospace Workers
Radican et al. Aircraft-maintenance workers with
(2008), Blair at least 1 year in 1952-1956 at Hill
et al. (1998) Air Force Base, Utah
Vital status as of 12-31-90; cancer
incidence between 1-1-73 and 12-
31-90 [Blair etal., 1998]
Vital status as of 12-31-2000
[Radican et al., 2008]
Krishnadasan Nested case-control study of
et al. (2007) prostate cancer incidence within a
cohort of 7,618 workers employed
for between 1950 and 1992, or
who had started employment
before 1980 at Boeing/Rockwell/
Rocketdyne (Santa Susana Field
Laboratory) [the UCLA cohort of
Morgenstern et al., 1997], Cancer
incidence ascertained between
1988 and 1999.
14,457 total (7,204 with TCE
exposures)
Incidence [Blairetal., 1998] and
mortality rates [Blair et al., 1998;
Radican et al., 2008] of non-
chemical exposed subjects
326 cases, 1,805 controls
Response rate:
Cases, 69%
Controls, 60%
Industrial hygienist assessment from interviews, surveys, hygiene files, position
descriptions. Cumulative TCE assigned to individual subjects using JEM.
Exposure-response patterns assessed using cumulative exposure, continuous or
intermittent exposures, and peak exposure. TCE replaced in 1968 with 1, 1, 1-
trichloroethane and was discontinued in 1978 in vapor degreasing activities.
Median TCE exposures were about 10 ppm for rag and bucket; 100-200
ppm for vapor degreasing.
Data from company records used to construct a job exposure matrix for
occupational chemical exposures, including TCE. Lifestyle factors obtained
from living subjects through mail and telephone surveys. Statistical analyses
controlled for possible confounders including other occupational exposure such
as hydrazine exposure.
Zhao et al. Aerospace workers with at least 2
(2005); Ritz et years of employment at
al. (1999) Boeing/Rockwell/Rocketdyne and
6,044 (2,689 with high cumulative
exposure to TCE). Mortality rates
of subjects in lowest TCE exposure
Industrial hygienist assessment from walk-through visits, interviews, and review
of historical facility reports. Each job title ranked for presumptive TCE
exposure as high (3), medium (2), low (1), or no (0) exposure. Cumulative TCE
6/22/2009
209
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Description Study Group (N)
Comparison Group (N)
had worked at at Santa Susana category.
Field Laboratory, Ventura, CA,
between 1950 and 1993 [the
UCLA cohort of Morgenstern et al.
(1997)]. Cancer mortality as of
December 31, 2001.
Aerospace workers with at least 2
years of employment at
Boeing/Rockwell/Rocketdyne
(Santa Susana Field Laboratory,
Ventura, CA) between 1950 and
1993 and who were alive as of
1988 [the UCLA cohort of
Morgenstern et al. (1997)] .
Cancer incidence was ascertained
between 1988 and 2000.
Exposure Assessment and Other Information
assigned to individual subjects using JEM. Quartile cut point value of
cumulative exposure scores based upon cumulative exposure scores among
exposed workers. Exposure-response patterns assessed using cumulative
exposure. Industrial hygiene monitoring data were not available and personnel
records did not identify work location for most employees. High exposure to
TCE occurred at rocket engine test stands that involved cleaning of rocket
engines. TCE use also used as a general degreasing solvent to clean metal parts
and mechanics, maintenance and utility workers, and machinists were presumed
with potential TCE exposure. All exposure assignments were made while
blinded to cancer diagnoses. Statistical analyses controlled for possible
confounders including other occupational exposure such as hydrazine exposure.
5,049 (2,227 with high cumulative
exposure to TCE). Incidence rates
of subjects in lowest TCE exposure
category.
Boice et al. Aerospace workers with 6 or more
(2006a) months of employment at
Boeing/Rockwell/Rocketdyne
(Santa Susana Field Laboratory
and nearby facilities) between 1-1-
48 and 12-31-99[IEI cohort, IEI
(2005)].
Vital status as of 12-31-99
41,351 total [1,138,610 P-Y]
1,642 [56,286] male hourly test
stand mechanics
1,111 [39,687] with potential TCE
exposure (TCE subcohort)
Mortality rates of US population
and California population.
Several internal referent groups
including male hourly non-
administrative Rocketdyne
Job title used to identify jobs with test stand work included test stand
mechanics, instrument mechanics, inspectors, test stand engineers and research
engineers. Company phone directories used to identify work location and
assignment to specific test stands and possible exposures in absence of work
history information in company personnel files. Industrial hygienist assessment
from walk-through surveys, interviews and review of medical records used to
identify work location and chemical exposures. Potential TCE exposure
assigned to test stands workers whose tasks included the cleaning or flushing of
rocket engines (engine flush) \n = 639] or for general utility cleaning [n = 472];
potential for exposure to large quantities of TCE was much greater during
engine flush than when TCE used as a utility solvent. No quantitative exposure
6/22/2009
210
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Description
Study Group (N)
Comparison Group (N)
Boice et al.
(1999)
Morgan et al.
(1998)
Aircraft-manufacturing workers
with at least 1 year on or after 1-1-
60 at Lockheed Martin (Burbank,
CA)
Vital status as of 12-31-96
Aerospace workers with at least 6
months employment at Hughes
(AZ plant) between 1-1-50 and 12-
31-85
Vital status as of 12-31-85
workers; male hourly,
nonadministrative SSFL workers;
and test stand mechanics with no
potential exposure to TCE for
intrachohort dose-response
analyses.
77,965 total (2,267 with potential
routine TCE exposures) [66,186 P-
Y] and 3,016 with routine or
intermittent TCE exposure [P-Y
not presented in published paper]
Mortality rates of US population
(routine TCE exposed subjects)
and mortality rates of all other
cohort subjects for analysis of
combined group of routine and
intermittent TCE exposures
20,508 total (4,733 with TCE
exposures) [105,852 P-Y],
Mortality rates of US population
for overall TCE exposure;
mortality rates of all-other cohort
subjects for semi-quantitative TCE
exposure-response analyses
Costa et al.
Workers employed between 1954 8,626 [118,606 P-Y]
6/22/2009
211
Exposure Assessment and Other Information
metric; exposure duration examined in statistical analyses.
Abstracted from walk-through surveys, hygiene files, and job descriptions.
TCE exposure (dichotomous variable) assigned to individual subjects using
JEM. Job title involving potential TCE exposure on routine basis included
process equipment operators and helpers, electroplaters, heat treaters, and sheet
metal forming jobs, straightening press operators, and stretch wrap-forming
machine operators. Job titles with potential TCE exposure on an intermittent
basis included metal bench workers, sheet metal hand formers, tube benders,
fabrication equipment operators, and fabrication and structures development
mechanics. Exposure-response patterns assessed by duration of exposure.
Exposure matrixes generated by employees and industrial hygienists.
Exposure-response patterns assessed using cumulative exposure (low versus
high) and job with highest TCE exposure rating (peak, medium/high exposure
versus no/low exposure). "High exposure" job classification defined as >50
ppm. No data were provided on the frequency of exposure-related tasks and
NRC (2006) noted medium and low rankings were likely highly misclassified
given exposure assignment did not fully consider temporal changes in exposure
intensity.
No exposure assessment was used in this study. Job title used to group jobs into
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Description
Study Group (N)
Comparison Group (N)
(1989)
and 1981 at an aircraft
manufacturing plant in Italy
Vital status ascertained on 06-30-
81
Mortality rates of the Italian
population
Garabrant Workers at an aircraft-
et al. (1988) manufacturing plant with at least 4
years of employment with
company and who had worked at
least 1 day at a plant in San Diego,
CA, between 1-1-58 and 12-31-82
14,067 total [222,100 P-Y]
Mortality rates of US population
Vital status as a 12-31-82
Cohorts Identified From Biological Monitoring (Urinary Trichloroacetic Acid, U-TCA)
Hansen et al. Workers biological monitored for 803 total [16,703 P-Y]
(2001) occupational exposure to TCE
between 1947 and 1989 using U- Cancer incidence rates of the Danish
TCA and air-TCE measurements population
Follow-up for cancer incidence
from 1-1-64 to 12-31-96
6/22/2009
212
Exposure Assessment and Other Information
the following categories: blue- and white-collar workers, technical staff, and
administrative clerks.
Exposure assessment for 70 of 14,067 cohort subjects; 14 cases of esophageal
cancer and 56 matched controls. An examination of company work records of
jobs held by these 70 subjects identified 37% with potential TCE exposure. No
information on TCE exposure potential to the remaining ~ 14,000 subjects.
Of the 803 subjects, 712 had U-TCA, 89 had air-TCE measurement records,
and 2 had records of both types. U-TCA covered period from 1947-1989; air
TCE measurements from 1974. Mean and median concentrations of U-TCA
were 250 |imol/L and 92 |imol/L: using the Ikeda et al. (1972) relationship for
TCE exposure to U-TCA, mean and median exposures were ~14 ppm and ~5
ppm. Historic median exposures estimated from the U-TCA concentrations
were low: 9 ppm for 1947 to 1964, 5 ppm for 1965 to 1973, 4 ppm for 1974 to
1979, and 0.7 ppm for 1980 to 1989. Air TCE measurements from 1974
onward were 19 ppm (mean) and 5 ppm (median). Overall, median TCE
exposure to cohort as extrapolated from air TCE and U-TCA
measurements was 4 ppm (arithmetic mean, 12 ppm).
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Description
Study Group (N)
Comparison Group (N)
Anttila et al. Workers biological monitored for
(1995) occupational exposure to TCE
between 1965 and 1982
3,974 total (3,089 withU-TCA
measurements) [59,905 P-Y]
Axelson et al.
(1994)
Follow-up for mortality through
1965 to 1991 and from 1967 to
1992 for incidence
Workers biological monitored for
occupational exposure to TCE
between 1955 and 1975
Mortality and cancer incidence rates
of the Finnish population
1,4,21 males
22,447 P-Y, mortality
23,517 P-Y, incidence
Other Cohorts
Clapp and
Hoffman
(2008)
Sung et al.
(2007, 2008)
Follow-up for mortality through
1986 and from 1958 to 1987 for
incidence
Deaths between 1969 and 2001
among employees > 5 year
employment duration at an IBM
facility in Endicott, NY
Female workers with first date of
employment from 1973 and 1997
at an electronics factory in
Taoyuan, Taiwan.
Follow-up for cancer incidence
from 1979 to 2001 (Sung et al.,
Mortality and cancer incidence rates
of Swedish male population
360 deaths among males and
females [Size and P-Y of worker
population are not known]
Proportion of deaths among New
York residents during 1979 to 1998
63,982 females [1,403,824 P-Y]
40, 647 with first live born offspring
Cancer incidence rates of Taiwan
population (Sung et al., 2007)
6/22/2009
213
Exposure Assessment and Other Information
Biological monitoring for U-TCA. Median U-TCA, 63 |imol/L for females
and 48 |imol/L for males; mean U-TCA was 100 |iol/L. There were on
average 2.5 U-TCA measurements per individual. Using the Ikeda et al.
(1972) relationship for TCE exposure to U-TCA, TCE exposures were
roughly 4 ppm (median) and 6 ppm (mean).
Biological monitoring for U-TCA from 1955 and 1975. Roughly % of cohort
had U-TCA concentrations equivalent to <20 ppm TCE.
No exposure assessment was used in this study.
No exposure assessment was used in this study. The electronic factory began
operations in May 1968 and closed in 1992. National Labor Department
inspection reports and the company's import/export statistics indicated use of
many chlorinated solvents including TCE and perchloroethylene. These
records indicated TCE was not used between 1975 and 1991 and
perchloroethylene was used after 1981. No information was available as to use
in other time periods. Published paper does not report TCE usage, potential
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Description
Study Group (N)
Comparison Group (N)
2007)
Childhood leukemia between
1979-2001 among first born of
female subjects in Sung et al.
(2007) (Sung et al., 2008)
Childhood leukemia incidence rates
of first born live births of Taiwan
population (Sung et al., 2007)
Chang et al. Male and female workers
(2005), Chang employed between 1978 and 1997
et al. (2003) at electronics factory as studied by
Sung et al. (2007)
86,868 [1,022,094 P-Y], mortality
86,868 [1,380,355 P-Y], incidence
Follow-up for mortality from 1985 Incidence (Chang et al., 2005) or
to 1997 and for cancer incidence
from 1979 to 1997.
ATSDR Workers employed between 1952
(2004) and 1980 at the View-Master
factory in Beaverton, OR.
mortality (Chang et al., 2003) rates
Taiwan population.
13,697 former employees identified
by plant owners in 1998 [Size and
P-Y of worker population are not
known]
Proportion of deaths between
1989-2001 in Oregon population
Raaschou- Blue-collar workers employed
Nielsen et al. since 1-1-68 at 347 Danish TCE-
(2003) using companies
40,049 total (14,360 with
presumably higher level exposure to
TCE) [339,486 P-Y]
Follow-up for cancer incidence
Cancer incidence rates of the Danish
6/22/2009
214
Exposure Assessment and Other Information
TCE exposure concentrations, or the percentage of study subjects whose job
titles indicated potential TCE exposure. A number of chlorinated solvents
were also found in soil and groundwater at factory site.
No exposure information on individual subjects, but study assumes TCE
exposure via drinking water to all employees. TCE and other VOCs detected
in well water at the time of the plant closure in 1998: TCE, 1,220-1,670 |ig/L;
1, 1,-DCE, up to 33 |ig/L: and, perchloroethylene up to 56 |ig/L. TCE used to
degrease metal equipment with most of degreasing occurring in one building,
the Paint Shop with disposal of waste TCE on plant grounds. Potential existed
for inhalation and dermal exposure associated with degreasing activities but
information is lacking on estimated exposure levels.
Employers had documented TCE usage. Blue-collar versus white-collar
workers and companies with <200 workers were variables identified as
increasing the likelihood for TCE exposure. Subjects were identified from the
following industries: iron and metal, electronics, painting, printing, chemical,
and dry cleaning. Median exposures to trichloroethylene were 40-60 ppm
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Description
Study Group (N)
Comparison Group (N)
through 12-31-97
population
Ritz (1999a) Male uranium-processing plant
workers with at least 3 months
employment at DOE facility in
Fernald, OH between 1-1-51 and
12-31-72.
Follow-up for cancer mortality
from 1-1-51 to 12-31-89.
Henschler et Male workers with at least 1 year
al. (1995) employment between 1956 and
1975 at cardboard factory
Vital status as of 12-31-92
Greenland et Cancer deaths known to employer
al. (1994) among pensioned GE workers at a
transformer manufacturing plant in
Pittsfield, MA, employed before
1984, who had died between
1969-1984, whose death was
reported to company pension plan,
and who had job history record;
controls were noncancer deaths
from same underlying cohort as
cases
3,814 white males[120,237 P-Y]
monitored for radiation with
2,971 with potential TCE exposure
Mortality rates of the US
population; Non-exposed internal
controls for TCE exposure-response
analyses
169 exposed [5,188 P-Y]
190 unexposed [6,100 P-Y]
Renal cancer incidence rates of
Danish and former German
Democratic Republic populations
512 cases, 1,202 controls
Response rate:
Cases, 69%
Controls, 60%
6/22/2009
215
Exposure Assessment and Other Information
for the years before 1970,10-20 ppm for 1970 to 1979, and approximately
4 ppm for 1980 to 1989.
Exposure matrixes for TCE, cutting fluids, kerosene, and radiation generated
by employees and industrial hygienists. Subjects were assigned potential TCE
according to intensity: light (2,792 subjects), moderate (179 subjects), heavy
(no subjects).
Walk-through surveys and employee interviews used to identify work areas
with TCE exposure. TCE-exposed renal cancer cases identified from national
workman's compensation files.
Industrial hygienist assessment from interviews and position descriptions.
TCE (no/any exposure) assigned to individual subjects using JEM.
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Description
Study Group (N)
Comparison Group (N)
Sinks et al. Deaths among workers employed
(1992) between 1-1-57 and 6-30-88 at a
paperboard container
manufacturing and printing plant
in Newnan, GA.
Vital status as of 6-30-88.
Kidney and bladder cancer
incidence using the Atlanta
Metropolitan Area Surveillance,
Epidemiology, and End Results
(SEER) registry, the Atlanta SEER
ineligible file and the Georgia
State Tumor Registry through 12-
31-90.
2,050 total [36,744 P-Y]
Mortality rates of the U.S.
population.
Bladder and kidney cancer
incidence rates from the Atlanta-
SEER registry for the years
1973-1977.
Nested case-control study of
kidney cancer cases or deaths
carried out to examine possible
association with work department..
Eight controls per cases were
randomly identified from all
employees and matched to cases on
date of birth (+ 5 years), age of case
at diagnosis or death, and sex;
control's age of first employment at
plant was less than that of case.
Shindell and Workers at a plant manufacturing
Ulrich (1985) trichloroethylene who were
employed fro three or more
2,646 males and females [16,332 P-
Y]
6/22/2009
216
Exposure Assessment and Other Information
No exposure assessment carried out for mortality and incidence analyses;
analyses of all plant employees including white- and blue-collar employees.
Assignment of work department in case-control study based upon work
history. Potential carcinogenic agents used in work departments based upon
material Safety Data Sheets and communication with product manufacturer.
Study does not assign potential exposures to individual subjects.
All employees (white-, blue-collar) at one plant manufacturing
trichloroethylene assumed to have potential TCE exposure regardless of job
title. No exposure assessment of trichloroethylene potential to individual
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Description
Study Group (N)
Comparison Group (N)
months between 1-1-57 and 7-31- Mortality rates of the US population
83.
Follow-up for mortality as defined
using broad categories to 7-31-83.
Wilcosky et al. Cancer deaths due to respiratory,
(1984) stomach, prostate,
lymphosarcoma, and lymphatic
leukemia among production
workers aged 40-79 employed
beginning in 1964 at a rubber plant
in Akron, Ohio; controls were a
20% age-stratified random sample
of the cohort.
183 cases of which 9 were due to
lymphosarcoma and 10 due to
lymphatic leukemia.
Response rate:
Not available in paper
6/22/2009
217
Exposure Assessment and Other Information
subjects and no information on TCE production processes or results of
industrial hygiene monitoring.
Worker exposure linked to job title, department and dates of employment for
the period through 1973. Plant documents on raw material and product
specifications and operating procedures used to identify TCE and other solvent
use by process area and calendar year. Industrial hygiene monitoring of
exposure concentrations did not support job exposure matrix.
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.0.2: Case-Control Epidemiologic Studies Examining Cancer and Trichloroethylene (TCE) Exposure
Reference Population
Cases Con- Response Exposure Assessment
trols Rates
Statistical Analyses
Bladder
Pesch et al.
(2000a)
Histologically confirmed urothelial 1,035
cancer (bladder, ureter, renal
pelvis) cases from German
hospitals (5 regions) in
1991-1995; controls randomly
selected from residency registries
matched on region, sex, and age
4,298 Cases, 84% In-person interview with case or next-of-
Controls, kin; questionnaire assessing occupational
71% history using job title or self-reported
exposure to assign TCE and other
exposures; exposure assigned using job-
exposure-matrix and job-task-exposure
matrix.
Logistic regression with
covariates for age, study
center, and smoking.
Siemiatycki et
al. (1994),
Siemiatycki
(1991)
Brain
DeRoos et al.
(2001)
Olshan et al.
(1999)
Male bladder cancer cases, age 484
35-75 years, diagnosed in 16 large
Montreal-area hospitals in
1979-1985 and histologically
confirmed; controls identified
concurrently at 18 other cancer
sites; age-matched, population-
based controls identified from
electoral lists and random digit
dialing (RDD).
Neuroblastoma cases in children of 504
<19 years selected from Children's
cancer Group and Pediatric
Oncology Group with diagnosis in
1992-1994; population controls
533
population
controls
and
740
subjects
with
other
cancers
504
Cases, 78%
Controls-
Population,
72%
Cases, 73%,
Controls,
74%
In-person interviews (direct or proxy)
with segments on work histories (job
titles and self-reported exposures);
analyzed and coded by a team of chemists
and industrial hygienists (294 exposures
on semi-quantitative scales).
Telephone interview with parent using
questionnaire to assess parental
occupation and self-reported exposure
history and judgment-based attribution of
exposure to TCE and other solvents.
Logistic regression adjusted
for age, ethnic origin,
socioeconomic status,
smoking, coffee
consumption, and
respondent status or
Mantel-Haenszel stratified
on age, income, index for
cigarette smoking, coffee
consumption, and
respondent status.
Logistic regression with
covariate for child's age and
material race, age, and
education
6/22/2009
218
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Population
Cases
Con-
trols
Response Exposure Assessment
Rates
Statistical Analyses
(random digit dialing) matched to
control on birth date
Heineman et White, male cases, age > 30 years,
al. (1994) identified from death certificates in
1978-1981; controls identified
from death certificates and
matched for age, year of death and
study area
300
386 Cases, 74% In-person interview with next-of-kin;
Controls, questionnaire assessing lifetime
63% occupational history using job title and
job-exposure matrix of Gomez et al.
(1994).
Logistic regression with
covariates for age and study
area
Colon and Rectum
Goldberg et Male colon cancer cases, age 497
al. (2001), 35-75 years, diagnosed in 16 large
Siemiatycki Montreal-area hospitals in
(1991) 1979-1985 and histologically
confirmed; controls identified
concurrently at 18 other cancer
sites; age-matched, population-
based controls identified from
electoral lists and random digit
dialing (RDD).
533
population
controls
and
740
subjects
with
other
cancers
Cases, 82% In-person interviews (direct or proxy)
Controls- with segments on work histories (job
Population, titles and self-reported exposures);
72% analyzed and coded by a team of chemists
and industrial hygienists (294 exposures
on semi-quantitative scales).
Logistic regression adjusted
for age, ethnic origin,
birthplace, education,
income, parent's
occupation, smoking,
alcohol consumption, tea
consumption, respondent
status, heating source
socioeconomic status,
smoking, coffee
consumption, and
respondent status or
Mantel-Haenszel stratified
on age, income, index for
cigarette smoking, coffee
consumption, and
respondent status.
6/22/2009
219
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Population
Cases
Con-
trols
Response
Rates
Exposure Assessment
Statistical Analyses
Dumas et al.
Male rectal cancer cases, age
292
533
Cases, 78%
In-person interviews (direct or proxy)
Logistic regression adjusted
(2000)
35-75 years, diagnosed in 16 large
population
Controls-
with segments on work histories (job
for age, education,
Montreal-area hospitals in
controls
Population,
titles and self-reported exposures);
respondent status, cigarette
1979-1985 and histologically
and
72%
analyzed and coded by a team of chemists
smoking, beer consumption
confirmed; controls identified
740
and industrial hygienists (294 exposures
and body mass index;
concurrently at 18 other cancer
subjects
on semi-quantitative scales).
Mantel-Haenszel stratified
sites; age-matched, population-
with
on age, income, index for
based controls identified from
other
cigarette smoking, coffee
electoral lists and random digit
cancers
consumption, ethnic origin,
dialing (RDD).
and beer consumption.
Fredriksson et
Colon cancer cases aged 30-75
329
658
Not
Mailed questionnaire assessing
Age, sex, physical activity
al. (1989)
years identified through the
available
occupational history with telephone
Swedish Cancer Registry among
interview follow-up.
patients diagnosed in 1980-1983;
population-based controls were
frequency-matched on age and sex
and were randomly selected from a
population register
Esophagus
Parent et al.
Male esophageal cancer cases, age
292
533
Cases, 78%
In-person interviews (direct or proxy)
Logistic regression adjusted
(2000a),
35-75 years, diagnosed in 19 large
population
Controls-
with segments on work histories (job
for age, education,
Siemiatycki
Montreal-area hospitals in
controls
Population,
titles and self-reported exposures);
respondent status, cigarette
(1991)
1979-1985 and histologically
and
72%
analyzed and coded by a team of chemists
smoking, beer consumption
confirmed; controls identified
740
and industrial hygienists (294 exposures
and body mass index;
concurrently at 18 other cancer
subjects
on semi-quantitative scales).
Mantel-Haenszel stratified
sites; age-matched, population- with on age, income, index for
6/22/2009
220
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Population
Cases Con- Response Exposure Assessment
trols Rates
Statistical Analyses
based controls identified from
electoral lists and random digit
dialing (RDD).
other
cancers
cigarette smoking, coffee
consumption, ethnic origin,
and beer consumption.
Lymphoma
Wang et al.
(2009)
Constantini et
al. (2008),
Miligi et al.
(2006)
Cases among females aged 21 and 601 717
84 years with non-Hodgkin's
lymphoma (NHL) in 1996 - 2000
and identified from Connecticut
Cancer Registry; population-based
female controls (1) if <65 years of
age, having Connecticut address
stratified by 5-year age groups
identified from random digit
dialing or (2) >65 years of age, by
random selection from Centers for
Medicare and Medicaid Service
files.
Cases aged 20-74 with NHL, 1,428 NHL
including chronic lymphocytic + CLL,
leukemia (CLL), all forms of
leukemia, or multiple myeloma 586
(MM) in 1991-1993 and identified Leukemia 1,278
through surveys of hospital and
pathology departments in study 263, MM
areas and in specialized
hematology centers in 8 areas in 1,100
Italy; population-based controls
Cases, 72%
Controls,
69% (<65
years) and
47% (>65
years)
Cases,
83%
Controls,
73%
Cases, 85%
Controls,
71%
Cases 83%
In-person interview with using
questionnaire assessment specific jobs
held for >1 year. Intensity and
probability of exposure to broad category
of organic solvents and to individual
solvents, including TCE, estimated using
job-exposure matrix of NCI (Gomez et al,
1994; Dosemeci et al., 1994) and
assigned blinded to case and control
status.
In-person interview primarily at
interviewee's home (non-blinded
interview) using questionnaire assessing
specific jobs, extraoccupational exposure
to solvents and pesticides, residential
history, and medical history.
Occupational exposure assessed by job-
specific or industry-specific
questionnaires. All NHL diagnoses and
20% sample of all cases confirmed by
Logistic regression adjusted
for age, family history of
hematopoietic cancer,
alcohol consumption and
race.
Logistic regression with
covariates for sex, age,
region, and education.
Logistic regression for NHL
cell types include an
additional covariate for
smoking.
6/22/2009
221
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Population
Cases Con- Response Exposure Assessment
trols Rates
Statistical Analyses
stratified by 5-year age groups and
by sex selected through random
sampling of demographic or of
National Health Service files.
Controls
76%
panel of 3 pathologists.
Seidler et al. NHL and Hodgkin's disease cases 710
(2007) aged 18-80 years identified
Mester et al. through all hospitals and
(2006) ambulatory physicians in six
Becker et al. regions of Germany between 1998
(2004) and 2003; population controls were
identified from population registers
and matched on age, sex, and
region
710 Cases, 87% In-person interview using questionnaire
Controls, assessing personal characteristics, lifestyle,
44% medical history, UV light exposure, and
occupational history of all jobs held for 1
year or longer. Exposure of a prior
interest were assessed using job task-
specific supplementary questionnaires.
Age, sex, region, smoking
and alcohol consumption
Perssonand Histologicallly confirmed cases of NHL, 199
Fredriksson B-cell NHL, age 20-79 years,
(1999) identified in two hospitals in
Combined Sweden: Oreboro in 1964-1986
analysis of (Persson et al., 1989) and in
NHL cases in Linkoping between 1975-1984
Persson et al. (Persson et al., 1993); controls
(1993), were identified from previous
Persson et al. studies and were randomly selected
(1989) from population registers
NHL, 479 Cases, 96% Mailed questionnaire to assess self
(Oreboro)
90%
(Linkoping)
Controls,
not reported
reported occupational exposures to TCE
and other solvents.
Unadjusted Mantel-
Haenszel chi-square,
Nordstrom et Histologically-confirmed cases in
al. (1998) males of hairy-cell leukemia
111
400 Cases, 91% Mailed questionnaire to assess self
Controls, reported working history, specific
Univariate analysis for
chemical-specific exposure
6/22/2009
222
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Population Cases Con- Response
trols Rates
reported to Swedish Cancer 83%
Registry in 1987-1992 (includes
one case latter identified with an
incorrect diagnosis date);
population-based controls
identified from the National
Population Registry and matched
(1:4 ratio) to cases for age and
county
Fritschi and
Male NHL cases, age 35-75 years, 215
533
Cases, 83%
Siemiatycki,
diagnosed in 16 large
population
Controls-
1996a),
Montreal-area hospitals in
controls
Population,
Siemiatycki
1979-1985 and histologically
(Group 1)
71%
(1991)
confirmed; controls identified
and
concurrently at 18 other cancer
1,900
sites; age-matched, population-
subjects
based controls identified from
with
electoral lists and random digit
other
dialing (RDD).
cancers
(Group 2)
Hardell et al.
Histologically-confirmed cases of 105
335
Not
(1994, 1981)
NHL in males, age 25-85 years,
available
admitted to Swedish (Umea)
hospital between 1974-1978;
living controls (1:2 ratio) selected
from the National Population
Register, matched to living cases
6/22/2009
223
Exposure Assessment
exposure, and leisure time activities.
Statistical Analyses
such as TCE.
In-person interviews (direct or proxy)
with
segments on work histories (job titles and
self-reported exposures); analyzed and
coded by a team of chemists and
industrial hygienists (294 exposures on
semi-quantitative scales).
Mantel-Haenszel stratified
by age, body mass index,
and cigarette smoking and
logistic regression adjusted
for age, proxy status,
income and ethic origin
Self-administered questionnaire assessing Unadjusted Mantel-
self-reported solvent exposure; phone Haenszel chi-square
follow-up with subject, if necessary.
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Population
Cases
Con-
trols
Response Exposure Assessment
Rates
Statistical Analyses
on sex, age, and place of residence;
deceased controls selected from the
National Registry for Causes of
Death, matched (1:2 ratio) to dead
cases for sex, age, place of
residence, and year of death
Persson et al. Histologicallly confirmed cases of
(1993), Hodgkin's disease, age 20-80
Persson et al. years, identified in two hospitals in
(1989) Sweden: Oreboro in 1964-1986
(Persson et al., 1989) and in
Linkoping between 1975-1984
(Persson et al., 1993); controls
randomly selected from population
registers
54 (1989 275 (1989 Not
study); study); available
31 (1993 204 (1993
study) study)
Mailed questionnaire to assess self
reported occupational exposures to TCE
and other solvents.
Logistic regression with
adjustment for age and other
exposure; unadjusted
Mantel-Haentzel chi-square.
Childhood Leukemia
Shu et al. Childhood leukemia cases, < 15
(2004, 1999) years, diagnosed between 1989 and
1993 by a Children's Cancer
Group member or affiliated
institute; population controls
(random digit dialing), matched for
age, race, and telephone area code
and exchange
1,842 1,986 Cases, 92% Telephone interview with mother, and
Controls, whenever available, fathers using
77% questionnaire to assess occupation using
job-industry title and self-reported
exposure history. Questionnaire included
questions specific for solvent, degreaser
or cleaning agent exposures.
Logistic regression with
adjustment for maternal or
paternal education, race, and
family income. Analyses of
paternal exposure also
included age and sex of the
index child.
Costas et al. Childhood leukemia (<19 years
19
37
Cases, 91% Questionnaire administered to parents
Logistic regression with
6/22/2009
224
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Population
Cases Con- Response Exposure Assessment
trols Rates
Statistical Analyses
(2002),
MADPH
(1997)
McKinney et
al. (1991)
Lowengart et
al. (1987)
age) diagnosed in 1969-1989 and
who were resident of Woburn.
MA; controls randomly selected
from Woburn public School
records, matched for age
Incident cases of childhood
leukemia and non-Hodgkin's
lymphoma, ages not identified,
identified in three geographical
areas in England in 1974 and 1988;
controls randomly selected from
children who were children of
residents in the three area at the
time of case diagnosis in area and
matched for sex and birth health
district.
Childhood leukemia cases aged <
10 years and identified from the
Los Angeles (CA) Cancer
Surveillance Program in
1980-1984; controls selected from
random digit dialing or from
friends of cases and matched on
age, sex, and race
109
206
123
123
Control, NA
Cases, 72%
Controls,
77%
Cases, 79%
Controls,
Not
available
separately assessing demographic and
lifestyle characteristics, medical history
information, environmental and
occupational exposure and use of public
drinking water in the home. Hydraulic
mixing model used to infer drinking
water containing TCE and other solvents
delivery to residence
In-person interview with questionnaire
with mother to assess maternal
occupational exposure history, and with
father and mother, as surrogate, to assess
paternal occupational exposure history.
No information provided in paper
whether interviewer was blinded as to
case and control status.
Telephone interview with questionnaire
to assess parental occupational and self-
reported exposure history.
composite covariate, a
weighted variable of
individual covariates
Matched pair design using
logistic regression for
univariate and multivariate
analysis
Matched (discordant) pair
analysis
6/22/2009
225
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Population
Cases Con- Response Exposure Assessment
trols Rates
Statistical Analyses
Melanoma
Fritschi and
Siemiatycki
(1996b),
Siemiatycki
(1991)
Male melanoma cases, age 35-75
years, diagnosed in 16 large
Montreal-area hospitals in
1979-1985 and histologically
confirmed; controls identified
concurrently at 18 other cancer
sites; age-matched, population-
based controls identified from
electoral lists and random digit
dialing (RDD).
103
533
population
controls
and
533
subjects
with
other
cancers
Cases, 78% In-person interviews (direct or proxy)
Controls- with segments on work histories (job
Population, titles and self-reported exposures);
72% analyzed and coded by a team of chemists
and industrial hygienists (294 exposures
on semi-quantitative scales).
Logistic regression adjusted
for age, education, and ethic
origin; Mantel-Haenszel
stratified on age, income,
index for cigarette smoking,
and ethnic origin.
Prostate
Aronson et al.
(1996),
Siemiatycki
(1991)
Renal Cell
Charbotel et
al. (2006,
2009)
Male prostate cancer cases, age 449
35-75 years, diagnosed in 16 large
Montreal-area hospitals in
1979-1985 and histologically
confirmed; controls identified
concurrently at 18 other cancer
sites; age-matched, population-
based controls identified from
electoral lists and random digit
dialing (RDD).
Cases from Arve Valley region in 87
France identified from local
urologists files and from area
533
population
controls
(Group 1)
and
other
cancer
cases from
same
study
(Group 2)
316
Cases, 81%
Controls-
Population,
72%
Cases, 74%
Controls,
78%
In-person interviews (direct or proxy)
with segments on work histories (job
titles and self-reported exposures);
analyzed and coded by a team of chemists
and industrial hygienists (294 exposures
on semi-quantitative scales).
Telephone interview with case or control,
or, if deceased, with next-of-kin (22%
cases, 2% controls). Questionnaire
Logistic regression adjusted
for age, ethnic origin,
socioeconomic status,
Quetlet, and respondent
status or
Mantel-Haenszel stratified
on age, income, index for
cigarette smoking, ethnic
origin, and respondent
status.
Conditional logistic
regression with covariates
for tobacco smoking and
6/22/2009
226
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Population Cases Con- Response
trols Rates
teaching hospitals; age- and sex-
matched controls chosen from file
of same urologist as who treated
case or recruited among the
patients of the case's general
practitioner.
Briining et al. Histologically-confirmed cases 134
(2003) from German hospitals (Arnsberg)
in 1992-2000; controls from
hospital with urology department
serving Arnsberg, and local
geriatric department, for older
controls, matched by sex and age
to cases
Peschetal. Histologically-confirmed cases 935
(2000b) from German hospitals (5 regions)
in 1991-1995; controls randomly
selected from residency registries
matched on region, sex, and age
401
Cases, 83%
4,298
Cases,
Controls
71%
Parent et al. Male renal cell carcinoma cases,
(2000b), age 35-75 years,
Siemiatycki diagnosed in 161arge
(1991) Montreal-area hospitals in
142
533 Cases, 82%
population Controls,
controls 71%
(Group 1)
6/22/2009
227
Exposure Assessment
Statistical Analyses
assessing occupational history,
particularly, employment in the screw
cutting jobs, and medical history. Semi-
quantitative TCE exposure assigned to
subjects using a task/TCE-Exposure
Matrix designed using information
obtained from questionnaires and routine
atmospheric monitoring of work shops or
biological monitoring (U-TCA) of
workers carried out since the 1960s.
In-person interviews with case or next-of-
kin; questionnaire assessing occupational
history using job title and job-exposure
matrix of Pannett et al. (1985) to assign
exposure to TCE and PERC.
body mass index.
Logistic regression with
covariates for age, sex, and
smoking
In-person interview with case or next-of-
kin; questionnaire assessing occupational
history using job title (JEM approach),
self-reported exposure, or job task (JTEM
approach) to assign TCE and other
exposures.
In-person interviews (direct or proxy)
with segments on work histories (job
titles and
self-reported exposures); analyzed and
Logistic regression with
covariates for age, study
center, and smoking
Mantel-Haenszel stratified
by age, body mass index,
and cigarette smoking;
logistic regression adjusted
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Population
Cases Con- Response Exposure Assessment
trols Rates
Statistical Analyses
1979-1985 and histologically
confirmed; controls identified
concurrently at 18 other cancer
sites; age-matched, population-
based controls identified from
electoral lists and random digit
dialing (RDD).
Dosemeci et Histologically-confirmed cases in
al. (1999) white males and females, age
20-85, identified through the
Minnesota Cancer Registry in
1988-1990; controls were
stratified for age and sex and were
randomly selected using random
digit dialing, age 21-64 years, or
from Health Care Financing
Administration records, for age
64-85 years
438
and
other
cancer
cases
(excluding
lung and
bladder
cancers)
(Group 2)
687
Cases, 87%
Controls,
86%
coded by a team of chemists and
industrial hygienists (about 300 exposures
on semi-quantitative scales).
In-person interviews with case or next-of-
kin; questionnaire assessing occupational
history of TCE using job title and job-
exposure matrix of Gomez et al. (1994).
for respondent status, age,
smoking, and body mass
index.
Logistic regression with
covariates for age, smoking,
hypertension, and body
mass index
Vamvakas et Cases who underwent
al. (1998) nephrectomy in 1987-1992 in a
hospital in Arnsberg region of
Germany; controls selected
accident wards from nearby
hospital in 1992
58
84
Cases, 83% In-person interview with case or next-of-
Controls, kin; questionnaire assessing occupational
75% history using job title or self-reported
exposure to assign TCE and
tetrachloroethylene exposures.
Logistic regression with
covariates for age, smoking,
body mass index,
hypertension, and diuretic
intake
6/22/2009
228
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Population
Cases Con- Response Exposure Assessment
trols Rates
Statistical Analyses
Multiple or Other Sites
Lee et al. Liver, lung, stomach, colorectal 53 liver 286
(2003) cancer deaths in males and females 39 stomach
between 1966-1997 from two 26 colorectal
villages in Taiwan; controls were 41 lung
cardiovascular and cerebral-
vascular disease deaths from same
underlying area as cases
Siemiatycki Male cancer cases, age 35-75 857 lung, 533
(1991) years, diagnosed in 16 large 117 population
Montreal-area hospitals in pancreas controls
1979-1985 and histologically (Group 1)
confirmed; controls identified and
concurrently at 18 other cancer other
sites; age-matched, population- cancer
based controls identified from cases from
electoral lists and random digit same
dialing (RDD). study
(Group 2)
Not reported Residence as recorded on death certificate Mantel-Haenszel stratified
by age, sex, and time period
Cases, 79% In-person interviews (direct or proxy)
(lung), 71% with segments on work histories (job
(pancreas) titles and self-reported exposures);
Controls- analyzed and coded by a team of chemists
Population, and industrial hygienists (294 exposures
72% on semi-quantitative scales).
Mantel-Haenszel stratified
on age, income, index for
cigarette smoking, ethnic
origin, and respondent status
(lung cancer) and age,
income, index for cigarette
smoking, and respondent
status (pancreatic cancer).
6/22/2009
229
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.0.3: Geographic-Based Studies Assessing Cancer and Trichloroethylene (TCE) Exposure
Reference
Description
Analysis Approach
Exposure Assessment
Broome County, NY Studies
ATSDR
Total, 22 site-specific, and
Standardized incidence ratios among all subjects
Two study areas, Eastern and Western study areas, identified
(2006a,
childhood cancer incidence
(ATSDR, 2006a) or among white subjects only
based on potential for soil vapor intrusion exposures as defined
2008)
from 1980-2001 among
(ATSDR, 2008) with expected numbers of cancers
by the extent of likely soil vapor contamination. Contour lines
residents in 2 areas in
derived using age-specific cancer incidence rates for
of modeled VOC soil vapor contamination levels based on
Endicott, NY.
New York State, excluding New York City. Limited
exposure model using GIS mapping and soil vapor sampling
assessment of smoking and occupation using medical
results taken in 2003. The study areas were defined by 2000
and other records in lung and kidney cancer subjects
Census block boundaries to conform to model predicted areas of
(ATSDR, 2008).
soil vapor contamination. TCE was the most commonly found
contaminant in indoor air in Eastern study area at levels ranging
from 0.18 to 140 |ig/m3, with tetrachloroethylene, cis-1,2-
dichloroethene, 1,1,1-trichloroethane, 1,1-dichloroethylene, 1,1-
dichloroethane, and Freon 113 detected at lower levels. PCE
was most common contaminant in indoor air in Western study
area with other VOCs detected at lower levels.
Maricopa County, AZ Studies
Aickin et al.
Deaths due to cancer,
Standardized rate ratios for mortality from Poisson
Location of residency in Maricopa County, AZ, at the time of
(1992)
including leukemia in
regression modeling. Childhood leukemia incidence
death as surrogate for exposure. Some analyses examined
Aickin
1966-1986 and childhood
data evaluated using Bayes methods and Poisson
residency in West Central Phoenix and cancer. Exposure
(2004)
(<19 years old) leukemia
regression modeling.
information is limited to TCE concentration in two drinking
incident cases (1965-1986)
water wells in 1982.
among residents of
Maricopa County, AZ.
Pima County, AZ Studies
ADHS
Cancer incidence in
Standardized incidence rate ratios from Poisson
Location of residency in Pima, County, AZ, at the time of
6/22/2009
230
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Description
Analysis Approach
Exposure Assessment
(1990, 1995)
children (<19 years old)
regression modeling using method of Aickin et al.
diagnosis or death as surrogate for exposure. Exposure
and testicular cancer in
(1992). Analysis compares incidence in Tucson
information is limited to monitoring since 1981 and include:
1970-1986 and 1987-1991
Airport Area to rate for rest of Pima County.
VOCs in soil gas samples (TCE, perchloroethylene, 1, 1-
among residents of Pima
dichloroethylene, 1, 1, 1-tirchloroacetic acid); PCBs in soil
County, AZ.
samples, and TCE in municipal water supply wells.
Other
Morgan and
Incident cancer cases
Standardized incidence rates for all cancer sites and 16
TCE and perchlorate detected in some county wells; no
Cassady
diagnosed between 1-1-88
site-specific cancers; expected numbers of cancers
information on location of wells to residents, distribution of
(2002)
and 12-31-98 among
using incidence rates of site-specific cancer of a four-
contaminated water, or TCE exposure potential to individual
residents of 13 census
county region between 1988-1992
residents in studied census tracts.
tracts in Redlands area, San
Bernardino County, CA.
Vartiainen et
Total cancer and site-
Standardized incidence ratios with expected number of
Monitoring data from 1992 indicated presence of TCE,
al. (1993)
specific cancer cases
cancers and site-specific cancers derived from
tetrachloroethylene and 1, 1,1,-trichloroethane in drinking water
(lymphoma sites and liver)
incidence of the Finnish population
supplies in largest towns in municipalities. Residence in town
from 1953-1991 in two
used to infer exposure to TCE.
Finnish municipalities.
Cohn et al.
Incident leukemia and
Logistic regression modeling adjusted for age
Monitoring data from 1984-1985 on TCE, THM, and VOCs
(1994)
NHL cases from
concentrations in public water supplies, and historical
Fagliano et
1979-1987 from 75
monitoring data conducted in 1978-1984.
al. (1990)
municipalities and
identified from the New
Jersey State Cancer
Registry. Histological type
was classified according to
WHO classification
scheme and the
classification of NIH
6/22/2009
231
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Description
Analysis Approach
Exposure Assessment
Working Formulation
Group was adopted to
grade NHL.
Mallin
Incident bladder cancer
Standardized incidence and mortality rates for bladder
Exposure data are lacking for the study population with the
(1990)
cases and deaths between
cancer by county of residence and zip code; expected
exception of noting one of two zip code areas with observed
1978-1985 among
numbers of bladder cancers using age-race-sex specific
elevated bladder cancer rates also had groundwater supplies
residents of 9 northwestern
incidence rates from SEER or bladder cancer mortality
contaminated with TCE, perchloroethylene and other solvents.
Illinois counties.
rates of the U.S. population from 1978-1985.
Isacson et al.
Incident bladder, breast,
Age-adjusted site-specific cancer incidence in Iowa
Monitoring data of drinking water at treatment plant in each
(1985)
prostate, colon, lung and
towns with populations of 1,000-10,000 and who were
Iowa municipality with populations of 1,000-10,000 used to
rectal cancer cases reported
serviced by a public drinking water supply
infer TCE and other volatile organic compound concentrations
to Iowa cancer registry
in finished drinking water supplies.
between 1969-1981
1
2
3
6/22/2009
232
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 4.0.4. Standards of Epidemiologic Study Design and Analysis Use for Evaluation
Category A: Study Design
Clear articulation of study objectives or hypothesis. The ideal is a clearly stated hypothesis or study objectives and the study is designed to achieve the identified objectives.
Selection and characterization in cohort studies of exposure and control groups and of cases and controls (case-control studies) is adequate. The ideal is for selection of cohort and referents from the
same underlying population and differences between these groups to be due to TCE exposure or level of TCE exposure and not to physiological, health status, or lifestyle factors. Controls or referents are
assumed to lack or to have background exposure to TCE. These factors may lead to a downward bias including one of which is known as "healthy worker bias," often introduced in analyses when mortality or
incidence rates from a large population such as the U.S. population are used to derive expected numbers of events. The ideal in case-control studies is cases and controls are derived from the same population
and are representative of all cases and controls in that population. Any differences between controls and cases are due to exposure to TCE itself and not to confounding factors related to both TCE exposure and
disease. Additionally, the ideal is for controls to be free of any disease related to TCE exposure. In this latter case, potential bias is toward the null hypothesis.
Category B: Endpoint Measured
Levels of health outcome assessed. Three levels of health outcomes are considered in assessing the human health risks associated with exposure to TCE: biomarkers of effects and susceptibility, morbidity,
and mortality. Both morbidity as enumerated by incidence and mortality as identified from death certificates are useful indicators in risk assessment for hazard identification. The ideal is for accurate and
predictive indicator of disease. Incidence rates are generally considered to provide an accurate indication of disease in a population and cancer incidence is generally enumerated with a high degree of accuracy
in cancer registries. Death certifications are readily available and have complete national coverage but diagnostic accuracy is reduced and can vary by specific diagnosis. Furthermore, diagnostic inaccuracies
can contribute to death certificates as a poor surrogate for disease incidence. Incidence, when obtained from population-based cancer registries, is preferred for identifying cancer hazards.
Changes in diagnostic coding systems for lymphoma, particularly non-Hodgkin's lymphoma. Classification of lymphomas today is based on morphologic, immunophenotypic, genotypic, and clinical
features and is based upon the World Health Organization (WHO) classification, introduced in 2001, and incorporation of WHO terminology into International Classification of Disease (ICD)-0-3. International
Classification of Disease (ICD) Versions 7 and earlier had rubrics for general types of lymphatic and hematopoietic cancer, but no categories for distinguishing specific types of cancers, such as acute leukemia.
Epidemiologic studies based on causes of deaths as coded using these older ICD classifications typically grouped together lymphatic neoplasms instead of examining individual types of cancer or specific cell
types. Before the use of immunophenotyping, these grouping of ambiguous diseases such as non-Hodgkin's lymphoma and Hodgkin's lymphoma may be have misclassified. With the introduction of ICD-10 in
1990, lymphatic tumors coding, starting in 1994 with the introduction of the Revised European-American Lymphoma classification, the basis of the current WHO classification, was more similar to that
presently used. Misclassification of specific types of cancer, if unrelated to exposure, would have attenuated estimate of relative risk and reduced statistical power to detect associations. When the outcome was
mortality, rather than incidence, misclassification would be greater because of the errors in the coding of underlying causes of death on death certificates (IOM, 2003). Older studies that combined all lymphatic
and hematopoietic neoplasms must be interpreted with care.
Category C: TCE-Exposure Criteria
Adequate characterization of exposure. The ideal is for TCE exposure potential known for each subject and quantitative assessment [j ob-exposure-matrix approach] of TCE exposure assessment for each
subject as a function of job title, year exposed, duration, and intensity. The assessment approach is accurate for assigning TCE intensity [TCE concentration or a time-weighted-average] to individual study
subjects and estimates of TCE intensity are validated using monitoring data from the time period. For the purpose of this report, the objective for cohort and case-controls studies is to differentiate TCE-exposed
subjects from subjects with little or no TCE exposure. A variety of dose metrics may be used to quantify or classify exposures for an epidemiologic study. They include precise summaries of quantitative
exposure, concentrations of biomarkers, cumulative exposure, and simple qualitative assessments of whether exposure occurred (yes or no). Each method has implicit assumptions and potential problems that
may lead to misclassification. Studies in which it was unclear that the study population was actually exposed to TCE are excluded from analysis.
Category D: Follow-up (Cohort)
Loss to follow-up. The ideal is complete follow-up of all subjects; however, this is not achievable in practice, but it seems reasonable to expect loss to follow-up not to exceed 10%. The bias from loss to
follow-up is indeterminate. Random loss may have less effect than if subjects who are not followed have some significant characteristics in common.
Follow-up period allows full latency period for over 50% of the cohort. The ideal to follow all study subjects until death. Short of the ideal, a sufficient follow-up period to allow for cancer induction period or
latency over 15 or 20 years is desired for a large percentage of cohort subjects.
Category E: Interview Type (Case-control)
6/22/2009
233
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Interview approach. The ideal interviewing technique is face-to-face by trained interviewers with more than 90% of interviews with cases and control subjects conduced face-to-face. The effect on the quality
of information from other types of data collection is unclear, but telephone interviews and mail-in questionnaires probably increase the rate of misclassification of subject information. The bias is toward the
null hypothesis if the proportion of interview by type is the same for case and control, and of indeterminate direction otherwise.
Blinded interviewer. The ideal is for the interviewer to be unaware whether the subject is among the cases or controls and the subject to be unaware of the purpose and intended use of the information
collected. Blinding of the interviewer is generally not possible in a face-to-face interview. In face-to-face and telephone interviews, potential bias may arise from the interviewer expects regarding the
relationship between exposure and cancer incidence. The potential for bias from face-to-face interviews is probably less than with mail-in interviews. Some studies have assigned exposure status in a blinded
manner using a job-exposure matrix and information collected in the unblinded interview. The potential for bias in this situation is probably less with this approach than for non-blinded assignment of exposure
status.
Category F: Proxy Respondents
Proxy respondents. The ideal is for data to be supplied by the subject because the subject generally would be expected to be the most reliable source; less than 10% of either total cases or total controls for
case-control studies. A subject may be either deceased or too ill to participate, however, making the use of proxy responses unavoidable if those subjects are to be included in the study. The direction and
magnitude of bias from use of proxies is unclear, and may be inconsistent across studies.
Category G: Sample Size
The ideal is for the sample size is large enough to provide sufficient statistical power to ensure that any elevation of effect in the exposure group, if present, would be found, and to ensure that the confidence
bounds placed on relative risk estimates can be well characterized.
Category H: Analysis Issues
Control for potentially confounding factors of importance in analysis. The ideal in cohort studies is to derive expected numbers of cases based on age-sex- and time-specific cancer rates in the referent
population and in case-control studies by matching on age and sex in the design and then adjusting for age in the analysis of data. Age and sex are likely correlated with exposure and are also risk factors for
cancer development. Similarly, other factors such as cigarette smoking and alcohol consumption are risk factors for several site-specific cancers reported as associative with TCE exposure. To be a confounder
of TCE, exposure to the other factor must be correlated, and the association of the factor with the site-specific cancer must be causal. The expect effect from controlling for confounders is to move the estimated
relative risk estimate closer to the true value.
Statistical methods are appropriate. The ideal is that conclusions are drawn from the application of statistical methods that are appropriate to the problem and accurately interpreted.
Evaluation of exposure-response. The ideal is an examination of a linear exposure-response as assessed with a quantitative exposure metric such as cumulative exposure. Some studies, absent quantitative
exposure metrics, examine exposure response relationships using a semi-quantitative exposure metric or by duration of exposure. A positive dose-response relationship is usually more convincing of an
association as causal than a simple excess of disease using TCE dose metric. However, a number of reasons have been identified for a lack of linear exposure-response finding and the failure to find such a
relationship mean little from an etiological viewpoint.
Documentation of results. The ideal is for analysis observations to be completely and clearly documented and discussed in the published paper, or provided in supplementary materials accompanying
publication.
6/22/2009
234
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
2 References
3
4 ADHS (Arizona Department of Health Services). 1990. The incidence of childhood leukemia
5 and testicular cancer in Pima County: 1970-1986. Prepared for the Arizona Department of
6 Environmental Quality under the Water Quality Assurance Revolving Fund and by the Arizona
7 Department of Health Services, Division of Disease Prevention, Office of Risk Assessment and
8 Investigation, Office of Chronic Disease Epidemiology. September 17, 1990.
9 ADHS (Arizona Department of Health Services). 1995. Update of the incidence of childhood
10 cancers and testicular cancer in Southwest Tucson, 1987 - 1991. Prepared by the Arizona
11 Department of Health Services, Office of Risk Assessment and Investigations, Disease
12 Prevention Services Arizona Department of Health Services. June, 6, 1995.
13 Aickin M. 2004. Bayes without priors. J Clin Epidemiol 57:4-13.
14 Anttila A, Pukkala E, Sallmen M, Hernberg S, Hemminki K. (1995) Cancer incidence among
15 Finnish workers exposed to halogenated hydrocarbons. J Occup Environ Med 37:797-806.
16 ATSDR (Agency for Toxic Substance and Disease Registry). 2004. Feasibility Investigation of
17 Worker Exposure to Trichloroethylene at the View-Master Factory in Beaverton, Oregon.
18 Prepared by the Oregon Department of Human Services. Report No. PB2005-102173.
19 ATSDR. (2006) Health Consultation: Cancer and Birth Outcome Analysis, Endicott Area, Town
20 of Union, Broome County, New York. U.S. Department of Health and Human Services, May
21 2006, Atlanta, GA.
22 ATSDR. (2008) Health statistics review follow-up: cancer and birth outcome analysis. Endicott
23 Area Investigation, Endicott Area, Town of Union, Broome County, New York. U.S.
24 Department of Health and Human Services, May 2008, Atlanta, GA.
25 Axelson O, Selden A, Andersson K, Hogstedt C. (1994). Updated and expanded 1 Swedish
26 cohort study on trichloroethylene and cancer risk. J Occup Environ 36:556-562.
27 Becker N, Deeg E, Meters A. 2004. Population-based study of lymphoma in Germany:
28 rationale, study design and first results. Leuk Res 28:713-724.
6/22/2009
235
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Blair A, Hartge P, Stewart PA, McAdams M, Lubin J. (1998) Mortality and cancer incidence of
2 aircraft maintenance workers exposed to trichloroethylene and other organic solvents and
3 chemicals: extended follow up. Occup Environ Med 55:161-171.
4 Boice JD, Marano DE, Fryzek JP, Sadler CJ, McLaughlin JK. (1999) Mortality among aircraft
5 manufacturing workers. Occup Environ Med 56:581-597.
6 Boice JD, Marano DE, Cohen SS, Mumma MR, Blott WJ, Brill AB, Fryzek JP, Henderson BE,
7 McLaughlin JK. (2006) Mortality among Rocketdyne worker who tested rocket engines, 1948-
8 1999. J Occup Environ Med 48:1070-1092.
9 Briining T, Pesch B, Wiesenhiitter B, Rabstein S, Lammert M, Baumiiller A, Bolt HM. 2003.
10 Renal cell cancer risk and occupational exposure to trichloroethylene: results of a consecutive
11 case-control study in Arnsberg, Germany. Am J Ind Med. 43:274-85.
12 Chang YM, Tai CF, Yang SC, Chen CJ, Shih TS, Lin RS, Liou SH. (2003) A cohort mortality
13 study of workers exposed to chlorinated organic solvents in Taiwan. Ann Epidemiol
14 13:652-660.
15 Chang Y-M, Tai C-F, Yang S-C, Lin R, Sung F-C, Shin T-S, Liou S-H. 2005. Cancer Incidence
16 among Workers Potentially Exposed to Chlorinated Solvents in An Electronics Factory. J Occup
17 Health 47:171-180.
18 Chang Y-M, Tai C-F, Yang S-C, Lin R, Sung F-C, Shin T-S, Liou S-H. 2005. Cancer Incidence
19 among Workers Potentially Exposed to Chlorinated Solvents in An Electronics Factory. J Occup
20 Health 47:171-180.
21 Chang Y-M, Tai C-F, Yang S-C, Lin R, Sung F-C, Shin T-S, Liou S-H. 2005. Cancer Incidence
22 among Workers Potentially Exposed to Chlorinated Solvents in An Electronics Factory. J Occup
23 Health 47:171-180.
24 Chang YM, Tai CF, Yang SC, Lin R, Sung FC, Shin TS, Liou SH. (2005) Cancer incidence
25 among workers potentially exposed to chlorinated solvents in an electronics factory. J Occup
26 Health 47:171-180.
27 Charbotel B, Fevotte J, Hours M, Martin JL, Bergeret A. (2006) Case-control study on renal cell
28 cancer and occupational exposure to trichloroethylene. Part II: Epidemiological aspects. Ann
29 Occup Hyg 50(8):777-87.
6/22/2009
236
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Cohn P, Klotz J, Bove F, Berkowitz M, Fagliano J. (1994) Drinking water contamination and the
2 incidence of leukemia and non-Hodgkin's lymphoma. Environ Health Perspect
3 102(6-7):556-61.
4 Costantini AS, Benvenuti A, Vineis P, Kriebel D, Tumino R, Ramazzotti V, et al. 2008. Risk of
5 leukemia and multiple myeloma associated with exposure to benzene and other organic solvents:
6 Evidence from the Italian Multicenter Case-control study. Am J Ind Med 51:803-811.
7 Costa G, Merletti F, Segnan N. 1989. A mortality cohort study in a north Italian aircraft factory.
8 Br J Ind Med 46:738-743.
9 Costas K, Knorr RS, Condon SK. (2002) A case-control study of childhood leukemia in Woburn,
10 Massachusetts: the relationship between leukemia incidence and exposure to public drinking
11 water. Sci Total Environ 300(l-3):23-35. Costas et al. 2002
12 Dosemeci, M; Cocco, P; Chow, WH. (1999) Gender differences in risk of renal cell carcinoma
13 and occupational exposures to chlorinated aliphatic hydrocarbons. Am J Ind Med 36(l):54-9.
14 Fagliano J, Berry M, Bove F, Burke T. (1990) Drinking water contamination and the incidence
15 of leukemia: an ecologic study. Am J Public Health 80(10): 1209-1212.
16 Garabrant DH, Held J, Langholz B, Bernstein L. (1988) Mortality of aircraft manufacturing
17 workers in southern California. Am J Ind Med 13:683-693.
18 Greenland S, Salvan A, Wegman DH, Hallock MF, Smith TH. (1994) A case-control study of
19 cancer mortality at the transformer-assembly facility. Int Arch Occup Environ Health 66:49-54.
20 Hansen J, Raaschou-Nielsen O, Christensen JM, Johansen I, McLaughlin JK, Lipworth L, Blot
21 WJ, Olsen JH. (2001) Cancer incidence among Danish workers exposed to trichloroethylene. J
22 Occup Environ Med 43:133-139.
23 Hardell L, Eriksson M, Lenner P, Lundgren E. 1981. Malignant lymphoma and exposure to
24 chemicals, especially organic solvents, chlorophenols and phenoxy acids: a case-control study.
25 Br J Cancer 43: 169-176.
26 Hardell L, Bengtsson NO, Johsson U, Eriksson S, Larsson LG. 1984. Aetiological aspects of
27 primary liver cancer with special regard to alcohol, organic solvents and acute intermittent
28 porphyria - an epidemiological investigation. Br J Cancer 50:389-297.
6/22/2009
237
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Henschler D, Vamvakas S, Lammert M, Dekant W, Kraus B, Thomas B, Ulm K. (1995)
2 Increased incidence of renal cell tumors in a cohort of cardboard workers exposed to
3 trichloroethene. Arch Toxicol 69:291-299.
4 IOM (Institute of Medicine). 2003. Gulf War and Health, Vol 2. Insecticides and Solvents.
5 Washington DC: The National Academies Press.
6 Isacson P, Bean JA, Splinter R, Olson DB, Kohler J. 1985. Drinking water and cancer incidence
7 in Iowa. III. Association of cancer with indices of contamination. Am J Epidemiol
8 121:856-869.
9 Krishnadasan A, Kennedy N, Zhao Y, Morgenstern H, Ritz B. (2007) Nested case-control study
10 of occupational chemical exposures and prostate cancer in aerospace and radiation workers. Am
11 J Ind Med 50:383-390.
12 Lee JH, Chung CW, Ma YC, Want GS, Chen PC, Hwang YH, et al. 2003. Increased mortality
13 odds ratio of male liver cancer in a community contaminated by chlorinated hydrocarbons in
14 groundwater. Occup Environ Med 60:364-369.
15 Lowengart RA, Peters JM, Cicioni C, Buckley J, Bernstein L, Preston-Martin S, Rappaport E.
16 (1987) Childhood leukemia and parents' occupational and home exposures. J Natl Cancer Inst
17 79(l):39-46.
18 MADPH (Massachusetts Department of Public Health). (1997) Woburn childhood leukemia
19 follow-up study. Volume 1: Analyses.
20 McKinney PA, Alexander FE, Cartwright RA, Parker L. (1991) Parental occupations of children
21 with leukaemia in west Cumbria, north Humberside, and Gateshead. BMJ 302(6778):681-7.
22 Mester B, Meters A, Deeg E, Eisner G, Becker N, Seidler A. 2006. Occupation and malignant
23 lymphoma: a population based case control study in Germany. Occup Environ Med 63:17-26.
24 Miligi L, Costantini AS, Benvenuti A, Kriebel D, Bolejack V, Tumino R, et al. 2006.
25 Occupational exposure to solvents and the risk of lymphomas. Epidemiology 17:552-561.
26 Morgan JW, Cassady RE. (2002) Community cancer assessment in response to long-time
27 exposure to perchlorate and trichloroethylene in drinking water. J Occup Environ Med.
28 44(7):616-621.
6/22/2009
238
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Morgan RW, Kelsh MA, Zhao K, Heringer S. (1998) Mortality of aerospace workers exposed to
2 trichloroethylene. Epidemiology 9:424-431.
3 Nordstrom M, Hardell L, Magnuson A, Hagberg H, Rask-Andersen A. 1998. Occupational
4 exposures, animal exposure and smoking as risk factors for hairy cell leukaemia evaluated in a
5 case-control study. Br J Cancer 77:2048-2052.
6 Persson B, Fredrikson M. 1999. Some risk factors for non-Hodgkin's lymphoma. Int J Occup
7 Med Environ Health 12:135-142.
8 Persson B, Dahlander A-M, Fredriksson M, Brage HN, Ohlson C-G, Alexson O. 1989.
9 Malignant lymphomas and occupational exposures. Br J Ind Med 46:516-520.
10 Persson B, Fredriksson M, Olsen K, Boeryd B, Axelson O. 1993. Some occupational exposure
11 as risk factors for malignant lymphomas. Cancer 72: 1773-1778.
12 Pesch B, Haerting J, Ranft U, Klimpel A, Oelschlagel B, Schill W. (2000) Occupational risk
13 factors for renal cell carcinoma: agent-specific results from a case-control study in Germany.
14 MURC Study Group. Multicenter urothelial and renal cancer study. Int J Epidemiol
15 29(6): 1014-24.
16 Raaschou-Nielsen O, Hansen J, McLaughlin JK, Kolstad H, Christensen JM, Tarone RE, Olsen
17 JH. (2003) Cancer risk among workers at Danish companies using trichloroethylene: a cohort
18 study. Am J Epidemiol 158:1182-1192.
19 Ritz B. (1999) Cancer mortality among workers exposed to chemicals during uranium
20 processing. J Occup Environ Med. 41:556-566.
21 Ritz B, Borgenstern H, Froines J, Young BB. 1999. Effects of exposure to external ionizing
22 radiation on cancer mortality in nuclear workers monitored for radiation at Rocketdyne/Atomics
23 International. Am J Ind Med 35:21-31.
24 Seidler A, Mohner M, Berger J, Mester B, Deeg E, Eisner G, et al. 2007. Solvent exposure and
25 malignant lymphoma: a population-based case-control study in Germany. J Occup Med Toxicol
26 2:2.
27 Shindell S, Ulrich S. 1985. A cohort study of employees of a manufacturing plant using
28 trichloroethylene. J Occup Med 27:577-579.
6/22/2009
239
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Shu XO, Stewart P, Wen WQ, Han D, Potter JD, Buckley JD, Heineman E, Robison LL. (1999)
2 Parental occupational exposure to hydrocarbons and risk of acute lymphocytic leukemia in
3 offspring. Cancer Epidemiol Biomarkers Prev 8(9):783-791.
4 Shu XO, Perentesis JP, Wen W, Buckley JD, Boyle E, Ross JA, et al. 2004. Parental exposure to
5 medications and hydrocarbons and ras mutations in children with acute lymphoblastic leukemia:
6 a report from the Children's Oncology Group. Cancer Epidemiol Biomarkers Prev
7 13:1230-1235.
8 Siemiatycki J. (1991) Risk factors for cancer in the workplace. Boca Raton: CRC Press.
9 Sinks T, Lushniak B, Haussler BJ, Sniezek J, Deng J-F, Roper P, Dill P, Coates R. (1992) Renal
10 cell cancer among paperboard printing workers. Epidemiol 3:483-489.
11 Sung TI, Chen PC, Jyuhn-Hsiarn Lee L, Lin YP, Hsieh GY, Wang JD. (2007) Increased
12 standardized incidence ratio of breast cancer in female electronics workers. BMC Public Health
13 7:102.
14 Sung TI, Wang JD, Chen PC. 2008. Increased risk of cancer in the offspring of female
15 electronics workers. Reprod Toxicol 25:115-119.
16 U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. Risk Assessment Forum,
17 Washington, DC; EPA/630/P-03/001B. Available from: .
19 Vamvakas S, Briining T, Thomasson B, Lammert M, Baumiiller A, Bolt HM et al. 1998. Renal
20 cell cancer correlated with occupational exposure to trichloroethene. J Cancer Res Clin Oncol
21 124:374-382.
22 Vartiainen T, Pukkala E, Rienoja T, Strandman T, Kaksonen K. (1993) Population exposure to
23 tri-and tetrachloroethene and cancer risk: two cases of drinking water pollution. Chemosphere
24 27:1171-1181.
25 Wilcosky TC, Checkoway H, Marshall EG, Tyroler HA. (1984) Cancer mortality and solvent
26 exposures in the rubber industry. Am Ind Hyg Assoc J 45(12):809-11.
27 Zhao Y, Krishnadasan A, Kennedy N, Morgenstern H, Ritz B. (2005) Estimated effects of
28 solvents and mineral oils on cancer incidence and Mortality in a cohort of aerospace workers.
29 Am J Ind Med 48:249-258.
6/22/2009
240
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
2
6/22/2009
241
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
4.1 Genetic toxicity
This section discusses the genotoxic potential of TCE and its metabolites. A summary is
provided at the end of each section for TCE or its metabolite for their mutagenic potential in
addition to an overall synthesis summary at the end of the genotoxicity section.
The application of genotoxicity data to predict potential carcinogenicity is based on the
principle that genetic alterations are found in all cancers. Genotoxicity is the ability of chemicals
to alter the genetic material in a manner that permits changes to be transmitted during cell
division. Although most tests for mutagenicity detect changes in DNA or chromosomes, some
specific modifications of the epigenome including proteins associated with DNA or RNA, can
also cause transmissible changes. Genetic alterations can occur through a variety of mechanisms
including gene mutations, deletions, translocations, or amplification; evidence of mutagenesis
provides mechanistic support for the inference of potential for carcinogenicity in humans.
TCE and its known metabolites trichloroacetic acid (TCA), dichloroacetic acid (DCA),
chloral hydrate (CH), trichloroethanol (TCOH), S-(l,2-dichlorovinyl)-L-Cysteine (DCVC) and
S-dichlorovinyl glutathione (DCVG) have been studied to varying degrees for their genotoxic
potential. The following section summarizes available data on genotoxicity for both TCE and its
metabolites for each potential genotoxic endpoints, when available, in different organisms.
Changes that occur due to the modifications in the epigenome are discussed in endpoint-specific
sections 4.2-4.8.
4.1.1 TCE
4.1.1.1 DNA binding Studies
Covalent binding of TCE to exogenous DNA and protein in cell-free systems has been
studied by several investigators. Incubation of 14C-TCE with salmon sperm DNA in the
presence of microsomal preparations from B6C3F1 mice resulted in dose-related covalent
binding of TCE to DNA. The binding was enhanced when the microsomes were taken from
mice pretreated with phenobarbital, suggesting the binding may be related to an oxidative
metabolite, or when 1, 2-epoxy-3,3,3-trichloropropane, an inhibitor of epoxide hydralase, was
added to the incubations (Banerjee and Van Duuren, 1978). In addition, covalent binding of 14C-
TCE with microsomal proteins was detected after incubation with microsomal preparations from
mouse lung, liver, stomach and kidney and rat liver (Baneijee and Van Duuren, 1978).
Furthermore, incubation of 14C-TCE with calf thymus DNA in the presence of hepatic
6/22/2009
242
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
microsomes from phenobarbital-pretreated rats yielded in significant covalent binding (Di Renzo
etal., 1982).
To determine the metabolic profile and adduct formation in mouse and rat systems, the
roles of rat p450 isosymes and human liver microsomes in TCE metabolism was evaluated.
Miller and Guengerich (1983) used liver microsomes from control, b-naphthoflavone- and
phenobarbital-induced B6C3F1 mice, Osborne-Mendel rats and human liver microsomes.
Significant covalent binding of TCE metabolites to calf thymus DNA and proteins was observed
in all experiments. Phenobarbital treatment increased the formation of chloral and TCE oxide
formation, DNA and protein addcts. In contrast, b-naphthoflavone treatment did not induce the
formation of any microsomal metabolite suggesting that the forms of P450 induced by
phenobarbital are primarily involved in TCE metabolism while the b-naphthoflavone-inducible
forms of P450 have only a minor role in TCe metabolism. TCE metabolism (based on TCE-
epoxide and DNA-adduct formation) was 2.5-3-fold higher in mouse than in rat microsomesdue
to differences in rates and clearance of metabolism (discussed in section 3.3.3.1) The levels of
DNA and protein adducts formed in human liver microsomal system approximated those
observed in liver microsomes prepared from untreated rats. It was also shown that whole
hepatocytes of both untreated mice and phenobarbital-induced rats and mice could activate TCE
into metabolites able to covalently bind exogenous DNA. A study by Cai and Guengerich
(2001) postulate TCE oxide (an intermediate in the oxidative metabolism of TCE in rat and
mouse liver microsomes) to be responsible for the covalent binding of TCE with protein, and to a
much lesser extent, DNA. Mass spectrometry was used to analyze the reaction of TCE oxide
(synthesized by m-chloroperbenzoic acid treatment of TCE) with nucleosides, oligonucleotides
and protein to understand the transient nature of the inhibition of enzymes in the context of
adduct formation. Protein amino acid adducts were observed during the reaction of TCE oxide
with the model peptides. The majority of these adducts were unstable under physiological
conditions. Results using other peptides also indicate that adducts formed from the reaction of
TCE oxide with macromolecules and their biological effects are likely to be relatively short-
lived.
Studies have been conducted using in vitro and in vivo systems to understand the DNA
and protein binding capacity of TCE. Binding of TCE was observed in calf thymus DNA. In a
study in male mice, after repeated intraperitoneal (i.p.) injections of 14C-TCE, radioactivity was
detected in the DNA and RNA of all organs studied (kidney, liver, lung, spleen, pancreas, brain
and testis) (Bergman, 1983). However, in vivo labeling was shown to be due to metabolic
incorporation of CI fragments, particularly in guanine and adenine, rather than to DNA-adduct
6/22/2009
243
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
formation. In another study (Stott et al., 1982), following i.p injection of 14C-TCE in male
Sprague-Dawley rats (10-100 mg/kg) and B6C3F1 mice (10-250 mg/kg), high liver protein
labeling was observed while very low DNA labeling was detected. Stott et al. (1982) also
observed very low levels of DNA binding (0.62 ± 0.43 alkylation/106 nucleotides) in mice
administered 1,200 mg/kg or TCE, which is reportedly tumorigenic upon chronic administration.
In addition, a dose-dependent binding of TCE to hepatic DNA and protein at low doses in mice
was demonstrated by Kautiainen et al. (1997). In their dose-response study (doses between
2|ig/kg and 200 mg/kg b.w), the highest level of protein binding (2.4 ng/g protein) was observed
lh after the treatment followed by a rapid decline, indicating pronounced instability of the
adducts and/or rapid turnover of liver proteins. Highest binding of DNA (120pg/g DNA) was
found between 24 and 72h following treatment. Dose response curves were linear for both
protein and DNA binding. In this study, the data suggest that TCE does bind to DNA and
proteins in a dose-dependent fashion, however, the type and structure of adducts were not
determined.
TCE was covalently bound in vivo to DNA, RNA and proteins of rat and mouse organs
22h after i.p injection. Labeling of proteins from various organs of both species was higher than
that of DNA. In vitro, trichloroethylene was bioactivated by microsomal fractions dependent on
cytochrome P450, mainly from liver of both species, to intermediate(s) capable of binding to
exogenous DNA. No particular species-specific difference was evident except for mouse lung
microsomes, which were more efficient than rat lung microsomes. . This also supports the results
described by Miller and Guengerich (1983). The authors suggest some binding ability of TCE to
interact covalently with DNA (Mazzullo et al., 1992).
In summary, studies demonstrate that TCE can lead to binding to nucleic acids and
proteins, and that such binding is likely predicted on conversion to one or more reactive
metabolites (e.g., TCE oxide). For instance, increased binding was observed in samples
bioactivated with mouse and rat microsomal fractions. In most studies that compared DNA and
protein labeling, covalent binding of protein was higher than that of DNA, though the reasons for
this preferential binding have not been determined.
4.1.1.2 Bacterial systems — Gene mutations
Gene mutation studies (Ames assay) in various Salmonella strains of bacteria exposed to
TCE both in the presence and absence of stabilizing agent have been conducted by different
laboratories (Henschler et al., 1977; Simmon et al., 1977; Waskell, 1978; Baden et al., 1979;
6/22/2009
244
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Crebelli et al., 1982; Shimada et al., 1985; Mortelmans et al., 1986; McGregor et al., 1989)
(Table 4.1.1). It should be noted that these studies have tested TCE samples of different purities
using various experimental protocols. Inconsistent results were obtained in the presence and
absence of both stabilizing agents and metabolic activation system (S9).
Waskell (1978) studied the mutagenicity of several anesthetics and their metabolites.
Included in their study was trichloroethylene (and its metabolites) using Ames assay. The study
was conducted both in the presence and absence of S9 and caution was exercised to perform the
experiment under proper conditions (incubation of reaction mixture in sealed desiccator vials).
This study was performed in both TA98 and TA100 S. typhimurium strains at a dose range of
0.5-10% between 4 and 48h. No change in revertant colonies was observed in any of the doses
or time courses tested.
In other studies highly purified, epoxide free TCE samples were not mutagenic in
experiments with and without exogenous metabolic activation in Salmonella strain TA100 using
the plate incorporation assay (Henschler et al., 1977). Furthermore, no mutagenic activity was
found in several other strains including TA1535, TA1537, TA97, TA98, and TAlOOusing the
preincubation protocol (Mortelmans et al., 1986). Simmon et al. (1977) observed a less than 2-
fold but reproducible and dose-related increase in his+ revertants in plates inoculated with S.
typhimurium TA 100 and exposed to a purified, epoxide-free TCE sample. However, the authors
observed no mutagenic response in strain TA1535 with S9 mix and in both 1535 and TA100
without rat or mouse liver S9. Similar results were obtained by Baden et al. (1979), Bartsch et
al. (1979) and Crebelli et al. (1982). In all these studies purified, epoxide-free TCE samples
induced slight but reproducible and dose-related increases in his+ revertants in S. typhimurium
TA100 only in the presence of S9. No mutagenic activity was detected without exogenous
metabolic activation or when liver S9 from un-induced rats, mice and hamsters (Crebelli et al.,
1982) was used for activation.
Shimada et al. (1985) tested a low-stabilized, highly purified TCE sample in a modified
Ames reversion test using vapor exposure to S. typhimurium TA1535 and TA100. No mutagenic
activity was observed—both in the presence and absence of S9 mix. However, at the same doses
(1, 2.5 and 5% concentration), a sample of lower purity, containing undefined stabilizers, was
directly mutagenic in TA 100 (>4-fold) and TA1535 (>37 fold) at 5% concentration regardless
of the presence of S9 mix. . It should be noted that the doses used in this study resulted in
extensive killing of bacterial population, particularly at 5% concentration, more thatn 95%
toxicity was observed.
6/22/2009
245
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
A series of carefully controlled studies evaluating TCE (with and without stabilizers) was
conducted by McGregor et al. (1989). The authors tested high purity and oxirane-stabilized TCE
samples for their mutagenic potential in S. typhimurium strains TA 1535, TA 98 and TA 100.
Stabilized TCE was tested using a preincubation protocol up to a dose level of 10,000|ig/plate.
No mutagenic response was observed in either the presence or absence of metabolic activation
(S9) derived from Aroclor 1254-induced male rat liver. TCE without oxirane stabilizers also
was nonmutagenic when tested in a vapor delivery system. However, TCE containing 0.5-0.6%
1,2 epoxybutane induced mutagenic response in strains TA1535 and TA100 both in the presence
and absence of S9 mix. Epichlorohydrin (another commonly used stabilizer) also induced
increases in mutant frequency at a concentration of 0.0009%.
A study on Escherichia coli K12 strain was conducted by Greim et al. (1975) using
analytical-grade TCE samples. Revertants were scored at two loci: arg56, sensitive to base-pair
substitution and nadII3, reverted by frameshift mutagens. In addition, forward mutations to 5-
methyltryptophan resistance and galactose fermentation were selected. Approximately two-fold
increase in arg+ colonies was observed. No change in other sites was observed. No definitive
conclusion can be drawn from this study due to lack of information on reproducibility and dose-
dependence.
In addition to the above studies, the ability of TCE to induce gene mutations in bacterial
strains has been reviewed and summarized by several authors (Fahrig et al., 1995; Crebelli and
Carere, 1989; Douglas et al., 1999; Moore and Harrington-Brock, 2000; Clewell and Andersen,
2004). In summary, the results of adequately and carefully performed studies indicate pure TCE
is incapable of inducing point mutations in various strains of S. typhimurium tested either in the
presence or absence of a metabolic activation system. Therefore, TCE, in its pure form as a
parent compound is unlikely to induce induce point mutations However, in the presence of
stabilizers that are contained in most technical grade TCE, mutations were observed in some
studies. It is possible that mutations observed in response to exposure to technical grade TCE
may be contributed by the contaminants/impurities such as 1,2 epoxybutane and epichlorohydrin,
which are known mutagens.
4.1.1.3 Fungal systems - Gene Mutations, conversions and recombination
6/22/2009
246
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Gene mutations, conversions, and recombinations using fungal systems have been studied
to identify the effect of TCE in different strains of fungi and yeast systems.
Crebelli et al. (1985) studied the mutagenicity of TCE in Aspergillus nidulans both for
gene mutations and mitotic segregation. No increase in mutation frequency was observed when
Aspergillus was plated on selective medium and then exposed to TCE vapors. A small but
statistically significant increase in mutations was observed when conidia of cultures were grown
in the presence of TCE vapors and then plated on selective media. Since TCE required actively
growing cells to exerts its genetic activity and previous studies (Bignami et al., 1980) have
shown weak activity in the induction of methGl suppressors by trichloroethanol and chloral
hydrate, it is possible that endogenous metabolic conversion of TCE into trichloroethanol or
chloral hydrate may have been responsible for the positive response.
To understand the cytochrome P-450 mediated genetic activity of TCE, Callen et al.
(1980) conducted a study in two yeast strains (D7 and D) with different P-450 contents in their
log-phases. The D7 strain in it log-phase had a cytochromo P-450 concentration up to 5 times
higher than a similar cell suspension of D4 strain. Two different concentrations (15 and 22mM)
at two different time points (lh and 4h) were used in this study. A significant increase in
frequencies of mitotic gene conversion and recombination was observed at 15mM concentrations
at lh exposure period in the metabolically more active D7 strain, however the 22mM
concentration was highly cytotoxic (only 0.3% of the total number of colonies survived). No
changes were seen in D4 strain, suggesting that metabolic activation played an important role in
both genotoxicity and cytotoxicity. However, marginal or no genetic activity was observed
when incubation of cells and test compounds were continued for 4h in either strain, possibly
because of increased cytotoxicity, or a destruction of the metabolic system.
Koch et al. (1988) studied the genetic effects of chlorinated ethylenes including TCE in
the yeast Saccharomyces cerevisiae, strain D7 both in stationary-phase cells without S9,
stationary-phase cells with S9 and logarithmic-phase cells using different concentrations (11.1,
16.6 and 22.2 mM). No significant change in mitotic gene conversion or reverse mutation was
observed in either absence or presence of S9. There was an increase in the induction of mitotic
aneuploidy in Strain D61.M, though it was not statistically significant.
Rossi et al. (1983) studied the effect of TCE on yeast species S. pombe both using in vitro
and host mediated mutagenicity studies and the effect of two stabilizers, epichlorohydrin and
1,2-Epoxybutane that are contained in the technical grade of TCE. The main goal of this study
6/22/2009
247
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
was to evaluate genetic activity of TCE samples of different purity and if the effect is due to the
additives present in the TCE or TCE itself. The induction of forward mutations at five loci (ade
1,3,4,5,9) of the adenine pathway in the yeast, strain PI was evaluated. The stationary-phase
cells were exposed to 25mM concentration of TCE for 2, 4, and 8h in the presence and absence
of S9. No change in mutation frequency was observed both in pure-grade samples and technical-
grade samples either in the presence or absence of S9 and at any of the time-points tested. In a
following experiment, the same authors studied the effect of different concentrations (0.22, 2.2
and 22.0 mM) in a host mediated assay using liver microsome preparations obtained from
untreated mice, from PB-pretreated and NF-pretreated mice and rats. The results of that study are
described in section 4.1.1.4.1.).
Furthermore, TCE was tested for its ability to induce both point mutation and mitotic
gene conversion in diploid strain of yeast. S. cerevisiae (strain D7) both with and without a
mammalian microsomal activation system. In suspension test with D7, TCE was active only
with microsomal activation (Bronzetti et al., 1980).
These studies indicate that pure TCE is not likely to cause mutations, gene conversions,
or recombinations in fungal or yeast systems. The data suggest that the observed genotoxic
activity in these systems is predominantly mediated by either TCE metabolites or contaminants
used as stabilizers in technical grade TCE.
4.1.1.4 Mammalian Systems and Human
4.1.1.4.1 Gene Mutations
Very few studies have been conducted to identify the effect of TCE, particularly on gene
(point) mutations using mammalian systems (Table 4.1.3). Overall summary of different
endpoints using mammalian systems will be provided at the end of this section. In order to
assess the potential mutagenicity of TCE and its possible contaminants, Rossi et al. (1983)
performed genotoxicity tests using two different host mediated assays with pure- and technical-
grade TCE. Male mice were administered with one dose of 2g/kg of pure or technical grade
TCE by gavage. Following the dosing, for intraperitoneal host-mediated assay, yeast cell
suspensions (2 x 109 cells/mL) were inoculated into the peritoneal cavity of the animals.
Following 16h, animals were sacrificed and yeast cells were recovered to detect the induction of
forward mutations at five loci (ade 1, 2, 4, 5, 9) of the adenine pathway. A second host-mediated
assay was performed by exposing the animals to 2g/kg of pure or technical grade TCE and
inoculating the cells into the blood system. After 4h, yeast cells were recovered from livers. No
forward mutations in the loci indicated above were observed in host-mediated assay either by
6/22/2009
248
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
intrasanguineous or intraperitoneal treatment either with pure or technical-grade TCE. When the
mutagenic epoxide stabilizers were tested for mutagenicity independently or in combination, no
genotoxic activity was detected either at the concentrations evaluated. To confirm the sensitivity
of the assay, the authors tested N-nitroso-dimethyl-nitrosamine (NDMA; lmg/kg), a mutagen
and observed an increase in the mutation frequency to more than 20 times the spontaneous level.
These results on mutagenic activity of stabilizers contradict other in vitro studies where it is
shown that stabilizers play a role in induction of mutations in TCE-exposed cells containing
stabilizers. The authors assume that the negative result cold have been due to an inadequate
incubation time of the sample with the yeast cells.
Male and female transgenic lac Z mice were exposed by inhalation to an actual
concentrations of 0, 203, 1,153 and 3,141ppm TCE, 6 h per day for 12 days. Following 14 and
60 days of last exposure, animals were sacrificed and the mutation frequencies were determined
in bone marrow, kidney, spleen, liver, lung, and testicular germ cells. No gene mutations (base-
changes or small-deletions) were observed at any of the doses teste in male or female lung, liver,
bone marrow, spleen, and kidney, or in male testicular germ cells when the animals were
samples 60 days after exposure. In addition, gene mutations were not observed in the lungs at 14
days after the end of exposure.(Douglas et al., 1999). The authors acknowledge that lacZ
bacteriophage transgenic assay does not detect large deletions. The authors also acknowledge
that their hypothesis doe not readily explain the increases in small deletions and base-change
mutations found in the von Hippel-Lindau tumor suppressor gene in renal cell carcinomas of the
TCE-exposed population. DCA, a TCE metabolite has been shown to increase lacl mutations in
transgenic mouse liver, however, only after 60 weeks of exposure to high concentration
(>1000ppm) in drinking water (Leavitt et al., 1997). Considering the fact the DCA induced a
small increase in lac I mutations when the animals were exposed to drinking water in the Leavitt
et (1997) and that DCA is a minor metabolite, it is unlikely that DCA would have reached
sufficient tissue concentration to elicit the mutagenic effect in the this study (Douglas et al.,
1999).
4.1.1.4.2 von Hippel-Lindau (VHL) Gene Mutations
Studies have been conducted to determine the role of VHL gene mutations in renal cell
carcinoma, with and without TCE exposure, and are summarized here. Most of these studies are
epidemiologic, comparing VHL mutation frequencies of TCE-exposed to non-exposed cases
from renal cell carcinoma case-control studies, or to background mutation rates among other
renal cell carcinoma case series (described in Section 4.3.3). Inactivation of the VHL gene
through mutations, loss of heterozygosity and imprinting has been observed in about 70% of
6/22/2009
249
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
renal clear cell carcinomas (Alimov et al., 2000; Kenck et al., 1996). Recent studies have also
examined the role of other genes or pathways in renal cell carcinoma subtypes, including c-myc
activation and vascular endothelial growth factor (VEGF) (Furge et al., 2007; Toma et al., 2008).
Several studies have examined the role of VHL gene inactivation in renal cell carcinoma,
including a recent study that measured not only mutations but also promoter hypermethylation
(Nickerson et al., 2008). This study focused on kidney cancer regardless of cause, and found that
91% of cc-RCC exhibited alterations of the VHL gene, suggesting a role for VHL mutations as
an early event in cc-RCC. A recent analysis of current epidemiological studies of renal cell
cancer suggests VHL gene alterations as a marker of cc-RCC, but that limitations of previous
studies may make the results difficult to interpret (Chow and Devesa, 2008). Conflicting results
have been reported in epidemiological studies of VHL mutations in TCE-exposed cases and are
described in detail in Sec 4.4.2. Both Briining et al. (1997) and Brauch et al. (1999, 2004)
associated increased VHL mutation frequency in TCE-exposed renal cell carcinoma cases. The
two other available studies of Schraml et al. (1999) and Charbotel et al. (2007) because of their
limitations and lower mutation detection rate in the case of Charbotel et al. (2007) neither add
nor detract to the conclusions from the earlier studies. Additional discussion of these data is in
Section 4.3.3.
Limited animal studies have examined the role of TCE and VHL mutations, although
Mally et al. (2006) have recently conducted both in vitro and in vivo studies using the Eker rat
model (section 4.3.6.1.1 ). The Eker rat model (Tsc-2+/ ) is at increased risk for the development
of spontaneous renal cell carcinoma and as such has been used to understand the mechanisms of
renal carcinogenesis (Stemmer et al., 2007; Wolf et al., 2000). One study has demonstrated
similar pathway activation in Eker rats as that seen in humans with VHL mutations leading to
renal cell carcinoma, suggesting Tsc-2 inactivation is analogous to inactivation of VHL in human
renal cell carcinoma (Liu et al., 2003). In Mally et al. (2006), male rats carrying the Eker
mutation were exposed to TCE (0, 100, 250, 500, 1,000 mg/kg BW by gavage, 5 days a week)
for 13 weeks to determine the renal effects (additional data from this study on in vitro DCVC
exposure is discussed below, Section 4.1.5). Significant increase in labeling index in kidney
tubule cells was observed, however, no enhancement of preneoplastic lesions or tumor incidence
was found in Eker rat kidneys compared to controls. In addition, no VHL gene mutations in
exons 1-3 were detected in tumors obtained from either control or TCE-exposed Eker rats.
Although no other published studies have directly examined VHL mutations following exposure
to TCE, two studies performed mutational analysis of archived formalin-fixed paraffin embedded
tissues from renal carcinomas from previous rat studies. These carcinomas were induced by the
6/22/2009
250
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
genotoxic carcinogens potassium bromate (Shiao et al., 2002) or N-nitrosodimethylamine (Shiao
et al., 1998). Limited mutations in the VHL gene were observed in all samples, but, in both
studies, these were found only in the clear cell renal carcinomas. Limitations of these two
studies include the small number of total samples analyzed, as well as potential technical issues
with DNA extraction from archival samples (Sec 4.3.3). However, analyses of VHL mutations in
rats may not be informative as to the potential genotoxicity of TCE in humans because the VHL
gene may not be the target for nephrocarcinogenesis in rats to the extent that it appears to be in
humans.
4.1.1.4.3 Chromosomal Aberrations
A few studies were conducted to investigate the ability of TCE to induce chromosomal
aberrations in mammalian systems (Table 4.1.3). Galloway et al. (1987) studied the effect of
TCE on chromosome aberrations in Chinese hamster ovary cells. When the cells were exposed
to TCE (499-14,900|i/mL) for 2h with metabolic activation, no chromosomal aberrations were
observed. Furthermore, without metabolic activation, no changes in chromosomal aberrations
were found when the cells were exposed to TCE concentrations of 745-14,900 |ag/m L for
8-14h. It should be noted that in this study, liquid incubation method was used and the
experiment was part of a larger study to understand the genotoxic potential of 108 chemicals.
Three inhalation studies were performed using mice and rats exposed to different
concentrations of TCE to determine if TCE could induce cytogenetic damage (Kligerman et al.,
1994). In the first and second study, rats or mice respectively, were exposed to 0-, 5-, 500-, or
5,000-ppm TCE for 6 h. Peripheral blood lymphocytes in rats and splenocytes in mice were
analyzed for induction of chromosomal aberrations, sister chromatid exchanges and
micronucleus formation. The results of micronucleus and sister chromatid exchanges will be
discussed in the next sections (4.1.1.4.4 and 4.1.1.4.5) . No significant increase in chromosomal
aberrations was observed in binucleated peripheral blood lymphocytes. In the third study, the
authors exposed the same strain of rats for 6h per day over 4 consecutive days. No statistically
significant concentration-related increases in chromosomal aberrations were observed. Based on
the results of the above studies, TCE does not appear to cause chromosomal aberrations either in
in vitro or in vivo mammalian systems.
4.1.1.4.4 Micronucleus Induction
Micronucleus is another endpoint that can demonstrate the genotoxic effect of a
chemical. When appropriate methods are used to identify the micronucleus formation
(kinetochore positive or kinetochore negative), this assay can provide information about a
6/22/2009
251
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
chemical's mechanism of action, i.e., if a chemical causes direct DNA damage resulting from
strand breaks (clastogen) or indirect DNA damage (aneugen) resulting from spindle poison.
Several studies have been conducted to identify if TCE can cause micronucleus formation (Table
4.1.4).
Wang et al. (2001) investigated an in vitro model to evaluate vapor toxicity of TCE in
CHO-K1 cells. Cells were grown in culture media with an inner petri dish containing TCE that
would evaporate into the media containing cells. The concentration of TCE in cultured medium
was determined by gas chromatography. The actual concentration of TCE ranged from 0.8 and
1.4 ppm after a 24h treatment. The effect of TCE on micronucleus formation was measured. A
significant dose-dependent increase in micronucleus formation was observed. A dose-dependent
decrease in cell growth and cell number was also observed. The authors did not test if the
micronucleus formed was due to damage to the DNA or spindle formation.
Robbiano et al. (2004) conducted an in vitro study on DNA damage and micronucleus
formation in rat and human kidney cells exposed to six carcinogenic chemicals including TCE.
The authors examined for the ability of TCE to induce DNA fragmentation and formation of
micronuclei in primary cultures of rat and human kidney cells derived from kidney cancer
patients with l-4mM TCE concentrations. A significant dose-dependent increase in the
frequency of micronucleus was obtained in primary kidney cells from both male rats and human
of both genders.
In the same study, Robbiano et al. (2004) administered rats with a single oral dose of
TCE (3,591 mg/kg) corresponding to V2 LD50 which had been exposed to folic acid for 48h and
the rats were euthanized 48 h later following exposure to TCE. The frequency of binucleated
cells was taken as an index of kidney cell proliferation. A statistically significant increase in the
average frequency of micronucleus was observed. The authors acknowledge that the
significance of the results should be considered in light of the limitations including (1)
examination of TCE on cells from only three rats, (2) considerable variation in the frequency of
DNA lesions induced in the cells, and (3) the possibility that kidney cells derived from kidney
cancer patients may be more sensitive to DNA-damaging activity due to a more marked
expression of enzymes involved in the metabolic activation of kidney procarcinogens and
suppression of DNA repair processes. Never the less, this study is important and provides
information of the possible genotoxic effects of TCE.
6/22/2009
252
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Hu et al. (2008) studied the effect of TCE on micronucleus frequencies using human
hepatoma HepG2 cells. The cells were exposed to 0.5, 1, 2, and 4 mM TCE for 24h. TCE
caused a significant increase in micronucleus frequencies at all concentrations tested. It is
important to note that similar concentrations that were used in Robbiano et al. (2004).
As described in the chromosomal aberration section (section 4.1.1.4.3), inhalation studies
were performed using male mice and rats (Kligerman et al., 1994) to determine if TCE could
induce micronuclei. In the first and second study, rats or mice respectively, were exposed to 0-,
5-, 500-, or 5,000-ppm TCE for 6 h. Peripheral blood lymphocytes in rats and splenocytes in
mice were analyzed for induction of micronucleus formation. TCE caused a statistically
significant increase in micronucleus formation in rat bone marrow polychromatic erythrocytes at
all concentrations but not in mice. The authors note that TCE was significantly cytotoxic at the
highest concentration tested as determined by significant concentration-related decrease in the
ratio of PCEs/normochromatic erythrocytes. In the third study, to confirm the results of the first
study, the authors exposed rats to one dose of 5,000ppm for 6h. A statistical increase in bone
marrow micronucleus-PCEs was observed confirming the results of the first study.
Male CD1 mice were treated with TCE (457 mg/kg bw) for 30h. Bone marrow cells
were harvested for determination of micronucleus frequencies in PCEs. An increase in
micronucleus frequency at 30h after treatment was observed. Linear regression analysis showed
that micronucleus frequency induced by TCE correlated with trichloroethanol concentrations in
urine, a marker of TCE oxidative metabolism (Hrelia et al., 1994).
In summary, based on the results of the above studies, TCE is capable of inducing
micronuclei in different in vitro and in vivo systems tested. Since specific methods were not
used in most studies to identify if the micronucleus formed was due to DNA damage or spindle
poison, one cannot definitively identify the mechanism of micronucleus formation. However,
Kligerman et al. (1994) demonstrate micronucleus induction without the presence of
chromosomal aberrations, suggestive of spindle damage. Never the less, these are important
findings that indicate TCE has genotoxic potential as measured by the micronucleus formation.
4.1.1.4.5 Sister Chromatid Exchanges (SCEs)
Studies have been conducted to understand the ability of TCE to induce SCEs both in
vitro and in vivo systems (Table 4.1.4). White et al. (1979) evaluated the possible induction of
SCE in CHO using a vapor exposure procedure by exposing the cells to TCE (0.17%) for one
hour in the presence of metabolic activation. No change in SCE frequencies were observed
6/22/2009
253
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
between the control and the treatment group. However, in another study by Galloway et al.
(1987) a small but dose-related increase in SCE frequency in repeated experiments both with and
without metabolic activation was observed. It should be noted that in this study, liquid
incubation was used, and the exposure times were 25h without metabolic activation at a
concentration between 17.9 to 700 |ag/mL and 2h in the presence of S9 at a concentration of 49.7
to 14,900 |ig/mL, Due to the difference in the dose, length of exposure and treatment protocol
(vapor exposure vs. liquid incubation), no direct comparison can be made to explain the apparent
discrepancy. It should also be noted that inadequacy of dose selection and the absence of
positive control in the White et al. (1979) makes it difficult to interpret the study. In another
study (Gu et al., 1981a), a small but positive response was observed in assays with peripheral
lymphocytes.
No statistically significant increase in SCEs was found when male mice or CD rats were
exposed to TCE at concentrations of 5, 500, or 5,000 ppm for 6h (Kligerman et al., 1994).
Furthermore, to detect genotoxic effects of TCE on humans, SCEs were analyzed in lymphocytes
of 22 workers occupationally exposed to TCE and 22 matched controls. Although urinalysis in
the workers revealed their obvious exposure to TCE, no increase in SCE frequencies was found
in lymphocytes of the workers (Nagaya et al., 1989).
In summary, data are limited and insufficient to draw a conclusion on induction of SCEs
when exposed to TCE. No clear positive responses (although two studies have shown a small
increase in SCEs) have been observed in SCEs as a result of exposure to TCE either in vitro or in
vivo. It should be noted that direct comparison of these studies is difficult because several
different protocols, doses and time were used and lack of positive controls in some studies.
4.1.1.4.6 Unscheduled DNA Synthesis
Perocco and Prodi (1981) studied unscheduled DNA synthesis in human lymphocytes
cultured in vitro (Table 4.1.5). Three doses of TCE (2.5, 5.0, 10 |iL/mL) were used as final
concentrations with and without S9 mix. The results indicate that there was an increase in UDS
only in the presence of S9, and in addition, the increase was maximal at the TCE concentration
of 5|iL/mL, Three chlorinated ethane and ethylene solvent products were examined for their
genotoxicity in hepatocyte primary culture DNA repair assays using vapor phase exposures.
Unscheduled DNA synthesis or DNA repair was not observed in samples exposed to TCE
(Shimada et al., 1985). The abilities of chlorinated ethylenes including TCE to induce
unscheduled DNA synthesis were assessed in isolated hepatocytes using a method that does not
6/22/2009
254
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
require the blocking of semi-conservative DNA synthesis by Costa and Ivanetich (1984), who
reported that TCE induced unscheduled DNA synthesis. Based on the limited studies available,
no definitive conclusions can be made as to whether TCE causes unscheduled DNA synthesis.
4.1.1.4.7 DNA Strand Breaks
DNA damage in response to TCE exposure was studied using comet assay in human
hepatoma HepG2 cells (Hu et al., 2008; Table 4.1.5). The cells were exposed to 0.5, 1, 2, and 4
mM for 24h. TCE increased the DNA migration in a significant dose-dependent manner at all
tested concentrations suggesting TCE caused DNA strand breaks and chromosome damage.
TCE (4-10 mmol/kg body wt) were given to male mice by i.p. injection. The induction
of single-strand breaks (SSB) in DNA of liver, kidney, and lung was studied by the DNA
unwinding technique. There was a linear increase of the level of single strand breaks in kidney
and liver DNA but not in lung DNA 1 h after administration. The damage was completely
repaired 24 h after injection (Walles, 1986).
Robbiano et al. (2004) conducted an in vitro study on DNA damage in rat and human
kidney cells exposed to six carcinogenic chemicals including TCE. The authors examined the
ability of TCE to induce DNA fragmentation in primary cultures of rat and human kidney cells
with l-4mM TCE concentrations. TCE was dissolved in ethanol with a maximum concentration
of 0.3% and the rat cultures were exposed to 20h. Primary human kidney cells were isolated
from fragments of kidney discarded during the course of surgery for carcinoma of both male and
female donors with an average age of 64.2years. Significant dose-dependent increases in the
ratio of treated/control tail length (average 4-7compared to control) was observed as
measured by Comet assay in primary kidney cells from both male rats and human of both
genders.
Clay et al. (2008) studied the DNA damage inducing capacity of TCE using comet assay
in rat kidney proximal tubules. Rats were exposed by inhalation to a range of TCE
concentrations (500, 1,000, or 2,000ppm) for 6h per day for 5 days. TCE did not induce DNA
damage (as measured by tail length and percent tail DNA and tail movement) in rat kidney
proximal tubules in any of the doses tested possibly due to study limitations (small number of
animals tested (n=5) and limited exposure dimte (6h/day for only 5d)). These results are in
contrast to the findings of Robbiano et al. (2004) which showed DNA damage and increased
micronuclei in the rat kidney 20h following a single dose (3,591 mg/kg bw) of TCE. Therefore,
6/22/2009
255
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
based on the above studies, while several studies reported DNA damage induced by TCE in
vitro, the data from in vivo studies are limited for making definitive conclusions.
4.1.1.4.8 DNA damage related to oxidative stress
A detailed description of studies related to lipid peroxidation of TCE is presented in
conjunction with discussion of liver toxicity (Section 4.4). Here, studies resulting from oxidative
damage pertaining to genotoxicity are described. The involvement of lipid peroxidation in the
genotoxic properties of TCE was confirmed by using immunoperoxidase staining for 8-
hydroxydeoxyguanosine (8-OHdG) and by measuring levels of thiobarbituric acid-reactive
substances (TBARS) (Hu et al., 2008). To elucidate the role of glutathione (GSH) in these
effects, the intracellular GSH level was modulated by pre-treatment with buthionine-(S,R)-
sulfoximine (BSO), a specific GSH synthesis inhibitor, and by co-treatment with N-
acetylcysteine (NAC), a GSH precursor. It was found that depletion of GSH in HepG2 cells with
BSO dramatically increased the susceptibility of HepG2 cells to TCE-induced cytotoxicity and
DNA damage, while when the intracellular GSH content was elevated by NAC, the DNA
damage induced by TCE was almost completely prevented. These results indicate that TCE
exerts genotoxic effects in HepG2 cells, probably through DNA damage by oxidative stress, and
that GSH plays an important role in modulating that damage.
The time courses of lipid peroxidation, free radical generation, and 80HdG formation
were used to assess the level of oxidative stress in the liver of B6C3F1 mice dosed orally once
daily, 5 days a week for 8 weeks at 0, 400, 800, and 1,200 mg/kg TCE in corn oil. Lipid
peroxidation, as measured by TBARS, was significantly elevated at the two highest dose levels
of TCE on days 6 through 14 of the study. 80HdG levels were statistically significant in the
1,200 mg/kg/day group on days 2, 3, 10, 28, 49, and 56 only. The highest measured free radical
load, 307% of oil control, occurred at day 6. Therefore, TCE administration at these doses
appears to induce oxidative stress and DNA damage in mice (Channel et al., 1998).
Toraason et al. (1999) examined the potential for TCE to induce oxidative DNA damage
in rats that was detectable as increased urinary excretion of 80HdG. TBARS and 8-
epiprostaglandin F2alpha (8epiPGF) were also measured as biomarkers of increased oxidative
stress. Male Fischer rats were administered a single i.p. injection of 0, 100, 500, or 1,000 mg/kg
of TCE. Rats were sacrificed 24 h after dosing. In rats exposed to TCE, TBARS and the
80HdG/dG ratios were significantly elevated in liver athough they were not significantly
affected in lymphocytes. Results indicate that a single high dose of TCE, can increase oxidative
6/22/2009
256
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
DNA damage in rat liver. The authors however, acknowledge that the usefulness of 80HdG as a
biomarker of TCE-induced oxidative DNA damage is questionable.
In summary, based on the above studies, it appears that TCE is capable of inducing
oxidative damage via lipid peroxidation and lead to DNA adduct formation.
4.1.1.4.9 Cell Transformation:
In vitro cell transformation using BALB/c-3T3 cells was conducted using TCE with
concentrations varying from 0-250 |ig/mL in liquid phase exposed for 72h (Table 4.1.5). A dose
dependent increase in type III foci was observed although the magnitude of increase was
minimal and no statistical analysis was conducted. The response was considered positive but the
increase was small compared to other chlorinated hydrocarbons tested such 1,1,1-trichloroethane
(Tu et al., 1985). In another study by Amacher and Zelljadt (1983), no significant change in
morphological transformation was obtained when Syrian hamster embryo cells were exposed to
5, 10, or 25 |ig/mL of TCE. In this experiment, two different serums (horse serum and fetal
bovine serum) were tested to understand the importance of serum quality in the transformation
assay. No significant changes were seen in transformation colonies when tested in different
serum.
6/22/2009
257
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Table 4.1.1. TCE Genotoxicity: Bacterial Assays
Test System/Endpoint Doses tested With
activation
Salmonella typhimurium (TA100)
Salmonella typhimurium (TA 1535,
TA100)
Salmonella typhimurium (TA 98, TA100) 0.5-10%
0.1 —1 Oul
(epoxide-free)
1 -2.5% (epoxide- + (TA 100)
free) -(TA 1535)
Salmonella typhimurium (TA100,
TA1535)
Salmonella typhimurium (TA100)
1-3% (epoxide-
free)
5-20 % (v/v)
+ (TA100);
+/-
(TA1535)
Salmonella typhimurium (TA100)
Salmonella typhimurium (TA 1535, TA
100)
Salmonella typhimurium (TA 98, TA100,
TA1535, TA1537, TA97)
Salmonella typhimurium (TA98, TA100,
TA1535)
Salmonella typhimurium (TA98, TA100,
TA1535)
0.33-1.33% +
(epoxide-free)
1-5% (higherand -(higher
lower purity) purity)
+ (lower
purity)
10-1000uL/plate -
<10,000 jjg/plate
(unstabilized)
<10,000 jjg/plate +
(oxirane-
stabilized)
6/22/2009
258
Without Comments
activation
References
plate incorporation assay; Henschleret al., 1977
reverse mutation
Reverse mutations Simmon et al., 1977
reverse mutation assay, Waskell, 1978
the study was conducted
in sealed dessicator vials
reverse mutation assay Baden et al., 1979
negative under normal Bartsch et al., 1979
conditions, but 2 fold
increase in mutations in a
preincubation assay
reverse mutation Crebelli et al., 1982
reverse mutation assay, Shimada et al., 1985
extensive cytotoxicity
preincubation protocol Mortelmans et al., 1986
ND vapor assay McGregor et al., 1989
+ vaporassay McGregor et al., 1989
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Salmonella typhimurium <10,000 jjg/plate ND
(epoxybutane
stabilized)
Salmonella typhimurium <10,000 jjg/plate ND
(epichlorohydrin
stabilized)
Escherichia Coli (K12) 0.9 mM +
(analytical grade)
1 ND: Not determined; NA: Not applicable
2
3
4 Table 4.1.2. TCE Genotoxicity: Fungal and Yeast Systems
Test System/Endpoint Doses tested With
activation
Gene Conversions
S. cerevisiae D7 and D4 15and22mM;1h ND
and 4h
S. cerevisiae D7 11.1,16.6,22.2
mM
S. pombe 0.2 to 200 mM
("pure" and
technical grade)
S. cerevisiae D7
6/22/2009 259
preincubation assay McGregor et al., 1989
vapor assay McGregor et al., 1989
revertants at arg56 but Greim et al, 1975
not nad113 or other loci
Without
activation
Comments
References
+ at 1h, D7
strain;
- at 4h, both
D7 and D4
gene conversion;
P450 content 5-fold
greater in D7 strain;
high cytotoxicity at 22 mM
both stationary and log
phase/production of
phototropic colonies
forward mutation,
different experiments with
different doses and time
Callen et al., 1980
Koch et al., 1988
Rossi et al., 1983
Bronzetti et al., 1980
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Aspergillus nidulans
no data
forward mutation
Crebelli et al., 1985
Recombination
S. cerevisiae
S. cerevisiae D7 and D4
A. nidulans
15 and 22mM; 1 h ND
and 4h
ND
gene conversion
gene cross over
Bronzetti et al., 1980
Callen et al., 1980
Crebellii et al., 1985
Mitotic aneupioidy
S. cerevisiae D61 .M
5.5, 11.1, 16.6
mM
loss of dominant color Koch et al., 1988
homolog
1
2
3
4
5
6
ND: Not determined; NA: Not applicable
Table 4.1.3. TCE Genotoxicity: Mammalian Systems - Gene mutations and chromosome aberrations
Test System/Endpoint Doses tested With Without Comments
activation activation
Gene Mutations (Forward Mutations)
Schizosaccharomyces pombe
2g/kg, 4h and
16h
ND
host-mediated:
intravenous and
intraperitoneal injections
of yeast cells
References
Rossi et al., 1983
Gene Mutations (Mutations Frequency)
lac Z transgenic mice
0,203,1,153, No base No base lung, liver, bone marrow, Douglas et al., 1999
3,141 ppm changes or changes or spleen, kidney, testicular
6/22/2009
260
-------
1
2
3
4
5
6
7
8
9
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
small
deletions
small
deletions
germ cells used
Chromosomal aberration
CHO
745-14,900 ND
ug/mL
499-14,900
ug/mL
5, 50, 500, 5,000
ppm (6 hr)
5, 50, 500, 5,000
ppm (6 hr, single
8-14h
ND
2h exposure
C57BL/6J mice
NA
splenocytes
CD rats
NA
peripheral blood
lymphocytes
and 4-day
exposure)
ND: Not determined; NA: Not applicable
Table 4.1.4. TCE Genotoxicity: Mammalian Systems - Micronucleus, sister chromatic exchanges
Test System/Endpoint Doses tested With Without Comments
activation activation
Micronucleus
Human hepatoma HepG2 cells
Primary cultures of human and rat
kidney cells
0.5-4 mM, 24 hr NA
1.0, 2.0, 4.0 mM NA
+
+
dose-dependent
significant increase
6/22/2009
261
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Sprague-Dawley rats
CHO-K1 cells
Male CD1 mice
C56BL/6J mice
CD rats
3,591 mg/kg +
0.8-1.4 ppm
457 mg/kg +
5, 50, 500, 5,000
ppm
5, 50, 500, 5,000 +
ppm
NA
NA
NA
dose-dependent
significant increase
bone marrow, correlated
with TCOH in urine
splenocytes
dose dependent;
peripheral blood
lymphocytes
Robbiano et al., 2004
Wang et al., 2001
Hrelin et al., 1994
Kligerman et al., 1994
Kligerman et al., 1994
Sister Chromatid Exchanges
CHO
CHO
CHO
Human lymphocytes
CD rats
Peripheral blood lymphocytes from
humans occupational^ exposed
C57BL/6J mice
0.17%
17.9-700ug/mL
49.7-14,900
ug/mL
178ug/mL
5, 50, 500, 5,000
ppm
occupational
exposure
5, 50, 500, 5,000
ppm
ND
+
ND
ND
+
ND
NA
NA
NA
1 hr (vapor)
25 hr (liquid)
2 hr
peripheral blood
lymphocytes
splenocytes
White et al., 1979
Galloway et al., 1987
Galloway et al., 1987
Gu et al., 1981a, b
Kligerman et al., 1994
Nagaya et al., 1989
Kligerman et al., 1994
1
2
3
4
5
6
7
ND: Not determined; NA: Not applicable
Table 4.1.5. TCE Genotoxicity: Mammalian Systems - unscheduled DNA synthesis, DNA strand breaks/protein crosslinks,
cell transformation
6/22/2009
262
-------
INTER-AGENCY REVIEW DRAFT
- DO NOT CITE OR QUOTE
Test System/Endpoint
Doses tested
With
Without
Comments
References
activation
activation
Unscheduled DNA synthesis
rat primary hepatocytes
ND
-
Shimada et al., 1985
human lymphocytes
2.5, 5, 10 uL/mL
+/-
-
increase was only in
Perocco and Prodi,
certain doses and
1981
maximum at 5 uL/mL
cone
Human WI-38
+
+
Beliles et al., 1980
phenobarbitol induced rat hepatocytes
2.8 mM
ND
+
Costa and Ivanetich,
1984
DNA strand breaks/protein crosslinks
Primary rat kidney cells
0.5, 1.0, 2.0, 4.0
NA
+
Dose dependent
Robbiano et al., 2004
mM
significant increase
Primary cultures of human kidney cells
1.0, 2.0, 4.0 mM
ND
+
dose dependent
Robbiano et al., 2004
significant increase
Sprague-Dawley rats
3,591 mg/kg
+
NA
single p.o. administration
Robbiano et al., 2004
Sprague-Dawley male CD rats
500, 1,000, 2,000
-
NA
Comet assay
Clay, 2008
ppm
Cell transformation
BALB/c3T3 mouse cells
4, 20, 100, 250
NA
+
weakly positive compared
Tu et al., 1985
ug/mL
to other halogenated
compounds tested in the
same experiment
Rat embryo cells
NA
+
Price et al., 1978
Syrian hamster embryo cells
5, 10 25 ug/mL
NA
-
Amacher and Zelljadt,
1983
1 ND: Not determined; NA: Not applicable
6/22/2009
263
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1
2
6/22/2009
264
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
4.1.1.5 Summary
There are several challenges in interpreting the genotoxicity results obtained from TCE
exposure.
(a) Purity of the test substance. Many studies were conducted using technical grade
TCE that contains trace amounts of stabilizers such as 1,2 epoxybutane and
epichlorohydrin, which are known mutagens and thus may confound the results.
(b) Conditions under which the assay is performed. For example, because of the
volatility of TCE, proper precautions need to be taken to limit the evaporation of
TCE, such as the use of a closed sealed system.
(c) Use of appropriate enzyme activation system. For example, it is not clear if the
S9 fractions used in many studies contain adequate CYP, GST, GSH, etc. to
adequately recapitulate in vivo metabolism, such as generation of short-lived
intermediates including TCE-epoxide, dichloroacetyl chloride, and down-stream
GSH conjugation products.
(d) Type of the assay performed. For example, if micronucleus assay is performed
using two different methods, different mechanisms can be inferred such as
whether TCE is a clastogen (DNA damage caused due to breaks in the genome) or
an aneugen (numerical changes in the chromosome caused due to spindle
damage).
(e) Furthermore, several TCE studies have been conducted along with numerous
other chlorinated compounds and the results interpreted as a comparison of the
group of compounds tested (relative potency). However, for the purposes of
hazard characterization, such comparisons are not informative—particularly
because they are not necessarily correlated with in vivo carcinogenic potency.
Considering the above challenges when interpreting the genotoxicity data of TCE,
evidence from a number of different analyses and a number of different laboratories using
various genetic endpoints indicates that TCE has a limited potential to be genotoxic, but some
effects have been reported at toxicologically relevant concentrations.
A series of carefully controlled studies evaluating TCE itself (without mutagenic
stabilizers and without metabolic activation) found it to be incapable of inducing gene mutations
in standard mutation bacterial assays. There is some evidence that TCE or its metabolites bind to
6/22/2009
265
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
DNA and can induce single strand DNA breaks in both hepatic and kidney cells (as measured
using comet assay). However, the dose required to cause these DNA breaks was very high and
the response was low.
Data are limited with respect to in vitro mammalian test systems for several other genetic
endpoints. For instance, several studies have shown that TCE is capable of inducing oxidative
damage to DNA via lipid peroxidation. Studies of sister chromatid exchanges, chromosomal
aberrations, unscheduled DNA synthesis, DNA damage, and cell transformation do not indicate
consistent positive responses. More consistent genotoxicity results, however, have been reported
with respect to micronucleus formation. In particular, several in vitro and rodent in vivo
genotoxicity assays showed increased frequency of micronucleus with TCE treatment. Because
of the absence of chromosomal aberrations in one study, these findings may be indicative of
spindle effects rather than DNA damage, though data to make this distinction is lacking in most
studies. Importantly, several of the in vivo studies were conducted at toxicologically relevant
exposures, with effects seen at doses as low as 5 ppm in air for 6 hr (Kligerman et al., 1994).
Below, the genotoxicity data for TCE metabolites TCA, DCA, TCOH, Chloral Hydrate,
DCVC, and DCVG are briefly reviewed. The contributions of these data are two-fold. First, to
the extent that these metabolites may be formed in the in vitro and in vivo test systems for TCE,
they provide insight into what agent or agents may contribute to the limited activity observed
with TCE in these genotoxicity assays. Second, because the in vitro systems do not necessarily
fully recapitulate in vivo metabolism, the genotoxicity of the known in vivo metabolites
themselves provide data as to whether one may expect genotoxicity to contribute to the toxicity
of TCE following in vivo exposure.
4.1.2 TCA (Trichloroacetic Acid)
The TCE metabolite TCA has been studied using a variety of genotoxicity assay for its
genotoxic potential (see IARC [2004] for additional information).
4.1.2.1 Bacterial Systems — Gene Mutations
TCA has been evaluated in a number of in vitro test systems including the bacterial
assays (Ames) using different Salmonella Strains such as TA98, TA100, TA104, TA1535, and
RSJ100. The majority of these studies did not report positive findings for genotoxicity. Waskell
6/22/2009
266
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
(1978) studied the effect of TCA (0.45mg/plate) on bacterial strains TA98 and TA100 both in the
presence and absence of S9. The author did not find any revertants at the maximum non-toxic
dose tested. Following exposure to TCA, Rapson et al. (1980) reported no change in mutagenic
activity in strain TA100 in the absence of metabolic activation (S9). DeMarini et al. (1994)
performed different studies to evaluate the genotoxicity of TCA, including the Microscreen
prophage-induction assay (TCA concentrations 0 to 10 mg/mL) and use of the Salmonella TA
100 strain using bag vaporization technique (TCA concentrations 0-100 ppm), neither of which
yielded positive results. Nelson et al. (2001) reported no positive findings with TCA using a
Salmonella microsuspension bioassay (S. typhimurium strain TA104) following incubation of
TCA for various lengths of time, with or without rat cecal microbiota. Similarly, no activity was
observed in a study conducted by Kargalioglu et al. (2002) where Salmonella strains TA98,
TA100, and RSJ100 were exposed to TCA (0.1-100 mM) either in the presence or absence of S9
(Kargalioglu et al., 2002).
TCA was also negative in other bacterial systems. The SOS chromotest (which measures
DNA damage and induction of the SOS repair system) in Escherichia coli PQ37, +/- S9 (Giller
et al., 1997) evaluated the genotoxic activity of TCA ranging from 10 to 10,000 |ag/m L and did
not find any response. Similarly, TCA was not genotoxic in the Microscreen prophage-induction
assay in E. coli with TCA concentrations ranging from 0 to 10,000 |ig/mL, with and without S9
activation (DeMarini et al., 1994).
However, TCA induced a small increase in SOS DNA repair (an inducible error-prone
repair system) in S. typhimurium strain TA1535 in the presence of S9 (Ono et al., 1991).
Furthermore, Giller et al. (1997) reported that TCA demonstrated genotoxic activity in an Ames
fluctuation test in S. typhimurium TA100 in the absence of S9 at noncytotoxic concentrations
ranging from 1,750 to 2,250 |ig/mL. The addition of S9 decreased the genotoxic response, with
effects observed at 3,000-7,500 |ig/mL. Cytotoxic concentrations in the Ames fluctuation assay
were 2,500 and 10,000 |ig/mL without and with microsomal activation, respectively.
4.1.2.2 Mammalian Systems
4.1.2.2.1 Gene Mutations
The mutagenicity of TCA has also been tested in cultured mammalian cells. Harrington-
Brock et al. (1998) examined the potential of TCA to induce mutations in L5178Y/TK+/" -3.7.2C
mouse lymphoma cells. In this study, mouse lymphoma cells were incubated in culture medium
6/22/2009
267
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
treated with TCA concentrations up to 2,150 |ag/m L in the presence of S9 metabolic activation
and up to 3,400 |ag/m L in the absence of S9 mixture. In the presence of S9, a doubling of mutant
frequency was seen at concentrations of 2,250 |ag/mL and higher, including several
concentrations with survival >10%. In the absence of S9, TCA increased the mutant frequency
by 2-fold or greater only at concentrations of 2,000 |ag/m L or higher. These results were
obtained at <11% survival rates. The authors noted that the mutants included both large-colony
and small-colony mutants. The small-colony mutants are indicative of chromosomal damage. It
should be noted that no rigorous statistical evaluation was conducted on these data.
4.1.2.2.2 Chromosomal Aberrations
Mackay et al. (1995) investigated the ability of TCA to induce chromosomal DNA
damage in an in vitro assay using cultured human cells. The authors treated the cells with TCA
as free acid, both in the presence and absence of metabolic activation. TCA induced
chromosomal damage in cultured human peripheral lymphocytes at concentrations (2,000 and
3,500 |ig/mL) that significantly reduced the pH of the medium. However, exposure of cells to
neutralized TCA did not have any effect even at a cytotoxic concentration of 5,000 |ig/mL. It is
possible that the reduced pH was responsible for the TCA-induced clastogenicity in this study.
To further evaluate the role of pH changes in the induction of chromosome damage, the authors
isolated liver-cell nuclei from B6C3Fi mice and suspended in a buffer at various pH levels. The
cells were stained with chromatin-reactive (fluorescein isothiocyanate) and DNA-reactive
(propidium iodide) fluorescent dyes. A decrease in chromatin staining intensity was observed
with the decrease in pH, suggesting that pH changes, independent of TCA exposure, can alter
chromatin conformation. It was concluded by the authors that TCA-induced pH changes are
likely to be responsible for the chromosomal damage induced by un-neutralized TCA. In
another in vitro study, Plewa et al. (2002) evaluated the induction of DNA strand breaks induced
by TCA (1-25 mM) in CHO cells and did not observe any genotoxicity.
4.1.2.2.3 Micronucleus
Relative genotoxicity of TCA was tested in a mouse in vivo system using three different
cytogenetic assay (bone marrow chromosomal aberrations, micronucleus and sperm-head
abnormalities) (Bhunya and Behera, 1987). TCA induced a variety of anomalies including
micronucleus in the bone marrow of mice. A small increase in the frequency of micronucleated
erythrocytes at 80 |ag/mL in a newt (Pleurodeles waltl larvae) micronucleus test was observed in
6/22/2009
268
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
response to TCA exposure (Giller et al., 1997). Mackay et al. (1995) investigated the ability of
TCA to induce chromosomal DNA damage in the in vivo bone-marrow micronucleus assay in
mice. C57BL mice were given TCA intraperitoneally at doses of 0, 337, 675, or 1,080 mg/kg-
day for males and 0, 405, 810, or 1,300 mg/kg-day for females for two consecutive days, and
bone-marrow samples were collected 6 and 24 hours after the last dose. The administered doses
represented 25, 50, and 80% of the median lethal dose, respectively. No treatment-related
increase in micronucleated polychromatic erythrocytes was observed.
4.1.2.2.4 DNA Damage
Studies on the ability of TCA to induce single-strand breaks have produced mixed results
(Chang et al., 1992; Styles et al., 1991; Nelson and Bull, 1988). Nelson and Bull (1988)
evaluated the ability of trichloroacetate and other compounds to induce single-strand DNA
breaks in vivo in Sprague-Dawley rats and B6C3Fi mice. Single oral doses were administered to
three groups of three animals, with an additional group as a vehicle control. Animals were
sacrificed after 4 hours, and 10% liver suspensions were analyzed for single-strand DNA breaks
by the alkaline unwinding assay. Dose-dependent increases in single-strand DNA breaks were
induced in both rats and mice, with mice being more susceptible than rats. The lowest dose of
TCA that produced significant SSBs was 0.6 mmol/kg (98 mg/kg) in rats but 0.006 mmol/kg
(0.98 mg/kg) in mice.
Styles et al. (1991) tested TCA for its ability to induce strand breaks in male B6C3Fi
mice in the presence and absence of liver growth induction. The test animals were given 1, 2, or
3 daily doses of neutralized TCA (500 mg/kg) by gavage and killed one hour after the final dose.
Additional mice were given a single 500-mg/kg gavage dose and sacrificed 24 hours after
treatment. Liver nuclei DNA were isolated, and the induction of single strand breaks was
evaluated using the alkaline unwinding assay. Exposure to TCA did not induce strand breaks
under the conditions tested in this assay. In a study by Chang et al. (1992), administration of
single oral doses of TCA (1 to 10 mmol/kg) to B6C3Fi mice did not induce DNA strand breaks
in a dose-related manner as determined by the alkaline unwinding assay. No genotoxic activity
(evidence for strand breakage) was detected in F344 rats administered by gavage up to 5
mmol/kg (817 mg/kg).
Studies have been conducted to evaluate the relationship between TCA-induced lipid
peroxidation and oxidative DNA damage (Austin et al., 1996; Parrish et al., 1996). In an acute
study by Austin et al. (1996), male B6C3Fi mice (six/group) were treated with a single oral dose
6/22/2009
269
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
of TCA (0, 30, 100, or 300 mg/kg), and 8-OHdG adducts were measured in liver DNA. A
significant increase in 8-OHdG levels was observed in the 300 mg/kg group at 8-10 hours post-
dosing. Parrish et al. (1996) expanded on this study by evaluating TCA-induced oxidative DNA
damage following repeated dosing. Male B6C3Fi mice (6/group) were exposed to 0, 100, 500,
or 2,000 mg/L TCA in drinking water for either 3 or 10 weeks (approximate doses of 0, 25, 125,
or 500 mg/kg-day). The levels of 8-OHdG levels were unchanged at both time periods. Thus,
oxidative damage to genomic DNA as measured by 8-OHdG adducts did not occur with
prolonged TCA treatment.
4.1.2.2.5 Cell Transformation
The initiating and promoting effects of TCA were investigated using a rat hepatic
enzyme-altered foci bioassay (Parnell et al., 1986). Twenty-four hours following partial
hepatectomy, rats either received a single oral dose (1,500 mg/kg) or 5,000 ppm TCA in drinking
water for 10, 20 or 30 days. Two weeks after the end of TCA exposure, the rats were promoted
for 3 or 6 months with 500 ppm Phenobarbital in drinking water. TCA failed to induce GGT-
positive foci using the initiation protocol. In the promotion protocol, TCA exposure resulted in a
significant increase in the number of GGT-positive foci. The authors indicate that the results
support the hypothesis that TCA may possess some promoting activity in the rat liver. Sprague-
Dawley rats were administered TCA by i.p injections and DNA was isolated from rat liver and
used to detect DNA damage of exon 7 of p53 gene (Yang and Heng, 2006). No change in the
p53 gene was observed in TCA treated rat livers DNA.
4.1.2.3 Summary
In summary, TCA has been studied using a variety of genotoxicity assays, including the
recommended battery. No mutagenicity was reported in S. typhimurium strains in the presence
or absence of metabolic activation or in an alternative protocol using a closed system, except in
one study on strain TA 100 using a modified protocol in liquid medium. This is largely
consistent with the results from TCE, which was negative in most bacterial systems except some
studies with the TA100 strain. Mutagenicity in mouse lymphoma cells was only induced at
cytotoxic concentrations. Measures of DNA-repair responses in bacterial systems have been
inconclusive, with induction of DNA repair reported in S. typhimurium but not in E. coli. TCA
induced oxidative DNA damage in the livers of mice following a single dose but not following
repeated dosing over 3 or 10 weeks. This is in contrast with TCE, which showed evidence of
6/22/2009
270
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
oxidative damage following both single and repeated in vivo exposures, and suggests minor, if
any, contribution from TCA to these effects. TCA-induced DNA strand breaks and chromosome
damage have been observed in in vivo but not in vitro although these effects have not been
uniformly reported, similar to the data from TCE. Furthermore, evidence suggests that TCA-
induced clastogenicity may be secondary to pH changes and not a direct effect of TCA. Finally,
a small number of micronucleus assays for TCA have shown inconsistent results, so the possible
contribution of TCA to the micronucleus activity of TCE is unclear.
4.1.3 Dichloroacetic Acid (DCA)
DCA is another metabolite of TCE that has been studied using a variety of genotoxicity
assay for its genotoxic potential (see IARC [2004] for additional information).
4.1.3.1 Bacterial and Fungal Systems — Gene Mutations
Studies were conducted to evaluate mutagenicity of DCA in different Salmonella strains
using Ames assay and E. coli (DeMarini et al., 1994; Giller et al., 1997; Waskell, 1978; Herbert
et al., 1980; Fox et al., 1996; Kargalioglu et al., 2002; Nelson et al., 2001; Fox et al., 1996).
DCA was mutagenic in three strains of S. typhimurium\ strain TA100 in three of five studies,
strain RSJ100 in a single study, and strain TA98 in two of three studies. DCA failed to induce
point mutations in other strains of S. typhimurium (TA 104, TA1535, TA1537, and TA1538) or
in E. coli strain WP2uvrA. In one study, DCA caused a weak induction of SOS repair in E. coli
strain PQ37 (Giller et al., 1997).
DeMarini et al. (1994), in the same study as described in TCA section of this chapter,
also studied DCA as one of their compounds for analysis. In the prophage-induction assay using
E. coli, DCA, in the presence of S9, was genotoxic producing 6.6-7.2 plaque-forming units
(PFU)/mM and slightly less than 3-fold increase in PFU/plate in the absence of S9. In the
second set of studies, which involved the evaluation of DCA at concentrations of 0-600 ppm for
mutagenicity in Salmonella TA100 strain, DCA was mutagenic both in the presence and absence
of S9, producing 3-5 times increases in the revertants/plate compared to the background. The
lowest effective concentration (LEC) for DCA without S9 was lOOppm and 50ppm in the
presence of S9. In the third and most important study, mutation spectra of DCA were
determined at the base-substitution allele hisG46 of Salmonella TA100. DCA induced revertants
were chosen for further molecular analysis at concentrations that produced mutant yields that
6/22/2009
271
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
were 2-5 fold greater than the background. The mutation spectra of DC A were significantly
different from the background mutation spectrum. Thus, despite the modest increase in the
mutant yields (3-5 times) produced by DC A, the mutation spectra confirm that DCA is
mutagenic. DCA primarily induced GC-AT transitions.
Kargalioglu et al. (2002) analyzed the cytotoxicity and mutagenicity of the drinking
water disinfection by-products (DBPs) including DCA in Salmonella typhimurium strains TA98,
TA100, and RSJ100 +/- S9. DCA was mutagenic in this test although the response was low
when compared to other DBPs tested in strain TA100. This study was also summarized in a
review by Plewa et al. (2002).Nelson et al. (2001) investigated the mutagenicity of DCA using a
Salmonella microsuspension bioassay following incubation of DCA for various lengths of time,
with or without rat cecal microbiota. No mutagenic activity was detected for DCA with S.
typhimurium strain TA104.
Although limited data, it appears that DCA has mutagenic activity in the Salmonella
strains, particularly TA 100 but not in other strains.
4.1.3.2 Mammalian Systems
4.1.3.2.1 Gene Mutations
Harrington-Brock et al. (1998) evaluated DCA for it mutagenic activity in L5178Y/TK
+/- (-) 3.7.2C mouse lymphoma cells. A dose-related increase in mutation (and cytotoxic)
frequency was observed at concentrations between 100 and 800 |ig/mL. Most mutagenic activity
of DCA at the Tk locus was due to the production of small-colony Tk mutants (indicating
chromosomal mutations). Precaution was taken to eliminate the role of pH (by testing different
pH) in induction of mutant frequencies and determined that the mutagenic effect observed was
due to the chemical and not pH effects.
Mutation frequencies were studied in male transgenic B6C3F1 mice harboring the
bacterial laclgene administered DCA at either 1.0 or 3.5g/L in drinking water (Leavitt et al.,
1997). No significant difference in mutant frequency was observed after 4 or 10 weeks of
treatment in both the doses tested as compared to control. However, at 60 weeks, mice treated
with l.Og/L DCA showed a slight increase (1.3-fold) in the mutant frequency over the control,
but mice treated with 3.5g/L DCA had a 2.3-fold increase in the mutant frequency. Mutational
spectra analysis revealed that -33% had G:C-A:T transitions and 21% had G:C - T:A
transversions and this mutation spectra was different than that was seen in the untreated animals,
indicating that the mutations were likely induced by the DCA treatment. The authors conclude
that these results are consistent with the previous observation that the proportion of mutations at
6/22/2009
272
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
T:A sites in codon 61 of the H-ras gene was increased in DCA-induced liver tumors in B6C3F1
mice (Leavitt et al., 1997).
4.1.3.2.2 Chromosomal Aberrations and Micronucleus
Harrington-Brock et al. (1998) evaluated DCA for its potential to induce chromosomal
aberrations in DCA-treated (0, 600 and 800 |ig/mL) mouse lymphoma cells. A clearly positive
induction of aberrations was observed at both concentrations tested. No significant increase in
micronucleus was observed in DCA-treated (0, 600 and 800 (ag/m L) mouse lymphoma cells
(Harrington-Brock et al., 1998).
Fuscoe et al. (1996) used the peripheral-blood-erythrocyte micronucleus assay (to detect
chromosome breakage and/or malsegregation) and the alkaline single cell gel electrophoresis
(SCG) technique to investigate the in vivo genotoxicity of DCA in bone marrow and blood
leukocytes, respectively. Mice were exposed to DCA in drinking water, available ad libitum, for
up to 31 weeks. The results indicate a small but statistically significant dose-related increase in
the frequency of micronucleated polychromatic erythrocytes (PCEs) after subchronic exposure to
DCA for 9 days. In addition, at the highest dose of DCA tested (3.5 g/L), a small but significant
increase in the frequency of micronucleated normochromatic erythrocytes (NCE) was detected
following exposure for > 10 weeks. The results indicated DNA cross-linking in blood
leukocytes in mice exposed to 3.5 g/L DCA for 28 days. These data provide evidence that DCA
may have some potential to induce chromosome damage when animals were exposed to
concentrations similar to those used in the rodent bioassay.
4.1.3.2.3 DNA Damage
In a series of experiments, male B6C3F1 mice and Sprague-Dawley rats treated with
DCA induced strand breaks in hepatic DNA in a dose-dependent manner in both species. Strand
breaks in DNA were observed at doses that produced no observable hepatotoxic effects as
measured by serum aspartate aminotransferase and alanine aminotransferase levels. The slopes
of the dose-response curves and the order of potency of the metabolites differed significantly
between rats and mice, suggesting that different mechanisms of single-strand break induction
may be involved in the two species (Nelson and Bull, 1988). Fuscoe et al. (1996), using single-
cell gel assay reported cross-linking in blood leukocytes in mice exposed to 3.5 g/L DCA for
28days.
6/22/2009
273
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
4.1.3.3 Summary
In summary, DCA has been studied using a variety but limited number of genotoxicity
assays. Within the available data, DCA has been demonstrated to be mutagenic in the
Salmonella assay, particularly in strain TA100, the in vitro mouse lymphoma assay and in vivo
cytogenetic and gene mutation assays. DCA can cause DNA strand breaks in mouse and rat liver
cells following in vivo administration by gavage. However, as with any in vitro or short term
studies, the concentration of DCA required to induce damage is high and the level of response is
generally low. Because of uncertainties as to the extent of DCA formed from TCE exposure,
inferences as to the possible contribution from DCA genotoxicity to TCE toxicity are difficult to
make.
4.1.4 Chloral Hydrate
Chloral hydrate has been evaluated for its genotoxic potential using a variety of
genotoxicity assays. These data are particularly important because it is known that a large flux
of TCE metabolism leads to chloral hydrate as an intermediate, so a comparison of their
genotoxicity profiles is likely to be highly informative.
4.1.4.1 DNA binding Studies
Limited analysis has been performed examining DNA binding potential of chloral
hydrate (Keller and Heck, 1988; Von Tungeln et al., 2002; Ni et al., 1995). Keller and Heck
(1988) conducted both in vitro and in vivo experiments using B6C3F1 mouse strain. The mice
were pretreated with l,500mg/kg TCE for 10 days and then given 800 mg/kg [14C] chloral. No
detectable covalent binding of 14C to DNA in the liver was observed. These results were
contradicted in another study with in vivo exposures to non-radioactive chloral hydrate at a
concentration of 1,000 and 2,000 nmol in mice B6C3F1 that demonstrated an increase in
malondialdehyde-derived and 8-oxo-2'-deoxyguanosine adducts in liver DNA (Von Tungeln et
al., 2002). Furthermore, while Keller and Heck (1988) observed no binding of chloral hydrate to
DNA in in vitro studies, Ni et al. (1995) observed malondialdehyde adducts in calf thymus DNA
when exposed to chloral hydrate and microsomes from male B6C3F1 mouse liver.
6/22/2009
274
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Mechanistic study was conducted to understand chloral toxicity in relationship to TCE
carcinogenesis (Keller and Heck, 1988). Chloral was investigated for its potential to form DNA-
protein cross-links in rat liver nuclei using concentrations 25, 100 or 250mM. No statistically
significant increase in percent interfacial DNA (IF DNA) containing DNA-protein cross-links
was observed. DNA and RNA isolated from the [14C] chloral-treated nuclei did not have any
detectable 14C bound. However, the proteins from choral-treated nuclei did have a
concentration-related binding of 14C.
4.1.4.2 Bacterial and Fungal Systems — Gene Mutations
Chloral hydrate induced gene mutations in S. typhimurium TA100 and TA104 strains, but
not in most other strains assayed. Four of six studies of chloral hydrate exposure in S.
typhimurium TA100 and two of two studies in S. typhimurium TA104 were positive for
revertants (Haworth et al., 1983; Ni et al., 1994; Giller et al., 1995; Beland, 1999). Waskell
(1978) studied the effect of chloral hydrate along with TCE and its other metabolites. Chloral
hydrate was tested at different doses (1.0-13mg/plate) in different S. typhimurium strains (TA
98, 100, 1535) for gene mutations using Ames assay. No revertant colonies were observed in
strains TA98 or 1535 both in the presence and absence of S9 mix. However, in TA 100, a dose-
dependent statistically significant increase in revertant colonies was obtained both in the
presence and absence of S9. It should be noted that chloral hydrate that was purchased from
Sigma was re-crystallized from one to six times from chloroform and the authors describe this as
crude chloral hydrate. Furthermore, Giller et al. (1995) studied chloral hydrate genotoxicity in
three short-term tests. Chloral induced mutations in strain TA100 of S. typhimurium (fluctuation
test). Similar results were obtained by Haworth et al. (1983). These are consistent with several
studies of TCE, in which low, but positive responses were observed in the TA100 strain in the
presence of S9 metabolic activation, even when genotoxic stabilizers were not present.
A significant increase in mitotic segregation was observed in Aspergillus nidulans when
exposed to 5 and lOmM chloral hydrate (Crebelli et al., 1985). Studies of mitotic crossing-over
in Aspergillus nidulans have been negative while these same studies were positive for
aneuploidy (Crebelli etal., 1985, 1991; Kafer, 1986; Kappas, 1989).
Limited analysis of chloral hydrate mutagenicity has been performed in Drosophila
(Zordan et al., 1994; Beland, 1999). Of these two studies, chloral hydrate was positive in the
6/22/2009
275
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
somatic mutation wing spot test (Zordan et al., 1994), equivocal in the induction of sex-linked
lethal mutation when in feed but negative when exposed via injection (Beland, 1999).
4.1.4.3 Mammalian Systems
4.1.4.3.1 Gene Mutations
Chloral hydrate induced concentration related cytotoxicity in TK+/- mouse lymphoma
cell lines without S9 activation. A non-statistical increase in mutant frequency was observed in
cells treated with chloral hydrate. The mutants were primarily small colony TK mutants,
indicating that most chloral hydrate induced mutants resulted from chromosomal mutations
rather than point mutations. It should be noted that in most concentrations tested (350-1,600
ug/mL), cytotoxicity was observed. Percent cell survival ranged from 96 to 4% (Harrington-
Brock, 1998).
4.1.4.3.2 Micronucleus Induction
Micronuclei induction following exposure to chloral hydrate is positive in most test
systems in both in vitro and in vivo assays, although some negative tests do also exist
(Harrington-Brock et al., 1998; Degrassi and Tanzarella, 1988; Beland, 1999; Lynch and Parry,
1993; Seelbach et al., 1993; Nesslany and Marzin, 1999; Russo and Levis, 1992a, b; Russo et al.,
1992; Leopardi et al., 1993; Allen et al., 1994; Nutley et al., 1996; Grawe et al., 1997; Giller et
al., 1995; Leuschner and Leuschner, 1991; Van Hummelen and Kirsch-Volders, 1992; Parry et
al., 1996; Bonatti et al., 1992; Ikbal et al., 2004).
Giller et al. (1995) studied chloral hydrate genotoxicity in three short-term tests. Chloral
hydrate caused a significant increase in the frequency of micronucleated erythrocytes following
in vivo exposure of the amphibian Pleurodeles waltl newt larvae.
Chloral hydrate induced aneuploidy in vitro in multiple Chinese hamster cell lines (Warr
et al., 1993; Furnus et al., 1990; Natarajan et al., 1993) and human lymphocytes (Vagnarelli et
al., 1990; Sbrana et al., 1993) but not mouse lymphoma cells (Harrington-Brock et al., 1998). In
vivo studies performed in various mouse strains led to increased aneuploidy in spermatocytes
(Russo et al., 1984; Liang and Pacchierotti, 1988; Miller and Adler, 1992) but not oocytes
(Mailhes et al., 1988) or bone marrow cells (Xu and Adler, 1990; Leopardi et al., 1993).
6/22/2009
276
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
The potential of chloral hydrate to induce aneuploidy in mammalian germ cells has been
of particular interest since Russo et al. (1984) first demonstrated that chloral hydrate treatment of
male mice results in significant increase in frequencies of hyperploidy in metaphase II cells.
This hyperploidy was thought to have arisen from chromosomal non-disjunction in
premeiotic/meiotic cell division and may be a consequence of chloral hydrate interfering with
spindle formation (reviewed by Russo et al. [1984] and Liang and Brinkley [1985]). Chloral
hydrate also causes meiotic delay, which may be associated with aneuploidy (Miller and Alder,
1992). Chloral hydrate has been shown to induce micronuclei but not structural chromosomal
aberrations in mouse bone-marrow cells. Micronucleus induced by non-clastogenic agents are
generally believed to represent intact chromosomes that failed to segregate into either daughter-
cell nucleus at cell division (Russo et al., 1992; Wang Xu and Adler, 1990). Furthermore,
chloral hydrate-induced micronuclei in mouse bone-marrow cells (Russo et al., 1992) and in
cultured mammalian cells (Degrassi and Tanzarella, 1988; Bonatti et al., 1992) have shown to be
predominantly kinetochore-positive in composition upon analysis with immunofluorescent
methods. The presence of a kinetochore in a micronucleus is considered evidence that the
micronucleus contains a whole chromosome lost at cell division (Degrassi and Tanzarella, 1988;
Hennig et al., 1988; Eastmond and Tucker, 1989). Therefore, both TCE and chloral hydrate
appear to increase the frequency of micronuclei.
Male C57B1/6J mice were given a single intraperitoneal injection of 0, 41, 83, or 165
mg/kg chloral hydrate. Spermatids were harvested at 22h, 11 days, 13.5 days, and 49 days
following exposure (Allen et al., 1994). Harvested spermatids were processed to identify both
kinetochore-positive micronucleus (aneugen) and kinetochore-negative micronucleus
(clastogen). All chloral hydrate doses administered 49 days prior to cell harvest were associated
with significantly increased frequencies of kinetochore-negative micronuclei in spermatids,
however, dose dependence was not observed. In animals treated for 13.5 days, only the 83
mg/kg dose caused a significant elevation of spermatid micronuclei. No increased frequencies of
were observed in animals treated with chloral hydrate for 11 days or 22h prior to spermatid
harvest (Allen et al., 1994). This study is in contrast with other studies (Degrassi and Tanzarella,
1988; Bonatti et al., 1992) who demonstrated predominantly kinetochore-positive micronucleus.
The ability of chloral hydrate to induce aneuploidy and polyploidy was tested in human
lymphocyte cultures established from blood samples obtained from two healthy non-smoking
donors (Sbrana et al., 1993). Cells were exposed for 72 and 96h at doses between 50 and
250|ig/mL, No increase in percent hyperdiploid, tetraploid, or endoreduplicated cells were
observed when cells were exposed to 72 h at any doses tested. However, at 96 h of exposure,
6/22/2009
277
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
significant increase in hyperdiploid was observed at one dose (150ug/mL) and was not dose
dependent. Significant increase in tetraploid was observed at dose 137mg/mL, again, no dose
dependence was observed.
Ikbal et al. (2004) assessed the genotoxic effects in cultured peripheral blood
lymphocytes of 18 infants (age range of 31-55days) before and after administration of a single
dose of chloral hydrate (50 mg/kg of body weight) for sedation before a hearing test for
micronucleus frequency. A significant increase in micronucleus frequency was observed after
administration of chloral hydrate. Although the authors indicate that the results were
significantly increased, the change in frequency was from 2.57 micronucleus/1,000 binucleate
(BN) cells before therapy to 3.56 micronucleus/1,000 BN cells.
4.1.4.3.3 Chromosomal Aberrations
Several studies have included chromosomal aberration analysis in both in vitro and in
vivo systems exposed to chloral hydrate and have resulted in positive in in vitro studies—
although not all studies had statistically significant increase (Furnus et al., 1990; Beland, 1999;
Harrington-Brock et al., 1998).
Analysis of chloral hydrate treated mouse lymphoma cell lines for chromosomal
aberrations resulted in a non-significant increase in chromosomal aberrations (Harrington-Brock
et al., 1998). However, it should be noted that the concentrations tested (1,250 and 1,300
|ig/mL) were cytotoxic (with a cell survival of 11 and 7%, respectively). Chinese hamster
embryo cells were also exposed to 0.001, 0.002, and 0.003% chloral hydrate for 1.5h (Furnus et
al., 1990). A non-statistically significant increase in frequency of chromosomal aberrations was
observed only at 0.002 and 0.003% concentrations, however the increase was not dose
dependent. In this study, it should be noted that the cells were only exposed for 1.5h to chloral
hydrate and cells were allowed to grow for 48h (two cell cycles) to obtain similar mitotic index
before analyzing for chromosomal aberrations.
In vivo studies have yielded mostly negative (Xu and Adler, 1990; Leuschner and
Leuschner, 1991; Russo and Levis, 1992a, b; Liang and Pacchierotti, 1988; Mailhes et al., 1993)
with the exception of one study (Russo et al., 1984) in an F1 cross of mouse strain between
C57B1/Cne X C3H/Cne. Hence, most studies suggest spindle effects rather than direct
clastogenicity when exposed to chloral hydrate.
6/22/2009
278
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
4.1.4.3.4 Sister Chromatid Exchanges
SCEs were assessed by Ikbal et al. (2004) in cultured peripheral blood lymphocytes of 18
infants (age range of 31-55days) before and after administration of a single dose of chloral
hydrate (50 mg/kg of body weight) for sedation before a hearing test for SCE frequencies.
Although the authors report a significant increase in SCEs, the average increase from before
administration (7.03 SCEs/cell) and after administration (7.90 SCEs/cell) was small. SCEs were
also assessed by Gu et al. (1981a) in human lymphocytes exposed in vitro with inconclusive
results, although positive results were observed by Beland (1999) in Chinese hamster ovary cells
exposed in vitro with and without an exogenous metabolic system.
4.1.4.3.5 DNA Damage
Single-strand DNA breaks were not observed in an in vitro assay in rat primary
hepatocytes (Chang et al., 1992). However, single-strand breaks (SSB) were observed both in
male Sprague-Dawley rat liver in vivo and male B6C3F1 mouse liver (Nelson and Bull, 1988).
4.1.4.3.6 Cell Transformation
Chloral hydrate was positive in the two studies designed to measure cellular
transformation (Gibson et al., 1995; Parry et al., 1996). Both studies exposed Syrian hamster
cells (embryo and dermal) to chloral hydrate and induced cellular transformation. Intercellular
communication was measured in two other studies, one positive for inhibition (Sprague-Dawley
rat liver cells) and one negative for inhibition (B6C3F1 mouse and Fisher 344 rat hepatocytes)
following in vitro exposure to CH (Klaunig et al., 1989; Benane et al., 1996).
4.1.4.4 Summary
Chloral hydrate has been shown to induce micronuclei formation, aneuploidy, and
mutations in multiple in vitro systems. In vivo studies have limited results to an increased
micronuclei formation mainly in mouse spermatocytes. CH is positive to in some studies in
in vitro genotoxicity assays that detect point mutations, micronuclei induction, chromosomal
aberrations, and/or aneuploidy. The in vivo data exhibit mixed results (Xu and Adler, 1990;
Russo et al., 1992; Mailhes et al., 1993; Allen et al., 1994; Alder, 1993; Nutley et al., 1996;
Leuschner and Beuscher, 1998). Most of the positive studies show that chloral hydrate induces
6/22/2009
279
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
aneuploidy rather than direct damage to DNA. Based on the existing array of data, CH has the
potential to be genotoxic, particularly when aneuploidy is considered in the weight of evidence
for genotoxic potential. Chloral hydrate appears to act through a mechanism of spindle
poisoning and resulting in numerical changes in the chromosomes, although some data suggest
induction of chromosomal aberrations. These results are consistent with TCE, albeit there are
more limited data on TCE for these genotoxic endpoints.
4.1.5 S-(l,2-dichlorovinyl)-L-Cysteine (DCVC) and S-dichlorovinyl glutathione (DCVG)
DCVC and DCVG have been studied for their genotoxic potential; however, since there
is limited number of studies to evaluate them based on each endpoint, particularly in mammalian
systems, the following section has been combined to include all the available studies for different
endpoints of genotoxicity. Study details can be found in Table 4.1.8.
DCVC and DCVG, cysteine intermediates of TCE formed by the GST pathway are
capable of inducing point mutations as evidenced by the fact that they are positive in the Ames
assay. Dekant et al. (1986) demonstrated mutagenicity of DCVC in S. typhimurium strains
(TA100, TA2638 and TA 98) using the Ames assay in the absence of S9. The effects were
decreased with the addition of a beta-lyase inhibitor aminooxyacetic acid, suggesting that
bioactivation by this enzyme plays a role in genotoxicity. Vamvakas et al. (1987) tested N-Ac-
DCVC for mutagenicity following addition of rat kidney cytosol and found genotoxic activity.
Furthermore, Vamvakas (1988a), in another experiment, investigated the mutagenicity of DCVG
and DCVC in Salmonella strain TA2638, using kidney subcellular fractions for metabolic
activation and AOAA (a beta-lyase inhibitor) to inhibit genotoxicity. DCVG and DCVC both
exhibited direct-acting mutagenicity, with kidney mitochondria, cytosol, or microsomes
enhancing the effects for both compounds and AOAA diminishing, but not abolishing the
effects. Importantly, addition of liver subcellular fractions did not enhance the mutagenicity of
DCVG, consistent with in situ metabolism playing a significant role in the genotoxicity of these
compounds in the kidney.
While additional data are not available on DCVG or NAc-DCVC, the genotoxicity of
DCVC is further supported by the predominantly positive results in other available in vitro and
in vivo assays. Jaffe et al. (1985) reported DNA strand breaks due to DCVC administered in
vivo, in isolated perfused kidneys, and in isolated proximal tubules of albino male rabbits.
Vamvakas et al. (1989) reported dose-dependent increases in unscheduled DNA synthesis in
6/22/2009
280
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
LLC-PK1 cell clones at concentrations without evidence of cytotoxicity. In addition, Vamvakas
et al. (1996) reported that 7-week DCVC exposure to LLC-PK1 cell clones at non-cytotoxic
concentrations induces morphological and biochemical de-differentiation that persists for at least
30 passages after removal of the compound. This study also reported increased expression of the
proto-oncogene c-fos in the cells in this system. In a Syrian hamster embryo fibroblast system,
DCVC did not induce micronuclei, but demonstrated an unscheduled DNA synthesis response
(Vamvakas et al., 1988b).
Two more recent studies are discussed in more detail. Mally et al. (2006) isolated
primary rat kidney epithelial cells from Tsc-2Ek/+ (Eker) rats, and reported increased
transformation when exposed to 10|iM DCVC, similar to that of the genotoxic renal carcinogens
N-methyl-N'-nitro-N-nitrosoguanidine (Horesovsky et al., 1994). The frequency was variable
but consistently higher than background. No loss-of-heterozygosity (LOH) of the Tsc-2 gene
was reported either in these DCVC transformants or in renal tumors (which were not increased in
incidence) from TCE-treated Eker rats, which Mally et al. (2006) suggested support a non-
genotoxic mechanism because a substantial fraction of spontaneous renal tumors in Eker rats
showed LOH at this locus (Kubo et al., 1994, Yeung et al., 1995) and because LOH was
exhibited both in vitro and in vivo with 2,3,4-tris(glutathion-S-yl)-hydroquinone treatment in
Eker rats (Yoon et al., 2001). However, 2,3,4-tris(glutathion-S-yl)-hydroquinone is not
genotoxic in standard mutagenicity assays (Yoon et al., 2001), and Kubo et al. (1994) also
reported that none of renal tumors induced by the genotoxic carcinogen N-ethyl-N-nitrosourea
showed LOH. Therefore, the lack of LOH at the Tsc-2 locus induced by DCVC in vitro, or TCE
in vivo, reported by Mally et al. (2006) is actually more similar to the response from the
genotoxic carcinogen N-ethyl-N-nitrosourea than the non-genotoxic carcinogen 2,3,4-
tris(glutathion-S-yl)-hydroquinone.
Finally, Clay (2008) evaluated the genotoxicity of DCVC in vivo using the comet assay
to assess DNA breakage in the proximal tubules of rat kidneys. Rats were exposed orally to a
single dose of DCVC (1 or 10 mg/kg). The animals were sacrificed either 2 or 16h after dosing
and samples prepared for detecting the DNA damage. DCVC (1 and lOmg/kg) induced no
significant DNA damage in rat kidney proximal tubules at the 16-h sampling time or after
lmg/kg DCVC at the 2-h sampling time. While Clay et al. (2008) concluded that these data
were insufficient to indicate a positive response in this assay, the study did report a statistically
significant increase in percent tail DNA 2 h after treatment with 10 mg/kg DCVC, despite the
small number of animals at each dose (n=5) and sampling time. Therefore, these data do not
substantially contradict the body of evidence on DCVC genotoxicity.
6/22/2009
281
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Overall, DCVC, and to a lesser degree DCVG and NAc-DCVC, have demonstrated
genotoxicity based on consistent results in a number of available studies. It is known that these
metabolites are formed in vivo following TCE exposure, specifically in the kidney, so they have
the potential to contribute to the genotoxicity of TCE, especially in that tissue. Moreover,
DCVC and DCVG genotoxic responses were enhanced when metabolic activation using kidney
subcellular fractions was used (Vamvakas et al., 1988a). Finally, the lack of similar responses in
genotoxicity assays with TCE, even with metabolic activation, is likely the result of the small
yield (if any) of DCVC under in vitro conditions, since in vivo, DCVC is likely formed
predominantly in situ in the kidney while S9 fractions are typically derived from the liver. This
hypothesis could be tested in experiments in which TCE is incubated with subcellular fractions
from the kidney, or from both the kidney and the liver (for enhanced GSH conjugation).
6/22/2009
282
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1 Table 4.1.6. TCE GSH conjugation metabolites genotoxicity
Test System/Endpoint Doses tested With Without Comments References
activation activation
Gene Mutations (Ames test)
S. typhimurium, TA100, 2638, 98 0.1-0.5 nmol ND + DCVC was mutagenic in Dekant et al., 1986
all three strains of S.
typhimurium without the
addition of mammalian
subcellular fractions
S. typhimurium, TA2638 50-300nmol + + Increase in number of Vamvakas et al., 1988a
revertants in DCVC alone
at low doses; further
increase in revertants
was observed in the
presence of microsomal
fractions. Toxicity as
indicated by decreased
revertants per plate were
seen at higher doses.
Mutation Analysis
In vitro - rat kidney epithelial cells, LOH 10uM
NA
only 1/9 transformed cells
Mally et al., 2006
in Tsc gene
showed LOH
In vitro - rat kidney epithelial cells, VHL 10uM
NA
No mutations in VHL
Mally et al., 2006
gene (exons1 -3)
aene. Note: VHL is not a
target gene in rodent
models of chemical-
induced or spontaneous
renal carcinogenesis
Unscheduled DNA synthesis
6/22/2009
283
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Porcine kidney tubular epithelial cell line
2.5uM-5, 10,
NA
+
Dose-dependent in UDS
Vamvakas et al., 1989
(LLC-PK1)
15,24h;
2.5uM-100uM
up to 24h tested at
2.5uM. Also, there was a
dose dependent increase
at lower conc. Higher
concentrations were
cytotoxic as determined
by LDH release from the
cells
Syrian hamster embryo fibroblasts
NA
+
Increase in UDS in
treatment groups
Vamvakas et al., 1988b
DNA strand breaks
Male rabbit renal tissue (perfused
0-100 mg/kg or
ND
+
Dose dependent increase
Jaffe et.al., 1985
kidneys and proximal tubules)
10uM to 10mM
SB in both iv and ip
injections (iv injections
were done only for 10
and 20 mg/kg).Perfusion
of rabbit kidney (45min
exposure) and proximal
tubules (30 min
exposure) expt. Resulted
in a dose dependent
difference in the amount
of single strand breaks
Primary kidney cells from both male rats
1-4mM; 20h
NA
+
Statistically significant
Robbiano, 2004
and human
exposure
increase in all doses
(1,2,4mM) both in rats
and human cells
In vivo - male Sprague-Dawley rats
TCE
+ (DCVC)
NA
No significant increase in
Clay, 2008
exposed to TCE or DCVC - comet
500-2,OOOppm,
- (TCE)
tail length in any of the
assay
inhalation, 6h per
TCE exposed groups. In
6/22/2009
284
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
day, 5 days
Expt. 1. 2h exposure -1
OR DCVC
or 10mg to DCVC
1 or 10mg/kg,
resulted in significant
single oral dose
increase with no dose
for 16h
response, but not at 16h.
In Expt. 2. ND for 1mg,
significant increase at
10mg
Micronucleus
Syrian hamster embryo fibroblasts
NA
"
No micronucleus
formation
Vamvakas et al., 1988b
Primary kidney cells from both male rats
1-4mM; 20h
NA
+
Statistically significant
Robbiano, 2004
and human
exposure
increase in all doses
(1,2,4mM) both in rats
and human cells
male Sprague-Dawley rats; proximal
4mmol/kg TCE
NA
+
Statistically significant
Robbiano et al., 1998
tubule cells (in vivo)
exposure, single
dose
increase in the average
frequency of
micronucleated kidney
cells was observed
Cell Transformation
Kidney tubular epithelial cell line (LLC-
1or 5 uM; 7
NA
+
Induced morphological
Vamvakas et al., 1996
PK1)
weeks
cell transformation at
both concentrations
tested. Furthermore,
cells maintained both
biochemical and
morphological alterations
remained stable for 30
passages
Rat kidney epithelial cells (in vitro)
10uM; 24h
NA
+
Cell transformation was
Mally et al., 2006
6/22/2009
285
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
exposure,
higher than control,
7weeks post
however cell survival
incubation
percent ranged from
39-64% indicating
cytotoxicity
Gene Expression
Kidney tubular epithelial cell line (LLC-
1 or 5 uM clones, NA
+
Increased c-fos Vamvakas et al., 1996
PK1)
30, 60, 90min
expression in land 5uM
exposed clones at three
different times tested
Kidney tubular epithelial cell line (LLC-
NA
+
Expression of c-fos and Vamvakas et al., 1993
PK1)
c-myc increased in a
time-dependent manner
ND: Not determined; NA: Not applicable
2
3
6/22/2009
286
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
4.1.6 Trichloroethanol (TCOH)
TCOH is negative in the Salmonella assay (Bignami et al., 1980; DeMarini et al., 1994).
SCEs were observed when human lymphocytes were exposed to trichloroethanol in vitro in
certain concentrations (Gu et al., 1981b). TCOH has not been evaluated in the other
recommended screening assays. Therefore, it is unclear whether TCOH is genotoxic.
4.1.7 Synthesis and Overall Summary
Trichloroethylene and its metabolites (TCA, DCA, CH, DCVC, DCVG, and TCOH)
have been evaluated for their genotoxic activity in several of in vitro systems such as bacteria,
yeast, and mammalian cells and, also, in in vivo systems (reviewed in ATSDR, 1997; IARC,
1995). Furthermore, a review of the mutagenicity of TCE contains a discussion of not only TCE
but also several of its metabolites such as TCA, DCA, chloral hydrate, DCVC and DCVG
(Moore and Harrington-Brock, 2000).
Due to the nature of TCE, it solubility and volatility, its metabolite(s) formation in vivo
and presence or absence of activation system and stabilizers, there are several challenges in
interpreting the genotoxicity results obtained from TCE exposure. For example, most studies
have been conducted using technical grade TCE which contains trace amounts of stabilizers such
as 1,2-epoxybutane and epichlorohydrin which are known mutagens. These stabilizers can
contribute to the results making interpretation of the data difficult with respect to the whether the
effect was caused by TCE exposure or the presence of stabilizers. Solubility and volatility of
TCE can be another factor. Because of the volatile nature of TCE, proper precautions should to
be taken to limit the evaporation of TCE, such as the use of a closed sealed system. If proper
care is not taken at this step of the experiment, then the results could be significant false
negatives. Use of inappropriate/inadequate enzyme activation system can also result in mis-
interpretation of the data. For example, it is not clear if the S9 fractions used in many studies
contain adequate amounts of CYP, GST, GSH, etc. to adequately recapitulate in vivo
metabolism, such as generation of short-lived intermediates including TCE epoxide,
dichloroacetyl chloride, and down-stream GSH conjugation products. Furthermore, the type of
the assay performed and the endpoint studied can greatly influence the conclusion. For instance,
bacterial mutation testing protocols typically specify the inclusion of cytotoxic concentrations of
the test compound, and the relative potency of the metabolites in vitro may not necessarily
6/22/2009
287
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
inform their relative contribution to the overall mechanistic effects of the parent compound,
TCE. This may be especially relevant when evaluating in vitro testing results for TCE, which
can undergo inter-organ metabolic processing involving multiple enzyme systems to yield highly
reactive species. Furthermore, if micronucleus assay is performed using two different methods,
different mechanisms can be inferred such as whether TCE is a clastogen (DNA damage caused
due to breaks in the genome) or an aneugen (numerical changes in the chromosome caused due
to spindle damage). In addition, such tests do not provide data for all effects that are relevant for
carcinogenesis. Also, type of samples used, methodology used for the isolation of genetic
material, and duration of exposure can particularly influence the results of several studies. This
is particularly true for human epidemiological studies. For example, while some studies use
tissues obtained directly from the patients others use formalin fixed tissues sections to isolate
DNA for mutation detection. Type of fixing solution, fixation time, and period of storage of the
tissue blocks often affect the quality of DNA. Formic acid contained in the formalin solution or
picric acid contained in Bouin's solution is known to degrade nucleic acids resulting in either
low yield or poor quality of DNA. In addition, during collection of tumor tissues, contamination
of neighboring normal tissue can easily occur if proper care is not exercised. This could lead to
the 'dilution effect' of the results, i.e., because of the presence of some normal tissue; frequency
of mutations detected in the tumor tissue can be lower than expected. Due to some of these
technical difficulties in obtaining proper material (DNA) for the detection of mutation, the results
of these studies should be interpreted cautiously. Furthermore, several TCE studies have been
conducted along with numerous other chlorinated compounds and the results interpreted as a
comparison of the group of compounds tested (relative potency). However, for the purposes of
hazard characterization, such comparisons are not informative—particularly because they are not
necessarily correlated with in vivo carcinogenic potency. Also, differentiating the effect of TCE
with respect to its potency can be influenced by many factors such as the type of cells, sensitivity
of the assay, need for greater concentration to show any effect, interpretation of data when the
effects are marginal, gradation of severity of the effects etc. Hence, caution should be exercised
when considering interpretations of genotoxicity data.
The following synthesis, summary, and conclusions focus on the available studies that
may provide some insight into the potential genotoxicity of TCE considering the above
challenges when interpreting the mutagenicity data for TCE.
Evidence from a number of different analyses and a number of different laboratories
using a fairly complete array of endpoints suggests that TCE and particularly its metabolites has
the potential to be genotoxic Based on a series of carefully controlled studies evaluating TCE
6/22/2009
288
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
itself (without mutagenic stabilizers and without metabolic activation) found it to be incapable of
inducing gene mutations in standard mutation bacterial assays (Waskell, 1978; Henschler et al.,
1977; Mortelmans et al., 1986; Simmon et al., 1977; Baden et al., 1979; Bartsch et al., 1979;
Crebelli et al., 1982, Shimada et al., 1985) except in TA 100 (Simmon et al., 1977). Therefore, it
appears that it is unlikely that TCE, as a pure compound, causes point mutations. In the presence
of stabilizers, which are contained in most technical grade TCE, mutations were observed in
some studies. It can be concluded that mutations observed in response to exposure to technical
grade TCE is probably contributed by the contaminants/impurities such as 1,2 epoxybutane and
epichlorohydrin which are known mutagens (McGregor et al., 1989, Rossi et al., 1983). In
fungal systems, no increase in mutation frequency was observed in some studies (Crebelli et al.,
1985; Koch et al., 1988, Rossi et al., 1983), however an increase in frequencies of mitotic gene
conversion and recombination was observed in some strains (Callen et al., 1980). Similar results
were obtained in mammalian systems (Rossi et al., 1983; Douglas et al., 1999). Data from
human epidemiological studies support the possible mutagenic effect of TCE leading to VHL
gene damage and subsequent occurrence of renal cell carcinoma in highly exposed population.
Association of increased VHL mutation frequency in TCE-exposed renal cell carcinoma cases
has been observed (Briining et al.,1997; Brauch et al., 1999, 2004).
Addition of enzyme systems capable of metabolizing TCE lead to a more relevant
response in genotoxicity tests. Studies have demonstrated that TCE can lead to binding to nucleic
acids and proteins (Di Renzo et al., 1982; Bergman, 1983; Miller and Guengerich, 1983;
Mazzullo et al., 1992; Kautiainen et al., 1997), and that such binding is likely predicted on
conversion to one or more reactive metabolites (e.g., TCE oxide). For instance, increased
binding was observed in samples bioactivated with mouse and rat microsomal fractions
(Baneijee and VanDuuren, 1978; Di Renzo et al., 1982; Miller and Guengerich, 1983; Mazzullo
et al., 1992). In most studies that compared DNA and protein labeling, covalent binding of
protein was higher than that of DNA, though the reasons for this preferential binding have not
been determined (Cai and Guengerich, 2001; Stott et al., 1982; Kautiainen et al., 1997).
TCE has also been shown to induce strand breaks (Robbiano et al., 2004; Hu et al., 2008)
but not in one study (Clay et al., 2008), oxidative damage via lipid peroxidation (Channel et al.,
1998; Toraason et al., 1999; Hu et al., 2008) and also causes micronuclei in different in vitro and
in vivo systems tested (Kligerman et al., 1994; Hrelia et al., 1994; Wang et al., 2001; Robbiano et
al., 2004; Hu et al., 2008). Since specific methods were not used in most studies to identify if the
micronucleus formed was due to DNA damage or spindle poison, one cannot definitively
identify the mechanism of micronucleus formation. However, Kligerman et al. (1994)
6/22/2009
289
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
demonstrate micronucleus induction without the presence of chromosomal aberrations that may
be indicative of spindle effects rather than DNA damage, though data to make this distinction is
lacking in most studies (Hrelia et al., 1994; Wang et al., 2001; Robbiano et al., 2004; Hu et al.,
2008). Nevertheless, these are important findings that indicate unmetabolized TCE has
genotoxic potential as measured by the micronucleus formation. On the contrary, TCE does not
appear to cause chromosomal aberrations either in in vitro or in vivo mammalian systems
(Galloway et al., 1987; Kligerman et al., 1994). Limited and insufficient data exists to draw a
conclusion on induction of SCEs as a result of exposure to TCE. No clear positive responses
have been observed in SCEs when exposed to TCE either in vitro or in vivo (White et al., 1979;
Gu et al., 1981a, b; Nagaya et al., 1989; Kligerman et al., 1994). It should be noted that direct
comparison of various studies is difficult because several different protocols, doses and times
were used and lack of positive controls in some studies. In addition, based on the limited studies
available, no definitive conclusions can be made as to whether TCE causes unscheduled DNA
synthesis (Perocco and Prodi, 1981; Costa and Ivanetich, 1984; Shimada et al., 1985), or cell
transformation (Amacher and Zelljadt, 1983; Tu et al., 1985).
TCA, an oxidative metabolite of TCE, exhibits little, if any genotoxic activity (Moore
and Harrington-Brock, 2000). TCA did not induce mutations in S. typhimurium strains in the
absence of metabolic activation or in an alternative protocol using a closed system (Waskell,
1978; Rapson et al., 1980; DeMarini et al., 1994; Giller et al., 1997; Nelson et al., 2001;
Kargalioglu et al., 2002) but a mutagenic response was induced in TA 100 in the Ames
fluctuation test (Giller et al., 1997). This is largely consistent with the results from TCE, which
was negative in most bacterial systems except some studies with the TA100 strain, but has not
been evaluated in the Ames fluctuation test. Mutagenicity in mouse lymphoma cells was only
induced at cytotoxic concentrations (Harrington-Brock et al., 1998). Measures of DNA-repair
responses in bacterial systems have been similarly inconclusive, with induction of DNA repair
reported in S. typhimurium but not in E. coli. TCA induced oxidative DNA damage in the livers
of mice following a single dose but not following repeated dosing over 3 or 10 weeks (Austin et
al., 1996; Parrish et al., 1996). This is in contrast with TCE, which showed evidence of
oxidative damage following both single and repeated in vivo exposures, and suggests minor, if
any, contribution from TCA to these effects. However, in vitro experiments with TCA should be
interpreted with caution if steps have not been taken to neutralize pH changes caused by the
compound. TCA was positive in some genotoxicity studies in vivo mouse and chick test systems
(Bhunya and Behera, 1987; Bhunya and Jena, 1996; Birner et al., 1994). TCA has been reported
to induce DNA SSB in hepatic DNA of mice (Nelson and Bull, 1988, 1989; Chang et al., 1992),
however other studies have failed to demonstrate this effect (Styles et al., 1991; Storer et al.,
6/22/2009
290
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1996). TCA-induced chromosomal damage has been observed in a few studies although these
effects have not been uniformly reported, similar to the data from TCE. Evidence suggests that
TCA-induced clastogenicity may be secondary to pH changes and not a direct effect of TCA
(Mackay et al., 1995). Finally, a small number of micronucleus assays for TCA have shown
inconsistent results (Bhunya and Behera, 1987; Giller et al., 1997, Mackay et al., 1995), so the
possible contribution of TCA to the micronucleus activity of TCE is unclear.
DCA, a chloroacid metabolite of TCE, has also been studied using different types of
genotoxicity assays. Although limited studies are conducted for different genetic endpoints,
DCA has been demonstrated to be mutagenic in the Salmonella assays, in vitro (DeMarini et al.,
1994; Kargalioglu et al., 2002; Plewa et al., 2002) in some strains, mouse lymphoma assay,
(Harrington-Brock et al., 1998) and in vivo cytogenetic (Leavitt et al., 1997; Fuscoe et al., 1996),
the micronucleus induction test, the Big Blue mouse system and other tests (Bignami et al., 1980;
Chang et al., 1989;DeMarini et al., 1994; Leavitt et al., 1997; Fuscoe et al., 1996; Nelson and
Bull, 1988; Harrington-Brock et al., 1998) in contrast to the parent compound, TCE. DCA can
cause DNA strand breaks in mouse and rat liver cells following in vivo mice and rats (Fuscoe et
al., 1996). However, with respect to in vitro or short-term studies, the concentration of DCA
required to induce damage is high and the level of response is generally low. Because of
uncertainties as to the extent of DCA formed from TCE exposure, inferences as to the possible
contribution from DCA genotoxicity to TCE toxicity are difficult to make.
Chloral hydrate is mutagenic in the standard battery of screening assays (Moore and
Harrington-Brock, 2000; Salmon et al., 1995). Effects include positive results in bacterial
mutation tests for point mutations and in the mouse lymphoma assay for mutagenicity at the Tk
locus (Haworth et al., 1983). In vitro tests showed that CH also induced micronuclei and
aneuploidy in human peripheral blood lymphocytes and Chinese hamster pulmonary cell lines.
Micronuclei were also induced in Chinese hamster embryonic fibroblasts. Several studies
demonstrate that chloral hydrate induces aneuploidy (loss or gain of whole chromosomes) in
both mitotic and meiotic cells, including yeast (Singh and Sinha, 1976, 1979; Kafer, 1986;
Gualandi, 1987; Sora and Agostini-Carbone, 1987), cultured mammalian somatic cells (Degrassi
and Tanzarella, 1988), and spermatocytes of mice (Russo et al., 1984; Liang and Pacchierotti,
1988). Chloral hydrate was negative for sex-linked recessive lethal mutations in drosophila
(Yoon et al., 1985). It induces SSB in hepatic DNA of mice and rats (Nelson and Bull, 1988)
and mitotic gene conversion in yeast (Bronzetti et al., 1984). Schatten and Chakrabarti (1998)
showed that chloral hydrate affects centrosome structure, which results in the inability to reform
normal microtubule formations and causes abnormal fertilization and mitosis of sea urchin
6/22/2009
291
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
embryos. Based on the existing array of data, CH has the potential to be genotoxic, particularly
when aneuploidy is considered in the weight of evidence for genotoxic potential. Chloral
hydrate appears to act through a mechanism of spindle poisoning and resulting in numerical
changes in the chromosomes. These results are consistent with TCE, albeit there are limited data
on TCE for these genotoxic endpoints.
DCVC, and to a lesser degree DCVG, has demonstrated bacterial mutagenicity based on
consistent results in a number of available studies (Dekant et al., 1986; Vamvakas et al., 1987;
Vamvakas, 1988a). DCVC has demonstrated a strong, direct-acting mutagenicity both with and
without the presence of mammalian activation enzymes. It is known that these metabolites are
formed in vivo following TCE exposure, so they have the potential to contribute to the
genotoxicity of TCE. The lack of similar response in bacterial assays with TCE is likely the
result of the small yield (if any) of DCVC under in vitro conditions, since in vivo, DCVC is
likely formed predominantly in situ in the kidney (S9 fractions are typically derived from the
liver). DCVC and DCVG have not been evaluated extensively in other genotoxicity assays, but
the available in vitro and in vivo data are predominantly positive. For instance, several studies
have reported the DCVC can induce primary DNA damage in mammalian cells in vitro and in
vivo (Jaffe et al., 1985; Vamvakas et al., 1989; Clay, 2008). Long-term exposure to DCVC
induced de-differentiation of cells (Vamavakas et al., 1996). It has been shown to induce
expression of the protooncogene c-fos (Vamvakas et al., 1996) and cause cell transformation in
rat kidney cells (Mally et al., 2006). In LLC-PK1 cell clones, DCVC was reported in induce
unscheduled DNA synthesis, but not micronuclei (Vamvakas et al., 1988b). Finally, DCVC
induced transformation in kidney epithelial cells isolated from Eker rats carrying the
heterozygous Tsc-2 mutations (Mally et al., 2006). Moreover, the lack of LOH at the Tsc-2 locus
observed in exposed cells does not constitute negative evidence of DCVC genotoxicity, as none
of renal tumors induced in Eker rats by the genotoxic carcinogen N-ethyl-N-nitrosourea showed
LOH (Kubo et al., 1994).
In support of the importance of metabolism, there is some concordance between effects
observed from TCE and those from several metabolites. For instance, both TCE and chloral
hydrate have been shown to induce micronucleus in mammalian systems, but chromosome
aberrations have been more consistently observed with chloral hydrate than with TCE. The role
of TCA in TCE genotoxicity is less clear, as there is less concordance between the results from
these two compounds. Finally, several other TCE metabolites show at least some genotoxic
activity, with the strongest data from DCA, DCVG, and DCVC. While quantitatively smaller in
6/22/2009
292
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1 terms of flux as compared to TCA and TCOH (for which there is almost no genotoxicity data),
2 these metabolites may still be toxicologically important.
3
4 Thus, uncertainties with regard to the characterization of TCE genotoxicity remain,
5 particularly because not all TCE metabolites have been sufficiently tested in the standard
6 genotoxicity screening battery to derive a comprehensive conclusion. However, the metabolites
7 that have been tested particularly DCVC have predominantly resulted in positive data although
8 to a lesser extent in DCVG and NAc-DCVC, supporting the conclusion that these compounds are
9 genotoxic, particularly in the kidney, where in situ metabolism produces and/or bioactivates
10 these TCE metabolites.
11 4.1.8 References:
12
13 Alder ID (1993) Synopsis of the in vivo results obtained with the 10 known or suspected
14 aneugens tested in the CEC collaborative study. Mutat Res 287(1): 131-137.
15
16 Allen JW, Collins BW, Evansky PA (1994) Spermatid micronucleus analyses of
17 trichloroethylene and chloral hydrate effects in mice. Mutat Res 323 (1-2): 81-88.
18
19 Alimov A, Kost-Alimova M, Liu J, Li C, Bergerheim U, Imreh S,
20 Klein G, Zabarovsky ER (2000) Combined LOH/CGH analysis proves the existence of
21 interstitial 3p deletions in renal cell carcinoma. Oncogene 19: 1392-1399.
22
23 Amacher DE, Zelljadt I. (1983). The morphological transformation of Syrian hamster embryo
24 cells by chemicals reportedly nonmutagenic to Salmonella typhimurium. Carcinogenesis 4:
25 291-295.
26 ATSDR (1997) Toxicological Profile for Trichloroethylene.
27 http://www.atsdr.cdc.gov/toxprofiles/tpl9.html
28 Austin, EW; Parrish, JM; Kinder, DH; et al. (1996) Lipid peroxidation and formation of 8-
29 hydroxydeoxyguanosine from acute doses of halogenated acetic acids. Fundam Appl Toxicol
30 31:77-82.
31
6/22/2009
293
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1 Baden JM, Kelley M, Mazze RI, et al. (1979). Mutagenicity of inhalation anaesthetics:
2 Trichloroethylene, divinyl ether, nitrous oxide, and cyclopropane. Br J Anaesth 51:417-421.
3
4 Beland FA. (1999) NTP technical report on the toxicity and metabolism studies of chloral
5 hydrate (CAS No. 302-17-0). Administered by gavage to F344/N rats and B6C3F1 mice.
6 Toxic Rep Ser. 59:1-66, A1-E7.
7
8 Baneijee S, Van Duuren BL (1978) Covalent binding of the carcinogen trichloroethylene to
9 hepatic microsomal proteins and to exogenous DNA in vitro. Cancer Res 38: 776-780.
10
11 Bartsch H, Malaveille C, Barbin A, Planche G. (1979) Mutagenic and alkylating metabolites of
12 halo-ethylenes, chlorobutadienes and dichlorobutenes produced by rodent or human liver tissues.
13 Evidence for oxirane formation by P450-linked microsomal mono-oxygenases. Arch Toxicol.
14 41(4):249-77.
15
16 Beliles RP, Brucik DJ, Mecler FJ, et al. (1980) Teratogenic-mutagenic risk of workplace
17 contaminants: Trichloroethylene, perchloroethylene, and carbon disulfide. U.S. Department of
18 Health, Education and Welfare. Contract No. 210-77-0047.
19
20 Benane SG, Blackman CF, House DE. (1996) Effect of perchloroethylene and its metabolites on
21 intercellular communication in clone 9 rat liver cells. J Toxicol Environ Health. 48(5):427-37.
22
23 Bhunya, SP; Behera, BC. (1987) Relative genotoxicity of trichloroacetic acid (TCA) as revealed
24 by different cytogenetic assays: bone marrow chromosome aberration, micronucleus and sperm-
25 head abnormality in the mouse. Mutat Res 188:215-221.
26
27 Bhunya SP, Jena GB.(1996) The evaluation of clastogenic potential of trichloroacetic acid
28 (TCA) in chick in vivo test system. Mutat Res. 367(4):254-9.
29
6/22/2009
294
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1 Bignami M, Conti G, Conti L, Crebelli R, Misuraca F, Puglia AM, Randazzo R, Sciandrello G,
2 Carere A. (1980) Mutagenicity of halogenated aliphatic hydrocarbons in Salmonella
3 typhimurium, Streptomyces coelicolor and Aspergillus nidulans. Chem Biol Interact.
4 30(l):9-23.
5
6 Birner G, Richling C, Henschler D, Anders MW, Dekant W. (1994) Metabolism of
7 tetrachloroethene in rats: identification of N epsilon-(dichloroacetyl)-L-lysine and N epsilon-
8 (trichloroacetyl)-L-lysine as protein adducts. Chem Res Toxicol. 7(6):724-32.
9
10 Bonatti S, Cavalieri Z, Viaggi S, Abbondandolo A. (1992) The analysis of 10 potential spindle
11 poisons for their ability to induce CREST-positive micronuclei in human diploid fibroblasts.
12 Mutagenesis. 7(2): 111-4
13
14 Bergman K (1983) Interaction of trichloroethylene with DNA in vitro and with RNA and DNA
15 of various mouse tissues in vivo. Arch Toxicol 54:181-193.
16
17 Brauch H, Weirich G, Hornauer MA, Storkel S, Wohl T, Briining T. (1999) Trichloroethylene
18 exposure and specific somatic mutations in patients with renal cell carcinoma. Natl Cancer Inst.
19 91(10):854—61.
20
21 Brauch H, Weirich G, Klein B, Rabstein S, Bolt HM, Briining T. (2004) VHL mutations in renal
22 cell cancer: does occupational exposure to trichloroethylene make a difference? Toxicol Lett.
23 151(1):301—10.
24
25 Bronzetti G, Corsi C, Nieri R. (1980) Use of the D7 strain of S. cerevisiae in the determination of
26 environmental risk. 1. Genetic effects of trichloroethylene. Boll Soc Ital Biol Sper.
27 56(12): 1315—21.
28
29 Bronzetti G, Galli A, Corsi C, Cundari E, Del Carratore R, Nieri R, Paolini M. (1984) Genetic
30 and biochemical investigation on chloral hydrate in vitro and in vivo. Mutat Res. 141(1): 19-22.
31
6/22/2009
295
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1 Briining T, Lammert M, Kempkes M, Thier R, Golka K, Bolt HM. (1997) Influence of
2 polymorphisms of GSTM1 and GSTT1 for risk of renal cell cancer in workers with long-term
3 high occupational exposure to trichloroethene. Arch Toxicol. 71(9):596-9.
4
5 Cai H, Guengerich P (2001) Reaction of trichloroethylene oxide with proteins and DNA:
6 instability of adducts and modulation of functions. Chem Res Toxicol 14:54-61.
7
8 Callen DF, Wolf CR, Philpot RM. (1980) Cytochrome P-450-mediated genetic activity and
9 cytotoxicity of seven halogenated aliphatic hydrocarbons in Succharomyces cerevisiae. Mutat
10 Res 77:55-63.
11
12 Channel SR, Latendresse JR, Kidney JK, Grabau JH, Lane JW, Steel-Goodwin L, Gothaus MC
13 (1998). A subchronic exposure to trichloroethylene causes lipid peroxidation and hepatocellular
14 proliferation in male B6C3F1 mouse liver, Toxicol. Sci. 43: 145-154.
15
16 Chang, LW; Daniel, FB; DeAngelo, AB. (1992) Analysis of DNA strand breaks induced in
17 rodent liver in vivo, hepatocytes in primary culture, and a human cell line by chloroacetic acids
18 and chloroacetaldehydes. Environ Mol Mutagen 20:277-288.
19
20 Charbotel B, Gad S, Caiola D, Beroud C, Fevotte J, Bergeret A, Ferlicot S, Richard S. (2007)
21 Trichloroethylene exposure and somatic mutations of the VHL gene in patients with Renal Cell
22 Carcinoma. J Occup Med Toxicol. 2:13.
23 Chow WH, Devesa SS. (2008) Contemporary epidemiology of renal cell cancer. Cancer J.
24 14(5):288-301.
25
26 Clay P (2008) Assessment of the genotoxicity of trichloroethylene and its metabolite, S-(l,2-
27 dichlorovinyl)-L-cysteine (DCVC), in the comet assay in rat kidney. Mutagenesis 23(1): 27-33.
28
29 Clewell HJ, Anderson ME. (2004) Applying mode-of-action and pharmacokinetic considerations
30 in contemporary cancer risk assessments: an example with trichloroethylene. Crit Rev Toxicol.
31 34(5):385-445.
6/22/2009
296
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1
2 Costa AK, Ivanetich KM (1984) Chlorinated ethylenes: their metabolism and effect on DNA
3 repair in rat hepatocytes. Carcinogenesis 5(12): 1629-36.
4
5 Crebelli R, Carere A. (1989) Genetic toxicology of 1,1,2-trichloroethylene. Mutation Res.
6 221:11-37.
7
8 Crebelli R, Bignami M, Conti L, Carere A. (1982) Mutagenicity of trichloroethylene in
9 Salmonella typhimurium TA100. Ann 1st Super Sanita. 18(1): 117—21.
10
11 Crebelli R, Conti G, Conti L, et al. (1985) Mutagenicity of trichloroethylene, trichloroethanol,
12 and chloral hydrate in Aspergillus nidulans. Mutat Res 155: 105-1 11.
13
14 Crebelli R, Conti G, Conti L, Carere A. (1991) In vitro studies with nine known or suspected
15 spindle poisons: results in tests for chromosome malsegregation in Aspergillus nidulans.
16 Mutagenesis. 6(2): 131-6.
17
18 DeMarini, DM; Perry, E; Shelton, ML. (1994) Dichloroacetic acid and related compounds:
19 induction of prophage in E. coli and mutagenicity and mutation spectra in Salmonella TA100.
20 Mutagenesis 9(5):429-37.
21
22 Dekant W, Vamvakas S, Berthold K, Schmidt S, Wild D, Henschler D. (1986) Bacterial beta-
23 lyase mediated cleavage and mutagenicity of cysteine conjugates derived from the
24 nephrocarcinogenic alkenes trichloroethylene, tetrachloroethylene and hexachlorobutadiene.
25 Chem Biol Interact. 60(1): 31-45.
26
27 Degrassi F, Tanzarella C (1988) Immunofluorescent staining of kinetochores in micronuclei: a
28 new assay for the detection of aneuploidy. Mutat Res. 203(5):339-45.
29
6/22/2009
297
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1 DiRenzo AB, Gandolfi AJ, Sipes IG (1982) Microsomal bioactivation and covalent binding of
2 aliphatic halides to DNA. Toxicol Lett 11:243-252.
3
4 Douglas GR, Gingerich JD, Soper LM, Potvin M, Bjarnason S. (1999) Evidence for the lack of
5 base-change and small-deletion mutation induction by trichloroethylene in lacZ transgenic mice.
6 Environ Mol Mutagen. 34(2-3): 190-4.
7
8 Eastmond DA, Tucker JD (1989) Identification of aneuploidy-inducing agents using cytokinesis-
9 blocked human lymphocytes and an antikinetochore antibody. Environ Mol Mutagen.
10 13(1):34—43.
11
12 Fahrig R, Madle S, Baumann H. (1995) Genetic toxicology of trichloroethylene (TCE). Mutation
13 Res. 340: 1-36.
14
15 Furnus CC, Ulrich MA, Terreros MC, Dulout FN. (1990) The induction of aneuploidy in
16 cultured Chinese hamster cells by propionaldehyde and chloral hydrate. Mutagenesis.
17 5(4):323-6.
18
19 Fuscoe JC, Afshari AJ, George MH, DeAngelo AB, Tice RR, Salman T, Allen JW.
20 (1996) In vivo genotoxicity of dichloroacetic acid: evaluation with the mouse peripheral blood
21 micronucleus assay and the single cell gel assay. Environ Mol Mutagen. 27(1): 1-9.
22
23 Galloway SM, Armstrong MJ, Reuben C, Colman S, Brown B, Cannon C, Bloom AD,
24 Nakamura F, Ahmed M, Duk S, et al. (1987) Chromosome aberrations and sister chromatid
25 exchanges in Chinese hamster ovary cells: evaluations of 108 chemicals. Environ Mol Mutagen.
26 10 Suppl 10:1-175.
27
28 Gibson DP, Aardema MJ, Kerckaert GA, Carr GJ, Brauninger RM, LeBoeuf RA. (1995)
29 Detection of aneuploidy-inducing carcinogens in the Syrian hamster embryo (SHE) cell
30 transformation assay. MutatRes. 343(l):7-24.
31
6/22/2009
298
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1 Giller S, Le Curieux F, Gauthier L, Erb F, Marzin D. (1995) Genotoxicity assay of chloral
2 hydrate and chloropicrine. Mutat Res. 348(4): 147-52.
3
4 Giller, S; Le Curieux, F; Erb, F; et al. (1997) Comparative genotoxicity of halogenated acetic
5 acids found in drinking water. Mutagenesis 12(5):321—328.
6
7 Grawe J, Musse M, Adler ID. (1997) Quantative and qualitative studies of micronucleus
8 induction in mouse erythrocytes using flow cytometry. I. Measurement of micronucleus
9 induction in peripheral blood polychromatic erythrocytes by chemicals with known and
10 suspected genotoxicity. Mutagenesis 12:1-8.
11
12 Greim H, Bonse G, Radwan Z, et al. (1975) Mutagenicity in vitro and potential carcinogenicity
13 of chlorinated ethylenes as a function of metabolic oxirane formation. Biochem Pharmacol
14 24:2013-2017.
15
16 Gu ZW, Sele B, Chmara D, Jalbert P, Vincent M, Vincent F, Marka C, Faure J (1981a) Effects
17 of trichloroethylene and its metabolites on the rate of sister chromatid exchange. In vivo and in
18 vitro study on the human lymphocytes [author's translation], Sem Hop 57(37—38): 1511—1512.
19
20 Gu ZW, Sele B, Jalbert P, Vincent M, Vincent F, Marka C, Chmara D, Faure J. (1981b)
21 [Induction of sister chromatid exchange by trichloroethylene and its metabolites (author's transl)]
22 Toxicol Eur Res. 3(2):63-7.
23
24 Gualandi G. (1987) Use of alpha- and beta-tubulin mutants for the study of spontaneous and
25 induced chromosomal mis-distribution in Aspergillus nidulans. Mutat Res. 178(1):33—41.
26
27 Harrington-Brock K, Doerr CL, Moore MM (1998) Mutagenicity of three disinfection by-
28 products: di- and trichloroacetic acid and chloral hydrate in L5178Y/TK+/- (-) 3.7.2C mouse
29 lymphoma cells. Mutat Res. 413(3):265-76.
30
6/22/2009
299
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1 Haworth S, Lawlor T, Mortelmans K, Speck W, Zeiger E. (1983) Salmonella mutagenicity test
2 results for 250 chemicals. Environ Mutagen. 5 Suppl 1:1-142.
3
4 Hennig UG, Rudd NL, Hoar DI. (1988) Kinetochore immunofluorescence in micronuclei: a
5 rapid method for the in situ detection of aneuploidy and chromosome breakage in human
6 fibroblasts. MutatRes. 203(6):405-14.
7
8 Henschler D, Eder E, Neudecker T, Metzler M. (1977) Carcinogenicity of trichloroethylene: fact
9 or artifact? Arch Toxicol. 37(3):233-6.
10
11 Herbert V, Gardner A, Colman N. (1980) Mutagenicity of dichloroacetate, an ingredient of some
12 formulations of pangamic acid (trade-named "vitamin B15"). Am J Clin Nutr. 33(6): 1179-82.
13
14 Horesovsky G, Ginsler J, Everitt J, Walker C. (1994) Increased susceptibility to in vitro
15 transformation of cells carrying the Eker tumor susceptibility mutation. Carcinogenesis.
16 10:2183-7.
17
18 Hrelia P, Maffei F, Vigagni F, Fimognari C, Flori P, Stanzani R, Cantelli Forti G. (1994)
19 Interactive effects between trichloroethylene and pesticides at metabolic and genetic level in
20 mice. Environmental Health Perspectives. 102 (Suppl 9): 31-34.
21
22 Hu C, Jiang L, Geng C, Zhang X, Cao J, Zhong L (2008) Possible involvement of oxidative
23 stress in trichloroethylene-induced genotoxicity in human GepG2 cells. Mutation Res.
24 652:88-94.
25
26 Ikbal M, Tastekin A, Dogan H, Pirim I, Ors R. (2004) The assessment of genotoxic effects in
27 lymphocyte cultures of infants treated with chloral hydrate. Mutation Res. 564: 159-164.
28
29 Jaffe DR, Hassall CD, Gandolfi AJ, Brendel K. (1985) Production of DNA single strand breaks
30 in rabbit renal tissue after exposure to 1,2-dichlorovinylcysteine. Toxicology. 35(1):25—33.
31
6/22/2009
300
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1 Kafer E. (1986) Tests which distinguish induced crossing-over and aneuploidy from secondary
2 segregation in Aspergillus treated with chloral hydrate or gamma-rays. Mutat Res.
3 164(3): 145-66.
4
5 Kappas A (1989) On the mechanisms of induced aneuploidy in Aspergillus nidulans and
6 validation of tests for genomic mutations. Prog Clin Biol Res. 318:377-84.
7
8 Kargalioglu, Y; McMillan, BJ; Minear, RA et al. (2002) Analysis of the cytotoxicity and
9 mutagenicity of drinking water disinfection by-products in Salmonella typhimurium. Teratog
10 Carcinog Mutagen 22:113-128.
11
12 Kautiainen A, Vogel JS, Turteltaub KW (1997) Dose-dependent binding of trichloroethylene to
13 hepatic DNA and protein at low doses in mice. Chemico-Biological Interactions 106: 109-121.
14
15 Keller DA, Heck HD. (1988) Mechanistic studies on chloral toxicity: Relationship to
16 trichloroethylene carcinogenesis. Toxicol Lett 42: 183-191.
17
18 Kenck C, Wilhelm M, Bugert P, Staehler G, Kovacs G. (1996) Mutation of the VHL gene is
19 associated exclusively with the development of non-papillary renal cell carcinomas. J Pathol.
20 179(2): 157-61.
21
22 Klaunig JE, Ruch RJ, Lin EL. (1989) Effects of trichloroethylene and its metabolites on rodent
23 hepatocyte intercellular communication. Toxicol Appl Pharmacol. 99(3):454-65.
24
25 Kligerman AD, Bryant MF, Doerr CL, Erexon GL, Evansky PA, Kwanyuen P, McGee JK
26 (1994) Inhalation studies of the genotoxicity of trichloroethylene to rodents. Mutation Res.
27 322(2):87-96.
28
29 Koch R, Schlegehnilch R, Wolf HU. (1988). Genetic effects of chlorinated ethylenes in the yeast
30 Saccharomyces cerevisiae. Mutat. Res. 206:209-216.
31
6/22/2009
301
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1 Kubo Y, Mitani H, Hino O. (1994) Allelic loss at the predisposing gene locus in spontaneous and
2 chemically induced renal cell carcinomas in the Eker rat. Cancer Res. 54(10):2633-5.
3
4 Leavitt SA, DeAngelo AB, George MH, Ross JA (1997) Assessment of the mutagenicity of
5 dichloroacetic acid in lacl transgenic B6C3F1 mouse liver. Carcinogenesis 18(11): 2101-2106.
6
7 Leopardi P, Zijno A, Bassani B, Pacchierotti F. (1993) In vivo studies on chemically induced
8 aneuploidy in mouse somatic and germinal cells. Mutat Res. 287(1): 119-30.
9
10 Leuschner J, Leuschner F. (1991) Evaluation of the mutagenicity of chloral hydrate in vitro and
11 in vivo. Arzneimittelforschung. 41(10): 1101-3.
12
13 Leuschner J, Beuscher N. (1998) Studies on the mutagenic and carcinogenic potential of chloral
14 hydrate. Arzneimittelforschung. 48(10):961-8.
15
16 Liang JC, Brinkley BR. (1985) Chemical probes and possible targets for the induction of
17 aneuploidy. Basic Life Sci. 36:491-505
18
19 Liang JC, Pacchierotti F. (1988) Cytogenetic investigation of chemically-induced aneuploidy in
20 mouse spermatocytes. Mutat Res. 201(2):325-35.
21
22 Liu E, Percy MJ, Amos CI, Guan Y, Shete S, Stockton DW, McMullin MF, Polyakova LA, Ang
23 SO, Pastore YD, Jedlickova K, Lappin TR, Gordeuk V, Prchal JT. (2004) The worldwide
24 distribution of the VHL 598C>T mutation indicates a single founding event. Blood.
25 103(5): 1937-40.
26
27 Lynch AM, Parry JM. (1993) The cytochalasin-B micronucleus/kinetochore assay in vitro:
28 studies with 10 suspected aneugens. Mutat Res. 287(l):71-86.
29
6/22/2009
302
-------
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
1 Mackay, JM; Fox, V; Griffiths, K; et al. (1995) Trichloroacetic acid: investigation into the
2 mechanism of chromosomal damage in the in vitro human lymphocyte cytogenetic assay and the
3 mouse bone marrow micronucleus test. Carcinogenesis 16(5): 1127-1133.
4
5 Maihes JB, Aardema MJ, Marchetti F (1993) Investigation of aneuploidy induction in mouse
6 oocytes following exposure to vinblastine-sulfate, pyrimethamine, diethylstilbestrol diphosphate,
7 or chloral hydrate. Environ Mol Mutagen 22(2) 107-114.
8
9 Mally A, Walker CL, Everitt JI, Dekant W, Vamvakas S. (2006) Analysis of renal cell
10 transformation following exposure to trichloroethene in vivo and its metabolite S-
11 (dichlorovinyl)-L-cysteine in vitro.: Toxicology. 224(1-2): 108-18.
12
13 Mazzullo M, Bartoli S, Bonora B, Colacci A, Lattanzi G, Niero A, Silingardi P, Grilli S. (1992)
14 In vivo and in vitro interaction of trichloroethylene with macromolecules from various organs of
15 rat and mouse. Res Commun Chem Pathol Pharmacol. 76(2): 192-208.
16
17 McGregor DB, Reynolds DM, Zeiger E. (1989) Conditions affecting the mutagenicity of
18 trichloroethylene in Salmonella. Environ Mol Mutagen 13:197-202.
19
20 Miller RE, Guengerich FP (1983) Metabolism of trichloroethylene in isolated hepatocytes,
21 microsomes, and reconstituted enzyme systems containing cytochrome P-450. Cancer Res
22 43:1145-1152.
23
24 Miller BM, Adler ID (1992) Aneuploidy induction in mouse spematocytes. Mutagenesis
25 7:37-43.
26
27 Moore MM, Harrington-Brock K. (2000) Mutagenicity of trichloroethylene and its metabolites:
28 implications for the risk assessment of trichloroethylene. Environ Health Perspect. 108 Suppl
29 2:215-23.
30
6/22/2009
303
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Mortelmans K, Haworth S, Lawlor T, Speck W, Tainer B, Zeiger E. (1986) Salmonella
mutagenicity tests: II. Results from the testing of 270 chemicals. Environ Mutagen. 8 Suppl
7:1-119.
Mazzullo M, Bartoli S, Bonora B, Colacci A, Lattanzi G, Niero A, Silingardi P, Grilli S. (1992)
In vivo and in vitro interaction of trichloroethylene with macromolecules from various organs of
rat and mouse. Res Commun Chem Pathol Pharmacol 76(2): 192-208.
Nagaya T, Ishikawa N, Hata H (1989) Sister chromatid exchanges in lymphocytes of workers
exposed to trichloroethylene. Mutat Res 222:279-282.
Natarajan AT, Duivenvoorden WC, Meijers M, Zwanenburg TS. (1993) Induction of mitotic
aneuploidy using Chinese hamster primary embryonic cells. Test results of 10 chemicals. Mutat
Res. 287(l):47-56.
Nelson, MA; Bull, RJ. (1988) Induction of strand breaks in DNA by trichloroethylene and
metabolites in rat and mouse liver in vivo. Toxicol Appl Pharmacol 94:45-54.
Nelson MA, Lansing AJ, Sanchez IM, Bull RJ, Springer DL. (1989) Dichloroacetic acid and
trichloroacetic acid-induced DNA strand breaks are independent of peroxisome proliferation.
Toxicology. 58(3):239-48.
Nelson, GM; Swank, AE; Brooks, LR; et al. (2001) Metabolism, microflora effects, and
genotoxicity in haloacetic acid-treated cultures of rat cecal microbiota. Toxicol Sci
60(2):232-241.
Ni YC, Wong TY, Kadlubar FF, Fu PP. (1994) Hepatic metabolism of chloral hydrate to free
radical(s) and induction of lipid peroxidation. Biochem Biophys Res Commun. 204(2): 93 7-43.
Ni YC, Kadlubar FF, Fu PP. (1995) Formation of malondialdehyde-modified 2'-deoxyguanosinyl
adduct from metabolism of chloral hydrate by mouse liver microsomes. Biochem Biophys Res
Commun. 216(3): 1110-7.
6/22/2009
304
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Nickerson ML, Jaeger E, Shi Y, Durocher JA, Mahurkar S, Zaridze D, Matveev V, Janout V,
Kollarova H, Bencko V, Navratilova M, Szeszenia-Dabrowska N, Mates D, Mukeria A,
Holcatova I, Schmidt LS, Toro JR, Karami S, Hung R, Gerard GF, Linehan WM, Merino M,
Zbar B, Boffetta P, Brennan P, Rothman N, Chow WH, Waldman FM, Moore LE. (2008)
Improved identification of von Hippel-Lindau gene alterations in clear cell renal tumors. Clin
Cancer Res. 14(15):4726-34.
Nesslany F, Marzin D. (1999) A micromethod for the in vitro micronucleus assay. Mutagenesis.
14(4):403-10.
Parry JM, Parry EM, Bourner R, Doherty A, Ellard S, O'Donovan J, Hoebee B, de Stoppelaar
JM, Mohn GR, Onfelt A, Renglin A, Schultz N, Soderpalm-Berndes C, Jensen KG, Kirsch-
Volders M, Elhajouji A, Van Hummelen P, Degrassi F, Antoccia A, Cimini D, Izzo M,
Tanzarella C, Adler ID, Kliesch U, Hess P, et al. (1996) The detection and evaluation of
aneugenic chemicals. MutatRes. 353(1-2): 11-46.
Nutley EV, Tcheong AC, Allen JW, Collins BW, Ma M, Lowe XR, Bishop JB, Moore DH,
Wyrobek AJ (1996) Micronuclei induced in round spermatids of mice after stem-cell treatment
with chloral hydrate: evaluations with centromeric DNA probes and kinetochore antibodies.
Environ Mol Mutagen 28(2):80-89.
Ono, Y; Somiya, I; Kawamura, M. (1991) The evaluation of genotoxicity using DNA repairing
test for chemicals produced in chlorination and ozonation processes. Water Sci Technol
23(l-3):329-338.
Parnell MJ, Koller LD, Exon JH, Arnzen JM. (1986) Trichloroacetic acid effects on rat liver
peroxisomes and enzyme-altered foci. Environ Health Perspect. 69:73-9.
Parrish, JM; Austin, EW; Stevens, DK; et al. (1996) Haloacetate-induced oxidative damage to
DNA in the liver of male B6C3Fi mice. Toxicology 110:103-111.
Perocco P, Prodi G. (1981) DNA damage by haloalkanes in human lymphocytes cultured in
vitro. Cancer Lett 13:213-218.
6/22/2009
305
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Plewa MJ, Kargalioglu Y, Vankerk D, Minear RA, Wagner ED. (2002) Mammalian cell
cytotoxicity and genotoxicity analysis of drinking water disinfection by-products. Environ Mol
Mutagen 40:134-142.
Price PJ, Hassett CM, Mansfield JI. (1978) Transforming activities of trichloroethylene and
proposed industrial alternatives. In Vitro 14:290-293.
Rapson, WH; Nazar, MA; Butsky, VV. (1980) Mutagenicity produced by aqueous chlorination
of organic compounds. Bull Environ Contam Toxicol 24:590-596.
Robbiano L, Mereto E, Migliazzi Morando A, Pastore P, Brambilla G. (1998) Increased
frequency of micronucleated kidney cells in rats exposed to halogenated anaesthetics. Mutat Res.
413(1): 1—6.
Robbiano L, Baroni D, Carrozzino R, Mereto E, Brambilla G. (2004) DNA damage and
micronuclei induced in rat and human kidney cells by six chemicals carcinogenic to the rat
kidney. Toxicology. 204(2-3): 187-95.
Rossi AM, Migliore L, Barali R, et al. (1983) In vivo and in vitro mutagenicity studies of a
possible carcinogen, trichloroethylene, and its two stabilizers, epichlorohydrin and 1,2-
epoxybutane. Teratogenesis Carcinog Mutagen 3:75-87.
Russo A, Pacchierotti F, Metalli P. N (1984) Nondisjunction induced in mouse spermatogenesis
by chloral hydrate, a metabolite of trichloroethylene. Environ Mutagen. 6: 695-703.
Russo A, Stocco A, Majone F (1992) Identification of kinetochore-containing (CREST+)
micronuclei in mouse bone marrow erythrocytes. Mutagenesis 7(3): 195-197.
Russo A, Levis AG. (1992a) Detection of aneuploidy in male germ cells of mice by means of a
meiotic micronucleus assay. Mutat Res. 281(3): 187—91.
Russo A, Levis AG. (1992b) Further evidence for the aneuploidogenic properties of chelating
agents: induction of micronuclei in mouse male germ cells by EDTA. Environ Mol Mutagen.
19(2): 125-31.
6/22/2009
306
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Salmon AG, Kizer KW, Zeise L, Jackson RJ, Smith MT. (1995) Potential carcinogenicity of
chloral hydrate—a review. J Toxicol Clin Toxicol. 33(2): 115—21.
Sbrana I, Di Sibio A, Lomi A, Scarcelli V. (1993) C-mitosis and numerical chromosome
aberration analyses in human lymphocytes: 10 known or suspected spindle poisons. Mutat Res.
287(l):57-70.
Schatten H, Chakrabarti A. (1998) Centrosome structure and function is altered by chloral
hydrate and diazepam during the first reproductive cell cycles in sea urchin eggs. Eur J Cell Biol.
75(l):9-20.
Schraml P, Kononen J, Bubendorf L, Moch H, Bissig H, Nocito A, Mihatsch MJ, Kallioniemi
OP, Sauter G. (1999) Tissue microarrays for gene amplification surveys in many different tumor
types. Clin Cancer Res. 5(8): 1966-75.
Seelbach A, Fissler B, Madle S. (1993) Further evaluation of a modified micronucleus assay
with V79 cells for detection of aneugenic effects. Mutat Res. 303(4): 163-9.
Shahin MM and von Borstel RC (1977). Mutagenic and lethal effects of a-benzene hexachloride,
dibutyl-phthalate and trichloroethylene in Saccharomyces cerevisiae, Mutation Res. 48:
173-180.
Shiao YH, Rice JM, Anderson LM, Diwan BA, Hard GC. (1998) von Hippel-Lindau gene
mutations in N-nitrosodimethylamine-induced rat renal epithelial tumors. J Natl Cancer Inst.
90(22): 1720-3.
Shiao YH, Kamata SI, Li LM, Hooth MJ, DeAngelo AB, Anderson LM, Wolf DC. (2002)
Mutations in the VHL gene from potassium bromate-induced rat clear cell renal tumors. Cancer
Lett. 87(l-2):207-14.
Shimada T, Swanson AF, Leber P, et al. (1985) Activities of chlorinated ethane and ethylene
compounds in the Salmonella microsome mutagenesis and rat hepatocyte/DNA repair assays
under vapor phase exposure conditions. Cell Biol Toxicol 1:159-179.
6/22/2009
307
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Singh M, Sinha U. (1976) Chloral hydrate induced haploidization in Aspergillus nidulans.
Experientia. 32(9): 1144-5.
Sora S, Agostini Carbone ML. (1987) Chloral hydrate, methylmercury hydroxide and ethidium
bromide affect chromosomal segregation during meiosis of Saccharomyces cerevisiae. Mutat
Res. 190(1): 13-7.
Stemmer K, Ellinger-Ziegelbauer H, Ahr HJ, Dietrich DR. (2007) Carcinogen-specific gene
expression profiles in short-term treated Eker and wild-type rats indicative of pathways involved
in renal tumorigenesis. Cancer Res. 67(9):4052-68.
Storer RD, McKelvey TW, Kraynak AR, Elia MC, Barnum JE, Harmon LS, Nichols WW,
DeLuca JG. (1996) Revalidation of the in vitro alkaline elution/rat hepatocyte assay for DNA
damage: improved criteria for assessment of cytotoxicity and genotoxicity and results for 81
compounds. Mutat Res. 368(2):59-101.
Stott WT, Quast JF, Watanabe PG (1982) The pharmacokinetics and macromolecular
interactions of trichloroethylene in mice and rats. Toxicol Appl Pharmacol. 62:137-151.
Styles, JA; Wyatt, I; Coutts, C. (1991) Trichloroacetic acid: studies on uptake and effects on
hepatic DNA and liver growth in mouse. Carcinogenesis 12(9): 1715-1719.
Toma MI, Grosser M, Herr A, Aust DE, Meye A, Hoefling C, Fuessel S, Wuttig D, Wirth MP,
Baretton GB. (2008) Loss of heterozygosity and copy number abnormality in clear cell renal cell
carcinoma discovered by high-density affymetrix 10K single nucleotide polymorphism mapping
array. Neoplasia. 10(7):634-42.
Toraason M, Clark J, Dankovic D, Mathias P, Skaggs S, Walker C, Werren D (1999) Oxidative
stress and DNA damage in Fischer rats following acute exposure to trichloroethylene or
perchloroethylene, Toxicology 138: 43-53.
Tu AS, Murray TA, Hatch KM, et al. (1985) In vitro transformation of BALBZC-3T3 cells by
chlorinated ethanes and ethylenes. Cancer Lett 28:85-92.
Vagnarelli P, De Sario A, De Carli L. (1990) Aneuploidy induced by chloral hydrate detected in
human lymphocytes with the Y97 probe. Mutagenesis. 5(6):591-2.
6/22/2009
308
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Vamvakas S, Dekant W, Berthold K, Schmidt S, Wild D, Henschler D. (1987) Enzymatic
transformation of metcapturic acids derived from halogenated alkenes to reactive mutagenic
intermediates. Biochem Pharmacol. 36(17):2741-8.
Vamvakas S, Elfarra AA, Dekant W, Henschler D, Anders MW (1988a) Mutagenicity of amino
acid and glutathione S-conjugates in the Ames test. Mutat Res. 206(l):83-90.
Vamvakas S, Dekant W, Schiffmann D, Henschler D. (1988b) Induction of unscheduled DNA
synthesis and micronucleus formation in Syrian hamster embryo fibroblasts treated with cysteine
S-conjugates of chlorinated hydrocarbons. Cell Biol Toxicol. (4):393-403.
Vamvakas S, Dekant W, Henschler D. (1989) Assessment of unscheduled DNA synthesis in a
cultured line of renal epithelial cells exposed to cysteine S-conjugates of haloalkenes and
haloalkanes. Mutat Res. 222(4):329-35.
Vamvakas S, Bittner D, Koster U. (1993) Enhanced expression of the protooncogenes c-myc and
c-fos in normal and malignant renal growth. Toxicol Lett. 67(1-3): 161-72.
Vamvakas S, Richter H, Bittner D. (1996) Induction of dedifferentiated clones of LLC-PK1 cells
upon long-term exposure to dichlorovinylcysteine. Toxicology. 106(l-3):65-74.
Van Hummelen P, Kirsch-Volders M. (1992) Analysis of eight known or suspected aneugens by
the in vitro human lymphocyte micronucleus test. Mutagenesis. 7(6):447-55.
Von Tungeln LS, Yi P, Bucci TJ, Samokyszyn VM, Chou MW, Kadlubar FF, Fu PP. (2002)
Tumorigenicity of chloral hydrate, trichloroacetic acid, trichloroethanol, malondialdehyde, 4-
hydroxy-2-nonenal, crotonaldehyde, and acrolein in the B6C3F(1) neonatal mouse. Cancer Lett.
185(1): 13—9.
Walles SA. (1986) Induction of single-strand breaks in DNA of mice by trichloroethylene and
tetrachloroethylene. Toxicol Lett. 31: 31-35.
Wang JL, Chen WL, Tsai SY, Sung PY, Huang RN. (2001) An in vitro model for evaluation of
vaporous toxicity of trichloroethylene and tetrachloroethylene to CHO-K1 cells. Chem Biol
Interact. 137(2): 139-54.
6/22/2009
309
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT CITE OR QUOTE
Waskell L. (1978) A study of the mutagenicity of anesthetics and their metabolites. Mutat Res
57:141-153.
Warr TJ, Parry EM, Parry JM.(1993) A comparison of two in vitro mammalian cell cytogenetic
assays for the detection of mitotic aneuploidy using 10 known or suspected aneugens. Mutat Res.
287(l):29-46.
White AE, Eger EI, Wolff S, Stevens WC (1979) Sister chromatid exchange induced by inhaled
anesthetics. Anesthesiology 50: 426-430.
Wolf DC, Goldsworthy TL, Janszen DB, Harden R, Donner EM, David CS, Everitt JI. (2000)
Promotion by sodium barbital induces early development but does not increase the multiplicity
of hereditary renal tumors in Eker rats. Carcinogenesis. 21(8): 1553-8.
Xu W, Adler ID (1990) Clastogenic effects of known and suspect spindle poisons studied by
chromosome analysis in mouse bone marrow cells. Mutagenesis 5(4): 371-374.
Yang Y, Heng, ZC (2006) Study on rat's DNA damage of p53 induced by chlorinated acetic
acids of drinking water disinfection by-products (article in Chinese). Wei Sheng Yan Jiu.
35(6):683-5.
Yeung RS, Xiao GH, Everitt JI, Jin F, Walker CL. (1995) Allelic loss at the tuberous sclerosis 2
locus in spontaneous tumors in the Eker rat. Mol Carcinog. 14(1):28—36.
Yoon JS, Mason JM, Valencia R, Woodruff RC, Zimmering S. (1985) Chemical mutagenesis
testing in Drosophila. IV. Results of 45 coded compounds tested for the National Toxicology
Program. Environ Mutagen. 7(3):349-67.
Yoon HS, Monks TJ, Walker CL, Lau SS. (2001) Transformation of kidney epithelial cells by a
quinol thioether via inactivation of the tuberous sclerosis-2 tumor suppressor gene. Mol
Carcinog. 31(1):37—45.
6/22/2009
310
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.2 Central Nervous System Toxicity
TCE exposure results in CNS effects in both humans and animals that can result from
acute, subchronic, or chronic exposure. There are studies indicating that TCE exposure results in
CNS tumors and this discussion can be found in Section 4.8. The studies discussed in this
section focus on the most critical neurological effects that were extracted from the
neurotoxicological literature. Although there are several studies and reports that have evaluated
TCE as an anesthetic, those studies were not included in this section because of the high
exposure levels in comparison to the selected critical neurological effects described below. The
critical neurological effects are nerve conduction changes, sensory effects, cognitive deficits,
changes in psychomotor function, and changes in mood and sleep behaviors. The selection
criteria that were used to determine study importance included study design and validity,
pervasiveness of neurological effect, and for animal studies, the relevance of these reported
outcomes in humans. More detailed information on human and animal neurological studies with
TCE can be found in Appendix D.
4.2.1 Alterations in Nerve Conduction
4.2.1.1 Trigeminal Nerve Function: Human Studies
A number of human studies have been conducted that examined the effects of
occupational or drinking water exposures to TCE on trigeminal nerve function (see Table 4.2.1).
Many studies reported that humans exposed to TCE present trigeminal nerve function
abnormalities as measured by blink reflex and masseter reflex test measurements (Feldman et al.,
1988, 1992; Kilburn and Warshaw, 1993; Kilburn, 2002a; Ruitjen et al., 2001). The blink and
masseter reflexes are mediated primarily by the trigeminal nerve and changes in measurement
suggest impairment in nerve conduction. Other studies measured the trigeminal somatosensory
evoked potential (TSEP) following stimulation of the trigeminal nerve and reported statistically
significantly delayed response on evoked potentials among exposed subjects compared to non-
exposed individuals (Barret et al., 1982, 1984, 1987; Mhiri et al., 2004). Two studies which also
measured trigeminal nerve function did not find any effect (El-Ghawabi et al., 1973; Rasmussen
et al., 1993c) but the methods were not provided in one study (El-Ghawabi et al., 1973) or an
appropriate control group was not included (Rasmussen et al., 1993c). These studies and results
are described below and summarized in detail in Table 4.2-1.
6/22/2009
311
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Integrity of the trigeminal nerve is commonly measured using blink and masseter
reflexes. Five studies (Barret et al., 1984; Feldman et al., 1988, 1992; Kilburn and Warshaw,
1993; Kilburn, 2002a) reported a significant increase in the latency to respond to the stimuli
generating the reflex. The latency increases in the blink reflex ranged from 0.4 ms (Kilburn,
2002a) to up to 3.44 ms (Feldman et al., 1988). The population groups in these studies were
exposed by inhalation occupationally (Barret et al., 1984) and through drinking water
environmentally (Feldman et al., 1988; Kilburn and Warshaw, 1993; Kilburn, 2002a). Feldman
et al. (1992) demonstrated persistence in the increased latency of the blink reflex response. In
one subject, exposure to TCE (levels not reported by authors) occurred through a degreasing
accident (high and acute exposure), and increased latency response times persisted 20 years after
the accident. Another two subjects, evaluated at 9 months and 1 month following a high
occupational exposure (exposure not reported by authors), also had higher blink reflex latencies
with an average increase of 2.8 ms over the average response time in the control group used in
the study. Although one study (Ruitjen et al., 1991) did not find these increases in male printing
workers exposed to TCE, this study did find a statistically significant average increase of 0.32
ms (p < 0.05) in the latency response time in TCE-exposed workers on the masseter reflex test,
another test commonly used to measure the integrity of the trigeminal nerve.
Three studies (Barret et al., 1982, 1987; Mhiri et al., 2004) adopting TSEPs to measure
trigeminal nerve function found significant abnormalities in these evoked potentials. These
studies were conducted on volunteers who were occupationally exposed to TCE through metal
degreasing operations (Barret et al., 1982, 1987) or through cleaning tanks in the phosphate
industry (Mhiri et al., 2004). Barret et al. (1982) reported that in eight of the eleven workers, an
increased voltage ranging from a 25 to a 45 volt increase was needed to generate a normal TSEP
and two of workers had an increased TSEP latency. Three out of 11 workers had increases in
TSEP amplitudes. In a later study, Barret et al. (1987) also reported abnormal TSEPs (increased
latency and/or increased amplitude) in 38% of the degreasers that were evaluated. The
individuals with abnormal TSEPs were significantly older (45 years vs. 40.1 years; p < 0.05) and
were exposed to TCE longer (9.9 years vs. 5.6 years;p < 0.01). Mhiri et al. (2004) was the only
study to evaluate individual components of the TSEP and noted significant increases in latencies
for all TSEP potentials (Nl, PI, N2, P2, N3;p < 0.01) and significant decreases in TSEP
amplitude (PI ,P< 0.02; N2,p < 0.05). A significant positive correlation was demonstrated
between exposure duration and increased TSEP latency (p < 0.02).
Two studies reported no statistically significant effect of TCE exposure on trigeminal
nerve function (El-Ghawabi et al., 1973; Rasmussen et al., 1993). El-Ghawabi et al. (1973)
6/22/2009
312
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
conducted a study on 30 money printing shop workers occupationally exposed to TCE.
Trigeminal nerve involvement was not detected, but the authors did not include the experimental
methods that were used to measure trigeminal nerve involvement and did not provide any data as
to how this assessment was made. Rasmussen et al. (1993c) conducted an historical cohort study
on 99 metal degreasers, 70 exposed to TCE and 29 to the fluorocarbon, CFC 113. It was
reported that 1 out of 21 people (5%) in the low exposure, 2 out of 37 (5%) in the medium
exposure and 4 out of 41 (10%) in the high exposure group experienced abnormalities in
trigeminal nerve sensory function, with a linear trend testp-walue of 0.42. The mean urinary
trichloroacetic acid concentration was reported for the high exposure group only and was 7.7
mg/L (maximum concentration, 26.1 mg/L). The trigeminal nerve function findings of high
exposure group subjects was compared to that of low exposure group since this study did not
include an unexposed or no TCE exposure group.
4.2.1.2Nerve Conduction Velocity — Human Studies
Two occupational studies assessed ulnar and median nerve function using tests of
conduction latencies (Triebig, 1982, 1983) (see Table 4.2-1). The ulnar nerve and median nerves
are major nerves located in the arm and forearm. Triebig (1982) studied twenty-four healthy
workers (20 males, 4 females) exposed to TCE occupationally (5-70 ppm) at three different
plants and did not find statistically significant differences in ulnar or median nerve conduction
velocities between exposed and unexposed subjects. This study has measured exposure data, but
exposures/responses are not reported by dose levels. The Triebig (1983) study is similar in
design to the previous study (Triebig, 1982). In this study, a dose response relationship was
observed between lengths of exposure to mixed solvents that included TCE (at unknown
concentration). A statistically significant reduction in nerve conduction velocities was observed
for the medium- and long-term exposure groups for the ulnar nerve and median nerves.
Table 4.2-1 Summary of human trigeminal nerve and nerve conduction velocity studies
Reference
Subjects
Exposure
Effect
Barret etal., 1982
11 workers with chronic
TCE exposure
Controls: 20 unexposed
subjects.
Presence of TCE and TCA
found through urinalysis.
Atmospheric TCE
concentrations and
Duration of exposure not
reported in paper
Following stimulation of the trigeminal
nerve, significantly higher voltage
stimuli was required to obtain a normal
response and there was a significant
increase in latency for response and
decreased response amplitude.
6/22/2009
313
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Barret et al., 1984
188 factory workers
No unexposed controls;
lowest exposure group
used as comparison
>150 ppm; n = 54 < 150
ppm; n = 134,
7 hrs/day for 7 years
Trigeminal nerve and optic nerve
impairment, asthenia and dizziness were
significantly increased with exposure.
Barret etal., 1987
104 degreaser machine
operators
Controls: 52 unexposed
subjects
Mean age 41.6 yrs
Mean duration, 8.2 yrs,
average daily exposure 7
hrs/day.
Average TCOH range =
162-245 mg/g creatinine
Average TCA range =
93-131 mg/g creatinine
Evoked trigeminal responses were
measured following stimulation of the
nerve and revealed increased latency to
respond, amplitude or both and
correlated with length of exposure
(p <0.01) and with age (p < 0.05), but
not concentration.
El-Ghawabi et al.,
1973
30 money printing shop
workers
Controls: 20 non-
exposed males
10 workers exposed to
inks not containing TCE
Mean TCE air
concentrations ranged
from 41 ppm to 163
ppm. Exposure
durations:
Less than 1 year: n = 3
1 year: n = 1
2 years: n = 2
3 years: n= 11
4 years: n = 4
5 years or greater: n = 9
No effect on trigeminal nerve function
was noted.
Feldman et al.,
1988
21 Woburn, MA
residents;
27 controls
TCE maximum
reported concentration
in well water was 267
ppb; other solvents
also present.
Exposure duration
ranged from 1-12
years.
Measurement of the blink reflex as
mediated by the trigeminal nerve
resulted in significant increases in the
latency of reflex components
(p< 0.001).
6/22/2009
314
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Feldman et al.,
18 workers;
TCE exposure
The blink reflex as mediated by the
1992
30 controls
categories of
trigeminal was measured. The
"extensive" group revealed latencies
"extensive",
greater than 3 SD above the non-
"occasional", and
exposed group mean on blink reflex
"chemical other than
components.
TCE"
"extensive" =
chronically exposed
(>1 yr) to TCE for 5
days/week and >50%
workday.
"occupational" =
chronically exposed to
TCE for 1-3
days/week and >50%
workday.
Kilburn and
160 residents living in
>500 ppb of TCE in
Significant impairments in sway speed
Warshaw, 1993
Southwest Tucson with
well water before 1981
with eyes open and closed and blink
TCE, other solvents, and
and 25 to 100 ppb
reflex latency (R-l) which suggests
chromium in
trigeminal nerve impairment.
groundwater.
afterwards.
Control: 113 histology
technicians from a
Duration ranged from
previous study (Kilburn
1 to 25 years.
et al., 1987; Kilburn and
Warshaw, 1992)
6/22/2009
315
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Kilburn, 2002a
236 residents near a
microchip plant in
Phoenix, AZ.
Controls: 161 regional
referents from
Wickenburg, AZ and 67
referents in northeastern
Phoenix.
<0.2-10,000 ppb of
TCE, <0.2-260,000
ppb TCA, <0.2-6,900
ppb 1, 1-DCE,
<0.2-1,600 1, 2-DCE,
<0.2-23,000 ppb PCE,
<0.02-330 ppb VC in
well water.
Exposure duration
ranged from 2 to 37
years.
Trigeminal nerve impairment as
measured by the blink reflex test; both
right and left blink reflex latencies (R-l)
were prolonged. Exposed group mean
14.2 + 2.1 ms (right) or 13.9 + 2.1 ms
(left) versus referent group mean of
13.4 +2.1 ms (right) orl3.5 +2.1ms
(left), p = 0.0001 (right) and 0.008 (left).
Mhiri et al., 2004
23 phosphate industry
workers
Controls: 23 unexposed
workers
Exposure ranged from
50-150 ppm, for 6
hrs/day for at least two
years.
Mean urinary
trichloroethanol and
trichloroacetic acid
levels were 79.3 ± 42
and 32.6 ± 22 mg/g
creatinine
Trigeminal somatosensory evoked
potentials (TSEPs) were recorded.
Increase in the TSEP latency was
observed in 15 out of 23 (65%) workers.
6/22/2009
316
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Rasmussen et al.,
1993c
96 Danish metal
degreasers
Age range: 19-68;
No unexposed controls;
low exposure group used
as comparison;
Average exposure
duration: 7.1 yrs.);
range of full-time
degreasing: 1 month to
36 yrs. Exposure to
TCE or to CFC 113
1) Low exposure: n =
19, average full-time
exposure 0.5 yrs
2) Medium exposure:
n = 36, average full-
time exposure 2.1 yrs.
3) High exposure: n =
41, average full-time
exposure 11 yrs. TCA
in high exposure group
= 7.7 mg/L
(max = 26.1 mg/L)
No statistically significant trend on
trigeminal nerve function, although
some individuals had abnormal function.
Ruitjen et al.,
1991
31 male printing
workers. Mean age 44
yrs; kean duration 16
years.
Controls: 28 unexposed;
Mean age 45 yrs
Mean cumulative
exposure = 704 ppm x
years (SD 583, range:
160-2,150 ppm x
years
Mean, 17 ppm at time
of study; historic TCE
levels from
1976-1981, mean of
35 ppm
Mean duration of 16
yrs.
Measurement of trigeminal nerve
function by using the blink reflex
resulted in no abnormal findings.
Increased latency in the masseter reflex
is indicative of trigeminal nerve
impairment.
6/22/2009
317
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Triebig et al.,
1982
24 workers (20 males, 4
females) occupationally
exposed—ages 17-56.
Controls: 144
individuals to establish
normal nerve conduction
parameters.
Matched group: 24
unexposed workers (20
males, 4 females)
Exposure duration of 1
month to 258 months
(mean 83 months).
Air exposures were
between 5-70 ppm
No statistically significant difference in
nerve conduction velocities between the
exposed and unexposed groups.
Triebig et al.,
1983
66 workers
occupationally exposed
Control: 66 workers not
exposed to solvents
Subjects were exposed
to a mixture of
solvents, including
TCE.
Exposure-response relationship
observed between length of solvent
exposure and statistically significant
reduction in ulnar nerve conduction
velocities.
4.2.1.3 Trigeminal Nerve Function: Laboratory Animal Studies
There is little evidence that TCE disrupts trigeminal nerve function in animal studies.
Two studies demonstrated TCE produces morphological changes in the trigeminal nerve at a
dose of 2,500 mg/kg-day for 10 weeks (Barret et al., 1991, 1992). However, dichloroacetylene, a
degradation product formed during the volatilization of TCE was found to produce more severe
morphological changes in the trigeminal nerve and at a lower dose of 17 mg/kg-day (Barret et
al., 1991,1992). Only one study (Albee et al., 2006) has evaluated the effects of TCE on
trigeminal nerve function and a subchronic inhalation exposure did not result in any significant
functional changes. A summary of these studies is provided in Table 4.2-2.
Barret et al. (1991,1992) conducted two studies evaluating the effects of both TCE and
dichloroacetylene on trigeminal nerve fiber diameter and internodal length as well as several
markers for fiber myelination. Female Sprague Dawley rats (n = 7/group) were dosed with 2,500
mg/kg TCE or 17 mg/kg-day dichloroacetylene by gavage for 5 days/week for 10 weeks. TCE-
dosed animals only exhibited changes in the smaller Class A fibers where internode length
increased marginally (<2%) and fiber diameter increased by 6%. Conversely, dichloroacetylene-
treated rats exhibited significant and more robust decreases in internode length and fiber
diameter in both fiber classes A (decreased 8%) and B (decreased 4%).
Albee et al. (2006) evaluated the effects of a subchronic inhalation TCE exposure in
Fischer 344 rats (10/sex/group). Rats were exposed to 0, 250, 800, and 2,500 ppm TCE for 6
hr/day, 5 days/week for 13 weeks. TCE exposures were adequate to produce permanent auditory
6/22/2009
318
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 impairment even though TSEPs were unaffected. While TCE appears to be negative in
2 disrupting the trigeminal nerve, the TCE breakdown product, dichloroacetylene, does impair
3 trigeminal nerve function. Albee et al. (1997) showed that a single inhalation exposure of rats to
4 300-ppm dichloroacetylene, for 2.25 hr, disrupted trigeminal nerve evoked potentials for at least
5 4 days post exposure.
6
Table 4.2-2 Summary of animal trigeminal nerve studies
Reference
Exposure
route
Species/strain/
sex/number
Dose level/
Exposure
duration
NOAEL;
LOAELa
Effects
Barret et
al., 1991
Direct Gastric
Administration
Rat, Sprague-
Dawley, female,
7/group
0, 2.5 g/kg, acute
administration
17 mg/kg
dichloroacetylene
LOAEL: 2.5
g/kg
Morphometric analysis was
used for analyzing the
trigeminal nerve. Increase in
external and internal fiber
diameter as well as myelin
thickness was observed in the
trigeminal nerve after TCE
treatment.
Barret et
al., 1992
Direct Gastric
Administration
Rat, Sprague-
Dawley, female,
7/group
0, 2.5 g/kg; 1
dose/day, 5
days/wk, 10 wks
17 mg/kg
dichloroacetylene
LOAEL: 2.5
g/kg
Trigeminal nerve analyzed
using morphometric analysis.
Increased internode length and
fiber diameter in class A fibers
of the trigeminal nerve
observed with TCE treatment.
Changes in fatty acid
composition also noted.
Albee et
al., 1997
Inhalation
Rat, Fischer 344,
male, 6
0, 300 ppm-
dichloro-
acetylene, 2.25
hours
LOAEL:
300 ppm
dichloro-
acetylene
Dichloroacetylene (TCE
byproduct) exposure impaired
the trigeminal somatosensory
evoked potential (TSEP) up to 4
days post-exposure.
Albee et
al., 2006
Inhalation
Rat, Fischer
344, male and
female,
10/sex/group
0, 250, 800,
2,500 ppm
NOAEL:
2,500
ppm
No effect on trigeminal
somatosensory evoked
potentials (TSEPs) was noted at
any exposure level
a NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level)
7
6/22/2009
319
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.2.1.4 Discussion and Conclusions: TCE-induced trigeminal nerve impairment
Epidemiologic studies of exposure to TCE found impairment of trigeminal nerve
function, assessed by the blink reflex test or the trigeminal somatosensory evoked potential
(TSEP), in humans exposed occupationally by inhalation or environmentally by ingestion (see
Table 4.2-1). Mean inhalational exposures inferred from biological monitoring or from a range
of atmospheric monitoring in occupational studies was approximately 50 ppm to <150 ppm TCE
exposure. Residence location is the exposure surrogate in geographical-base studies of
contaminated water supplies with several solvents. Well water contaminant concentrations of
TCE ranged from <0.2 ppb to 10,000 ppb and do not provide an estimate of TCE concentrations
in drinking water to studied individuals. Two occupational studies, each including more than
100 subjects, reported statistically significant dose-response trends based on ambient TCE
concentrations, duration of exposure, and/or urinary concentrations of the TCE metabolite TCA
(Barret et al., 1984, 1987). Three geographical-based studies of environmental exposures to
TCE via contaminated drinking water are further suggestive of trigeminal nerve function
decrements; however, these studies are more limited than occupational studies due to questions
of subject selection. Both exposed subjects who were litigants and control subjects who may not
be representative of exposed (Kilburn and Warshaw, 1993; Kilburn et al., 2002a); referents in
Kilburn and Warshaw (1993) were histology technicians and subjects in a previous study of
formaldehyde and other solvent exposures and neurobehavioral effects (Kilburn et al., 1987;
Kilburn and Warshaw, 1992). Results were mixed in a number of smaller studies. Two of these
studies reported changes in trigeminal nerve response (Mhiri et al., 2004; Barret et al., 1982),
including evidence of a correlation with duration of exposure and increased latency in one study
(Mhiri et al., 2004). Ruitjen et al. (1991) reported no significant change in the blink reflex, but
did report an increase in the latency of the masseter reflex, which also may reflect effects on the
trigeminal nerve. Two other studies reported no observed effect on trigeminal nerve impairment,
but the authors failed to provide assessment of trigeminal nerve function (El-Ghawabi et al.,
1973) or there was not a control (nonexposed) group included in the study (Rasmussen, 1993c).
Therefore, because of limitations in statistical power, the possibility of exposure
misclassification, and differences in measurement methods, these studies are not judged to
provide substantial evidence against a causal relationship between TCE exposure and trigeminal
nerve impairment. Overall, the weight of evidence supports a relationship between TCE
exposure and trigeminal nerve dysfunction in humans.
Impairment of trigeminal nerve function is observed in studies of laboratory animal
studies. Although one subchronic animal study demonstrated no significant impairment of
trigeminal nerve function following TCE exposure up to 2,500 ppm (NOAEL; Albee et al.,
6/22/2009
320
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
2006), morphological analysis of the nerve revealed changes in its structure (Barret et al., 1991,
1992). However, the dose at which an effect was observed by Barret et al. (1991, 1992) was
high (2,500 mg/kg-day—LOAEL) compared to any reasonable occupational or environmental
setting, although no lower doses were used. The acute or subchronic duration of these studies, as
compared to the much longer exposure duration in many of the human studies, may also
contribute to the apparent disparity between the epidemiologic and (limited) laboratory animal
data.
The subchronic study of Barret et al. (1992) and the acute exposure study of Albee et al.
(1997) also demonstrated that dichloroacetylene, a (ex-vivo) TCE degradation product, also
induces trigeminal nerve impairment, at much lower doses than TCE. It is possible that under
some conditions, co-exposure to dichloroacetylene from TCE degradation may contribute to the
changes observed to be associated with TCE exposure in human studies, and this issue is
discussed further below in Section 4.2.10.
Overall evidence from numerous epidemiologic studies supports a conclusion that TCE
exposure induces trigeminal nerve impairment in humans. Laboratory animal studies provide
limited additional support, and do not provide strong contradictory evidence. Persistence of
these effects after cessation of exposure cannot be determined since exposure was ongoing in the
available human and laboratory animal studies.
4.2.2 Auditory Effects
4.2.2.1 Auditory Function: Human Studies
The TCE Subregistry from the National Exposure Registry (NER) developed by the
Agency for Toxic Substances Disease Registry (ATSDR) was the subject of three studies (Burg
et al., 1995, 1999; ATSDR, 2003). A fourth study (Rasmussen et al., 1993c) of degreasing
workers exposed to either TCE or CFC 113 also indirectly evaluated auditory function. These
studies are discussed below and presented in detail in Table 4.2-3.
Burg et al. (1995, 1999) reviewed the effects of TCE on 4,281 individuals (TCE
Subregistry) residentially exposed to this solvent for more than 30 consecutive days. Face-to-
face interviews were conducted with the TCE subregistry population and self-reported hearing
loss was evaluated based on personal assessment through the interview (no clinical evaluation
was conducted). TCE registrants that were 9 years old or younger had a statistically significant
increase in hearing impairment as reported by the subjects. The relative risk (RR) in this age
6/22/2009
321
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
group for hearing impairments was 2.13 (95% CI: 1.12-4.06) which decreased to 1.12 (95% CI:
0.52-2.24) for the 10-17 age group and 0.32 (95% CI: 0.10-1.02) for all older age groups. A
statistically significant association (when adjusted for age and sex) was found between duration
of exposure, in these studies this was length of residency, and reported hearing impairment. The
odds ratio (OR) was 2.32 (95% CI: 1.18-4.56) for subjects exposed to TCE > 2 years and < 5
years, 1.17 (95% CI: 0.55-2.49) for exposure > 5 years and < 10 years, 2.46 (95% CI:
1.30-5.02) for exposure durations greater than 10 years.
ATSDR (2003) conducted a follow-up study to the TCE subregistry findings (Burg et al.,
1995, 1999) and focused on the subregistry children located in Elkhart, IN, Rockford, IL and
Battle Creek, MI using clinical tests for oral motor, speech, and hearing function. Exposures
were modeled using tap water TCE concentrations and geographic information system (GIS) for
spatial interpolation, and LaGrange for temporal interpolation to estimate exposures from
gestation to 1990 across the area of subject residences. Modeled data were used to estimate
lifetime exposures (ppb-years) to TCE in residential wells. The median TCE exposure for the
children was estimated from drinking water as 23 ppb per year of exposure (ranging from 0-702
ppb per year). Approximately 20 percent (ranged from 17-21%) depending on ipsilateral or
contralateral test reflex) of the children in the TCE subregistry and 5—7% in the control group
exhibited an abnormal acoustic reflex (involuntary muscle contraction that measures movement
of the stapedius muscle in the middle ear following a noise stimulus) which was statistically
significant (p = 0.003). Abnormalities in this reflex could be an early indicator of more serious
hearing impairments. No significant decrements were reported in the pure tone and typanometry
screening.
Rasmussen et al. (1993b) used a psychometric test to measure potential auditory effects
of TCE exposure in an occupational study. Results from 96 workers exposed to TCE and other
solvents were presented in this study. Details of the exposure groups and exposure levels are
provided in Table 4.2-3. The acoustic motor function test was used for evaluation of auditory
function. Significant decrements (p < 0.05) in acoustic motor function performance scores
(average decrement of 2.5 points on a 10 point scale) was reported for TCE exposure.
6/22/2009
322
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.2-3 Summary of human auditory function studies
Reference
Subjects
Exposure
Effect
ATSDR, 2003
116 children, under 10
yrs of age, residing near
6 Superfund sites.
Further study of children
in Burg et al. (1995,
1999).
Control: 182 children
TCE and other solvents in
ground water supplies.
Exposures were modeled
using tap water TCE
concentrations and
geographic information
system (GIS) for spatial
interpolation, and
LaGrange for temporal
interpolation to estimate
exposures from gestation to
1990 across the area of
subject residences.
Control = 0 ppb; low
exposure group = 0 < 23
ppb-years; and high
exposure group = >23 ppb-
years
Auditory screening revealed increased
incidence of abnormal middle ear
function in exposed groups as indicated
from acoustic reflex test. Adjusted odds
ratios for right ear ipsilateral acoustic
reflects: control, OR = 1.0, low exposure
group, OR = 5.1, p < 0.05; high
exposure group, OR = 7.2, p < 0.05.
ORs adjusted for age, sex, medical
history and other chemical
contaminants. No significant
decrements reported in the pure tone and
typanometry screening.
Burg et al., 1995
From an NHIS TCE
subregistry of 4,281
(4,041 living & 240
deceased) residents
Environmentally exposed
to TCE and other solvents
via well water in Indiana,
Illinois, & Michigan;
Increase in self-reported hearing
impairments for children < 9 yrs.
Burg et al., 1999
3,915 white registrants
Mean age 34 yrs (SD =
19.9 yrs.);
Cumulative TCE exposure
subgroups: <50 ppb,
n = 2,867; 50-500 ppb,
n = 870; 500-5,000 ppb,
n= 190; >5,000 ppb,
n = 35;
Exposure duration
subgroups: <2 yrs, 2-5
yrs, 5-10 yrs., > 10
yrs.;
A statistically significant
association (adjusted for age and sex)
between duration of exposure and self-
reported hearing impairment
was found.
6/22/2009
323
-------
1
2
3
4
5
6
7
8
9
10
11
12
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Rasmussen et al.,
1993b
96 Danish metal
degreasers. Age range:
19-68 yrs;
No unexposed controls;
low exposed group is
referent
Average exposure
duration: 7.1 yrs.);
range of full-time
degreasing: 1 month to
36 yrs. Exposure to
TCE or and CFC 113.
1) Low exposure:
n = 19, average full-
time exposure 0.5 yrs
2) Medium exposure:
n = 36, average full-
time exposure 2.1 yrs.
3) High exposure:
n = 41, average full-
time exposure 11 yrs.
Mean U-TCA in high
exposure group = 7.7
mg/L (max = 26.1
mg/L);
Auditory impairments noted through
several neurological tests.
Significant relationship of exposure was
found with Acoustic-motor function
(p < 0.001), Paced Auditory Serial
Addition Test (p < 0.001), Rey Auditory
Verbal Learning Test (p< 0.001),
4.2.2.2 Auditory Function: Laboratory Animal Studies
The ability of trichloroethylene (TCE) to permanently disrupt auditory function and
produce abnormalities in inner ear histopathology has been demonstrated in several studies using
a variety of test methods. Two different laboratories have identified NOAELs following
inhalation exposure for auditory function of 1,600 ppm for 12 hr/day for 13 weeks in Long
Evans rats (n = 6-10) (Rebert et al., 1991) and 1,500 ppm for 18 hr/day, 5 days/week for 3
weeks in Wistar-derived rats (n = 12) (Jaspers et al., 1993). The LOAELs identified in these and
similar studies are 2,500-4,000 ppm TCE for periods of exposure ranging from 4 hr/day for 5
days to 12 hr/day for 13 weeks (e.g. Muijser et al., 2000; Rebert et al., 1995, 1993; Crofton et al.,
1994; Crofton and Zhao, 1997; Fechter et al., 1998; Boyes et al., 2000; Albee et al., 2006).
Rebert et al. (1993) estimated acute blood TCE levels associated with permanent hearing
6/22/2009
324
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
impairment at 125 [j,g/mL by methods that probably underestimated blood TCE values (rats were
anaesthetized using 60% CO2). A summary of these studies is presented in Table 4.2-4.
Reflex modification was used in several studies to evaluate the auditory function in TCE-
exposed animals (Jaspers et al., 1993; Muijser et al., 2000; Fechter et al., 1998; Crofton and
Zhao, 1993; Crofton et al., 1994; Crofton and Zhou, 1997; Boyes et al., 2000; Yamamura et al.,
1983). These studies collectively demonstrate significant decreases in auditory function at mid-
frequency tones (8-20 kHz tones) for TCE exposures greater than 1,500 ppm after acute, short-
term, and chronic durations. Only one study (Yamamura et al., 1983) did not demonstrate
impairment in auditory function from TCE exposures as high as 17,000 ppm for 4 hours/day over
5 days. This was the only study to evaluate auditory function in guinea pigs, whereas the other
studies used various strains of rats. Despite the negative finding in Yamamura et al. (1983),
auditory testing was not performed in an audiometric sound attenuating chamber and extraneous
noise could have influenced the outcome. It is also important to note that the guinea pig has
been reported to be far less sensitive than the rat to the effects of ototoxic aromatic hydrocarbons
such as toluene.
Crofton and Zhao (1997) also presented a benchmark dose for which the calculated dose
of TCE would yield a 15 dB loss in auditory threshold. This benchmark response was selected
because a 15 dB threshold shift represents a significant loss in threshold sensitivity for humans.
The benchmark concentrations for a 15 dB threshold shift are 5,223 ppm for 1 day, 2,108 ppm
for 5 days, 1,418 ppm for 20 days and 1,707 ppm for 65 days of exposure. While more sensitive
test methods might be used and other definitions of a benchmark effect chosen with a strong
rationale, these data provide useful guidance for exposure concentrations that do yield hearing
loss in rats.
Brainstem auditory-evoked potentials (BAERs) were also measured in several studies
(Rebert et al., 1991, 1993, 1995; Albee et al., 2006) following at exposures ranging from 3-13
weeks. Rebert et al. (1991) measured BAERs in male Long Evans rats (n = 10) and F344 rats
(n = 4-5) following stimulation with 4, 8, and 16 kHz sounds. The Long-Evans rats were
exposed to 0, 1,600, or 3,200 ppm TCE, 12 hour/day for twelve weeks and the F344 rats were
exposed to 0, 2,000, or 3,200 ppm TCE, 12 hours/day for three weeks. BAER amplitudes were
significantly decreased at all frequencies for F344 rats exposed to 2,000 and 3,000 ppm TCE and
for Long Evans rats exposed to 3,200 ppm TCE. These data identify a LOAEL at 2,000 ppm for
the F344 rats and a NOAEL at 1,600 ppm for the Long Evans rats. In subsequent studies Rebert
et al. (1993, 1995) again demonstrated TCE significantly decreases BAER amplitudes and also
6/22/2009
325
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 significantly increases the latency of appearance. Similar results were obtained by Albee et al.
2 (2006) for male and female F344 rats exposed to TCE for 13 weeks. The NOAEL for this study
3 was 800 ppm based on ototoxicity at 2,500 ppm.
4
5 Notable physiological changes were also reported in a few auditory studies. Histological
6 data from cochleas in Long-Evans rats exposed to 4,000 ppm TCE indicated that there was a loss
7 in spiral ganglion cells (Fechter et al., 1998). Similarly, there was an observed loss in hair cells
8 in the upper basal turn of the cochlea in F344 rats exposed to 2,500 ppm TCE (Albee et al.,
9 2006).
10
Table 4.2-4 Summary of Animal Auditory Function Studies
Reference
Exposure
route
Species/strain/
sex/number
Dose level/
Exposure
duration
NOAEL;
LOAELa
Effects
Rebert et al.,
1991
Inhalation
Rat, Long Evans,
male, 10/group
Long Evans: 0,
1,600, 3,200
ppm; 12 hr/day,
12 weeks
Long Evans:
NOAEL:
1,600 ppm;
LOAEL:
3,200 ppm
Brainstem auditory evoked
responses (BAERs) were
measured. Significant decreases
in B AER amplitude and an
increase in latency of
appearance of the initial peak
(PI).
Rat, F344, male,
4-5/group
F344: 0, 2,000,
3,200 ppm; 12
hr/day, 3 weeks
F344:
LOAEL:
2,000 ppm
Rebert et al.,
1993
Inhalation
Rat, Long Evans,
male, 9/group
0, 2,500, 3,000,
3,500 ppm; 8
hr/day, 5 days
NOAEL:
2,500 ppm
LOAEL:
3,000 ppm
BAERs were measured 1-2
weeks post-exposure to assess
auditory function. Significant
decreases in BAERs were noted
with TCE exposure.
Rebert et al.,
1995
Inhalation
Rat, Long Evans,
male, 9/group
0, 2,800 ppm; 8
hr/day, 5 days
LOAEL:
2,800 ppm
BAER measured 2-14 days
post-exposure at a 16 kHz tone.
Hearing loss ranged from 55-85
dB.
Crofton et
al., 1994
Inhalation
Rat, Long Evans,
male, 7-8/group
0, 3,500 ppm
TCE; 8 hr/day, 5
days
LOAEL:
3,500 ppm
BAER measured and auditory
thresholds determined 5-8
weeks post-exposure. Selective
impairment of auditory function
for mid-frequency tones (8 and
16 kHz)
6/22/2009
326
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Crofton and
Zhou, 1997;
Boyes et al.,
2000
Inhalation
Rat, Long Evans,
male, 9-12/group
0, 4,000, 6,000,
8,000 ppm; 6
hours
NOAEL:
6,000 ppm
LOAEL:
8,000 ppm
Auditory thresholds as measured
by BAERs for the 16 kHz tone
increased with TCE exposure.
Measured 3-5 weeks post
exposure.
Rat, Long Evans,
male, 8-10/group
0, 1,600, 2,400,
3,200 ppm; 6
hr/day, 5 days
NOAEL:
2,400 ppm
LOAEL:
3,200 ppm
Rat, Long Evans,
male, 8-10/group
0, 800, 1,600,
2,400, 3,200
ppm; 6 hr/day, 5
days/wk, 4
weeks
NOAEL:
2,400 ppm
LOAEL:
3,200 ppm
Rat, Long Evans,
male, 8-10/group
0, 800, 1,600,
2,400, 3,200
ppm; 6 hr/day, 5
days/wk, 13
weeks
NOAEL:
1,600 ppm
LOAEL:
2,400 ppm
Fechter et
al., 1998
Inhalation
Rat, Long Evans,
male, 12/group
0, 4,000 ppm; 6
hr/day, 5 days
LOAEL:
4,000 ppm
Cochlear function measured 5-7
weeks after exposure. Loss of
spiral ganglion cells noted.
Three weeks post-exposure,
auditory function was
significantly decreased as
measured by compound action
potentials and reflex
modification.
Jaspers et al.,
1993
Inhalation
Rat, Wistar derived
WAG-Rii/MBL,
male, 12/group
0, 1,500, 3,000
ppm; 18 hr/day,
5 days/wk, 3
wks
NOAEL:
1,500 ppm
Auditory function assessed
repeatedly 1-5 weeks post-
exposure for 5, 20, and 35 kHz
tones; No effect at 5 or 35 kHz;
Decreased auditory sensitivity at
20 kHz, 3,000 ppm.
Muijser et
al., 2000
Inhalation
Rat, Wistar derived
WAG-Rii/MBL,
male, 8
0, 3,000 ppm;
18 hr/day, 5
days/wk, 3 wks
LOAEL:
3,000 ppm
Auditory sensitivity decreased
with TCE exposure at 4, 8, 16,
and 20 kHz tones. White noise
potentiated the decrease in
auditory sensitivity.
6/22/2009
327
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Albee et
al., 2006
Inhalation
Rat, Fischer 344,
male and
female,
10/sex/group
0, 250, 800,
2,500 ppm; 6
hr/day, 5
days/wk, 13
wks
NOAEL:
800 ppm
LOAEL:
2,500 ppm
Mild frequency specific hearing
deficits; Focal loss of cochlear
hair cells.
Yamamura
et al., 1983
Inhalation
Guinea Pig,
albino Hartley,
male,
7-10/group
0, 6,000, 12,000,
17,000 ppm; 4
hr/day, 5 days
NOAEL:
17,000
ppm
No change in auditory
sensitivity at any exposure level
as measured by cochlear action
potentials and microphonics.
Study was conducted in guinea
pig and species is less sensitive
to auditory toxicity than rats.
Studies were also not conducted
in a sound-isolation chamber
and effects may be impacted by
background noise.
a NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level)
4.2.2.3 Summary and Conclusion of Auditory Effects
Human and animal studies indicated that TCE produces persistent decrements in auditory
function. In the human epidemiological studies (ATSDR, 2003; Burg et al., 1995, 1999;
Rasmussen et al., 1993c) it is suggested that auditory impairments result from both an inhalation
and oral TCE exposure. A LOAEL of approximately 23 ppb-years TCE (extrapolated from
< 23 ppb-years group in the ATSDR, 2003) from oral intake is noted for auditory effects in
children. The only occupational study where auditory effects were seen reported mean urinary
trichloroacetic acid concentration, a non-specific metabolite of TCE, of 7.7 mg/L for the high
cumulative exposure group only (Rasmussen et al., 1993c). A NOAEL or a LOAEL for auditory
changes resulting from inhalational exposure to TCE cannot be interpolated from average U-
TCA concentration of subjects in the high exposure group because of a lack of detailed
information on long-term exposure levels and duration (Rasmussen et al., 1993c). Two studies
(Burg et al., 1995, 1999) evaluated self-reported hearing effects in people included in the TCE
subregistry comprised of people residing near Superfund sites in Indiana, Illinois, and Michigan.
In Burg et al. (1995), interviews were conducted with the TCE exposed population and it was
found that children aged 9 years or younger had statistically significant hearing impairments in
comparison to non-exposed children. This significant increase in hearing impairment was not
observed in any other age group that was included in this epidemiological analysis. This lack of
6/22/2009
328
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
effect in other age groups may suggest a common exposure such as drinking water to residents;
however, it may also suggest that children may be more susceptible than adults. In a follow-up
analysis, Burg et al. (1999) adjusted the statistical analysis of the original data (Burg et al., 1995)
for age and sex. When these adjustments were made, a statistically significant association was
reported self-reported for auditory impairment and duration of residence. These epidemiological
studies provided only limited information given their use of an indirect exposure metric of
residence location, no auditory testing of this studied population and self-reporting of effects.
ATSDR (2003) further tested the findings in the Burg studies (Burg et al., 1995, 1999) by
contacting the children that were classified as having hearing impairments in the earlier study
and conducting several follow-up auditory tests. Significant abnormalities were reported for the
children in the acoustic reflex test which suggested effects to the lower brainstem auditory
pathway with the large effect measure, the odds ratio, was reported for the high cumulative
exposure group. Strength of analyses was its adjustment for potential confounding effects of
age, sex, medical history and other chemical contaminants in drinking water supplies. The
ATSDR findings were important in that the results supported Burg et al. (1995, 1999).
Rasmussen et al. (1993b) also evaluated auditory function in metal workers with inhalation
exposure to either TCE or CFC 113. Results from tasks including an auditory element suggested
that these workers may have some auditory impairment. However, the tasks did not directly
measure auditory function.
Animals strongly indicated that TCE produces deficits in hearing and provides biological
context to the epidemiological study observations. Although there is a strong association
between TCE and ototoxicity in the animal studies, most of the effects began to occur at higher
inhalation exposures. NOAELs for ototoxicity ranged from 800-1,600 ppm for exposure
durations of at least 12 weeks (Albee et al., 2006; Crofton and Zhou, 1997; Boyes et al., 2000;
Rebert et al., 1991). Inhalation exposure to TCE was the route of administration in all the animal
studies. These studies either used reflex modification audiometry (Jaspers et al., 1993; Crofton
et al., 1994; Crofton and Zhou, 1997; Muijser et al., 2000) procedures or measured brainstem
auditory evoked potentials (Rebert et al., 1991, 1993, 1995) to evaluate hearing in rats.
Collectively, the animal database demonstrates that TCE produces ototoxicity at mid-frequency
tones (4-24 kHz) and no observed changes in auditory function were observed at either the low
(<4 kHz) or high (>24 kHz) frequency tones. Additionally, deficits in auditory effects were
found to persist for at least 7 weeks after the cessation of TCE exposure (Rebert et al., 1991;
Jaspers et al., 1993; Crofton and Zhou, 1997; Fechter et al., 1998; Boyes et al., 2000). Decreased
amplitude and latency were noted in the BAERs (Rebert et al., 1991, 1993, 1995) suggesting that
TCE exposure affects central auditory processes. Decrements in auditory function following
reflex modification audiometry (Jaspers et al., 1993; Crofton et al., 1994; Crofton and Zhou,
6/22/2009
329
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1997; Muijser et al., 2000) combined with changes observed in cochlear histopathology (Fechter
et al., 1998; Albee et al., 2006) suggest that ototoxicity is occurring at the level of the cochlea
and/or brainstem.
4.2.3 Vestibular function
4.2.3.1 Vestibular Function: Human Studies
The earliest reports of neurological effects resulting from TCE exposures focused on
subjective vestibular system symptoms, such as headaches, dizziness, and nausea. These
symptoms are subjective and self-reported. However, as they have been reported extensively in
the literature, there is little doubt that these effects can be caused by exposures to TCE.,
occupational exposures (Grandjean et al., 1955; Liu et al., 1988; Rasmussen et al., 1986; Smith
et al., 1970), environmental exposures (Hirsch et al., 1996), and in chamber studies (Stewart et
al., 1970; Smith etal., 1970).
Kylin et al. (1967) exposed 12 volunteers to 1,000 ppm (5,500mg/m3) TCE for two hours
in a 1.5x2x2 meters chamber. Volunteers served as their own controls since 7 of the 12 were
pre-tested prior to exposure and the remaining 5 were post-tested days after exposure. Subjects
were tested for optokinetic nystagmus, which was recorded by electronystogmography, that is,
"the potential difference produced by eye movements between electrodes placed in lateral angles
between the eyes." Venous blood was also taken from the volunteers to measure blood TCE
levels during the vestibular task. The authors concluded that there was an overall reduction in
the limit ("fusion limit") to reach optokinetic nystagmus when individuals were exposed to TCE.
Reduction of the "fusion limit" persisted for up to 2 hours after the TCE exposure was stopped
and the blood TCE concentration was 0.2 mg/100 mL.
4.2.3.2 Vestibular function: Laboratory animal data:
The effect of TCE on vestibular function was evaluated by either (i) promoting
nystagmus (vestibular system dysfunction) and comparing the level of effort required to achieve
nystagmus in the presence and absence of TCE or (ii) using an elevated beam apparatus and
measuring the balance. Overall, it was found that TCE disrupts vestibular function as presented
below and summarized in Table 4.2-5.
Niklasson et al. (1993) showed acute impairment of vestibular function in male- and
female-pigmented rats during acute inhalation exposure to TCE (2,700--7,200 ppm) and to
tricholoroethane (500-2,000 ppm). Both of these agents were able to promote nystagmus during
6/22/2009
330
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
optokinetic stimulation in a dose related manner. While there were no tests performed to assess
persistence of these effects, Tham et al. (1979, 1984) did find complete recovery of vestibular
function in rabbits (n= 19) and female Sprague-Dawley rats (// = 11) within minutes of
terminating a direct arterial infusion with TCE solution.
The finding that trichloroethylene can yield transient abnormalities in vestibular function
is not unique. Similar impairments have also been shown for toluene, styrene, along with
trichloroethane (Niklasson et al., 1993) and by Tham et al. (1984) for a broad range of aromatic
hydrocarbons. The concentration of TCE in blood at which effects were observed for TCE (0.9
mM/L) was quite close to that observed for most of these other vestibulo-active solvents.
Table 4.2-5 Summary of mammalian sensory studies—vestibular and visual systems
Reference
Exposure route
Species/strain/
sex/number
Dose level/
Exposure duration
NOAEL;
LOAEL3
Effects
Vestibular System Studies
Tham et
al., 1979
Intravenous
Rabbit, strain
unknown, sex
unspecified, 19
1-5 mg/kg /min
Positional nystagmus
developed once blood
levels reached 30 ppm
Tham et
al., 1984
Intravenous
Rat, Sprague-
Dawley, female,
11
80 |ig/kg/min
Excitatory effects on the
vestibule-oculomotor
reflex. Threshold effect at
blood [TCE] of 120 ppm or
0.9 mM/L.
Niklasson
et al., 1993
Inhalation
Rat, strain
unknown, male
and female, 28
0, 2,700, 4,200,
6,000, 7,200 ppm; 1
hour
LOAEL:
2,700 ppm
Increased ability to produce
nystagmus.
Umezu et
al., 1997
Intraperitoneal
Mouse, ICR,
male, 116
0, 250, 500, 1,000
mg/kg, single dose
and evaluated 30 min
post-administration
NOAEL:
250 mg/kg
LOAEL:
500 mg/kg
Decreased equilibrium and
coordination as measured by
the Bridge test (staying time
on an elevated balance
beam).
11 NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level)
4.2.3.3 Summary and Conclusions for the Vestibular Function Studies
Studies of TCE exposure in both humans and animals reported abnormalities in vestibular
function. Headaches, dizziness, nausea, motor incoordination, among other subjective symptoms
are reported in occupational epidemiological studies of TCE exposure (Grandjean et al., 1955;
Liu et al., 1988; Rasmussen et al., 1986; Smith et al., 1970; Hirsch et al., 1996; Stewart et al.,
6/22/2009
331
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1970). One human exposure study (Kylin et al., 1967) found that vestibular function was
affected following an acute exposure to 1,000-ppm TCE (LOAEL). Individuals had a decreased
threshold to reach nystagmus than when exposed to TCE than to air. Animal studies also
evaluated the threshold to reach nystagmus and reported that TCE decreased the threshold to
produce nystagmus in rats (LOAEL: 2,700 ppm; Tham et al., 1984; Niklasson et al., 1993) and
rabbits (Tham et al., 1983).
4.2.4 Visual Effects
4.2.4.1 Visual Effects: Human Studies
Visual impairment in humans has been demonstrated following exposures through
groundwater (Kilburn, 2002a; Reif et al., 2003), from occupational exposure through inhalation
(Rasmussen et al., 1993b; Troster and Ruff, 1990) and from a controlled inhalation exposure
study (Vernon and Ferguson, 1969). Visual functions such as color discrimination and
visuospatial learning tasks are impaired in TCE-exposed individuals. Additionally, an acute
exposure can impair visual depth perception. Details of the studies are provided below and
summarized in Table 4.2-6.
Geographical-based studies utilized color discrimination and contrast sensitivity tests to
determine the effect of TCE exposure on vision. In these studies it was reported that TCE
exposure significantly increased color discrimination errors (Kilburn, 2002a) or decreases in
contrast sensitivity tests approached statistical significance after adjustments for several possible
confounders (p = 0.06 or 0.07; Reif et al., 2003). Exposure in both studies is poorly
characterized, TCE is one of several contaminants in drinking water supplies and neither study
provides an estimate of an individual's exposure to TCE.
Rasmussen et al. (1993b) evaluated visual function in 96 metal workers, working in
degreasing at various factories and with exposure to TCE or CFC 113. Visual function was
tested through the visual gestalts test (visual perception) and a visual recall test. In the visual
gestalts test, the number of total errors significantly increased from the low group (3.4 errors) to
the high exposure group (6.5 errors; p = 0.01). No significant changes were observed in the
visual recall task. Troster and Ruff (1990) presented case studies conducted on two
occupationally exposed workers to TCE. Both patients presented with a visual-spatial task and
neither could complete the task within the number of trials allowed suggesting visual function
deficits as a measure of impaired visuospatial learning.
6/22/2009
332
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
2 In a chamber exposure study (Vernon and Ferguson, 1969), eight male volunteers (ages
3 21-30) were exposed to 0, 100, 300, and 1,000 ppm TCE for 2 hours. Each individual was
4 exposed to all TCE concentrations and a span of at least three days was given between
5 exposures. When the individuals were exposed to 1,000-ppm TCE (5,500 mg/m3), significant
6 abnormalities were noted in depth perception as measured by the Howard-Dolman test
7 (p< 0.01). There were no effects on the flicker fusion frequency test (threshold frequency at
8 which the individual sees a flicker as a single beam of light) or on the form perception illusion
9 test (volunteers presented with an illusion diagram).
10
Table 4.2-6 Summary of human visual function studies
Reference
Subjects
Exposure
Effect
Kilburn, 2002a
236 residents near a
microchip plant in
Phoenix, AZ;
Controls: 67 local
referents from Phoenix,
AZ and 161 regional
referents from
Wickenburg, AZ
TCE, TCA, 1, 1-DCE,
1, 2-DCE, PCE, and
VC detected in well
water up to 260,000
ppm; TCE
concentrations in well
water were 0.2-10,000
ppb. Exposure
duration ranged from
2-37 years.
Exposure duration
ranged from 2 to 37
years.
Color discrimination errors were
increased among residents compared to
regional referents (p < 0.01). No
adjustment for possible confounding
factors.
6/22/2009
333
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reif et al., 2003
143 residents of the
Rocky Mountain Arsenal
community of Denver
Referent group at lowest
concentration (<5 ppb).
Exposure modeling of TCE
concentrations in
groundwater and in
distribution system to
estimate mean TCE
concentration by census
block of residence. High
exposure group >15 ppb,
Medium exposure group >
5 ppb and <15 ppb
Low exposure referent
group <5 ppb
Contrast sensitivity test performances (C
and D) was marginally statistically
significant (p = 0.06 and 0.07,
respectively). No significant effects
reported for the Benton visual retention
test. Significant decrements (p = 0.02)
were reported in the Benton visual
retention test when stratified with
alcohol consumption.
Rasmussen et al.,
1993b
96 Danish metal
degreasers. Age range:
19-68; No unexposed
controls; low exposure
group was referent
Average exposure
duration: 7.1 yrs.);
range of full-time
degreasing: 1 month to
36 yrs. Exposure to
TCE orCFC 113.
1) Low exposure:
n= 19, average full-
time expo 0.5 yrs
2) Medium exposure:
n = 36, average full-
time exposure 2.1 yrs.
3) high exposure:
n = 41, average full-
time exposure 11 yrs.
TCA in high exposure
group = 7.7 mg/L
(max = 26.1 mg/L);
Statistically significant relationship of
exposure was found with the Visual
Gestalts learning and retention test
(cognitive test) indicating deficits in
visual performance.
Troster and Ruff,
1990
2 occupationally TCE-
exposed workers;
Controls: 2 groups of
n = 30 matched controls;
(all age & education
matched)
Exposure
concentration
unknown; Exposure
duration, 3-8 months.
Both workers experienced impaired
visuospatial learning.
6/22/2009
334
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Vernon and
Ferguson, 1969
8 male volunteers age
range 21-30; self
controls
0, 100 ppm, 300 ppm
and 1,000 ppm of TCE
for two hours
Statistically significant effects on visual
depth perception as measured by the
Howard-Dolman test. NOAEL: 300
ppm; LOAEL: 1,000 ppm; No
significant changes in any of the other
visual test measurements.
4.2.4.2 Visual Effects: Laboratory animal data
Changes in visual function have been demonstrated in animal studies during acute (Boyes
et al., 2003, 2005) and subchronic exposure (Rebert et al., 1991; Blain et al., 1994). In these
studies, the effect of TCE on visual evoked responses to patterns (Boyes et al., 2003, 2005;
Rebert et al., 1991) or a flash stimulus (Rebert et al., 1991; Blain et al., 1994) were evaluated.
Overall, the studies demonstrated that exposure to TCE results in significant changes in the
visual evoked response, which is reversible once TCE exposure is stopped. Details of the studies
are provided below and are summarized in Table 4.2-7.
Boyes et al. (2003, 2005) exposed adult, male Long-Evans rats were to TCE in a head-
only exposure chamber while pattern onset/offset visual evoked potentials (VEPs) were
recorded. Exposure conditions were designed to provide concentration x time products of 0
ppm/h (0 ppm for 4 h) or 4,000 ppm/h (see Table 4.2-7 for more details). VEP amplitudes were
depressed by TCE exposure during the course of TCE exposure. The degree of VEP depression
showed a high correlation with the estimated brain TCE concentration for all levels of
atmospheric TCE exposure.
In a subchronic exposure study, Rebert et al. (1991) exposed male Long Evans rats to
1,600 or 3,200-ppm TCE, for 12 weeks, 12 hours/day. No significant changes in flash evoked
potential measurements were reported following this exposure paradigm. Decreases in pattern
reversal visual evoked potentials (N1P1 amplitude) reached statistical significance following 6,
9, and 12 weeks of exposure. The drop in response amplitude ranged from approximately 20%
after 8 weeks to nearly 50% at week 14 but recovered completely within 1 week post-exposure.
This transient effect of TCE on the peripheral visual system has also been reported by
Blain (1994) in which New Zealand albino rabbits were exposed by inhalation to 350 ppm and
700-ppm TCE 4 hrs/day, 4 days/week for 12 weeks. Electroretinograms (ERG) and oscillatory
potentials (OPs) were recorded weekly under mesopic conditions. Recordings from the 350 and
700-ppm exposed groups showed a significant increase in the amplitude of the a- and b-waves
6/22/2009
335
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 (ERG). The amplitude of the OPs was significantly decreased at 350 ppm (57%) and increased
2 at 700 ppm (117%). These electroretinal changes returned to pre-exposure conditions within six
3 weeks after the inhalation stopped.
4
Table 4.2-7 Summary of Animal Visual System Studies
Reference
Exposure route
Species/strain/
sex/number
Dose level/
Exposure duration
NOAEL;
LOAELa
Effects
Rebert et
al., 1991
Inhalation
Rat, Long Evans,
male, 10/group
0, 1,600, 3,200 ppm;
12 hr/day, 12 weeks
NOAEL:
1,600 ppm
Significant amplitude
decreases in pattern reversal
evoked potentials (N1P1
amplitude) at 6, 9, and 12
weeks.
Boyes et
al., 2003
Inhalation
Rat, Long Evans,
male,
9-10/group
0 ppm, 4 hours; 1,000
ppm, 4 hours; 2,000
ppm, 2 hours;
3,000 ppm, 1.3 hours
4,000 ppm, 1 hour
LOAEL:
1,000 ppm,
4 hours
Visual function
significantly affected as
measured by decreased
amplitude (F2) in Fourier-
transformed visual evoked
potentials. Peak brain TCE
concentration correlated
with dose response.
Boyes et
al., 2005
Inhalation
Rat, Long Evans,
male,
8-10/group
0 ppm, 4 hours;
500 ppm, 4 hours;
1,000 ppm, 4 hours;
2,000 ppm, 2 hours;
3,000 ppm, 1.3 hours
4,000 ppm, 1 hour;
5,000 ppm, 0.8 hour
LOAEL:
500 ppm, 4
hours
Visual function
significantly affected as
measured by decreased
amplitude (F2) in Fourier-
transformed visual evoked
potentials. Peak brain TCE
concentration correlated
with dose response.
Blain et al.,
1994
Inhalation
Rabbit, New
Zealand albino,
male, 6-8/group
0, 350, 700 ppm; 4
hr/day, 4 days/wk, 12
wks
LOAEL:
350 ppm
Significant effects noted in
visual function as measured
by electroretinogram (ERG)
and oscillatory potentials
(OP) immediately after
exposure. No differences in
ERG or OP measurements
were noted at 6 weeks post-
TCE exposure.
11 NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level)
5
6/22/2009
336
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.2.4.3 ,Vwmmary and Conclusion of Visual Effects
Changes in visual function are reported in human studies. Although central visual
function was not evaluated in the human studies (such as electroretinograms, evoked potential
measurements), clinical tests indicated deficits in color discrimination (Kilburn, 2002a), visual
depth perception (Vernon and Ferguson, 1969) and contrast sensitivity (Reif et al., 2003). These
changes in visual function were observed following both an acute exposure (Vernon and
Ferguson, 1969) and residence in areas with groundwater contamination with TCE and other
chemicals (Kilburn, 2002a; Reif et al., 2003). The exposure assessment approach of Reif et al.
(2003), who adopted exposure modeling and information on water distribution patterns, is
considered superior to that of Kilburn (2002a) who used residence location as a surrogate for
exposure. In the one acute, inhalation study (Vernon and Ferguson, 1969), a NOAEL of 300
ppm and a LOAEL of 1,000 ppm for 2 hours was reported for visual effects. A NOAEL is not
available from the drinking water studies since well water TCE concentration is a poor surrogate
for an individual's TCE ingestion (Kilburn, 2002a) and limited statistical analysis comparing
high exposure group to low exposure group (Reif et al., 2003).
Animal studies have also demonstrated changes in visual function. All of the studies
evaluated central visual function by measuring changes in evoked potential response following a
visual stimulus that was presented to the animal. Two acute exposure inhalation studies (Boyes
et al., 2003, 2005) exposed Long Evans rats to TCE based on a concentration x time schedule
(Haber's law) and reported decreases in visual evoked potential amplitude. All of the exposures
from these two studies resulted in decreased visual function with a LOAEL of 500 ppm for 4
hours. Another important finding that was noted is the selection of the appropriate dose metric
for visual function changes following an acute exposure. Boyes et al. (2003, 2005) found that
among other potential dose metrics, brain TCE concentration was best correlated with changes in
visual function as measured by evoked potentials under acute exposure conditions. Two
subchronic exposure studies (Rebert et al., 1991; Blain et al., 1994) demonstrated visual function
changes as measured by pattern reversal evoked potentials (Rebert et al., 1991) or
electroretinograms/oscillatory potentials (Blain et al., 1994). Unlike the other three visual
function studies conducted with rats, Blain et al. (1994) demonstrated these changes in rabbits.
Significant changes in ERGs and oscillatory potentials were noted following a 12 week exposure
at 350 ppm (LOAEL) in rabbits (Blain et al., 1994) and in rats exposed to 3,200 ppm TCE for 12
weeks there were significant decreases in pattern reversal evoked potentials but no effect was
noted in the 1,600 ppm exposure group (Rebert et al., 1991). Both subchronic studies examined
visual function following an exposure-free period of either 2 weeks (Rebert et al., 1991) or 6
6/22/2009
337
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
weeks (Blain et al., 1994) and found that visual function returned to pre-exposure levels and the
changes are reversible.
4.2.5 Cognitive function
4.2.5.1 Cognitive Effects: Human Studies
Effects of TCE on learning and memory have been evaluated in populations
environmentally exposed to TCE through well water, in workers occupationally exposed through
inhalation and under controlled exposure scenarios. Details of the studies are provided in Table
4.2-8 and discussed briefly below. In the geographical-based studies (Kilburn and Warshaw,
1993; Kilburn, 2002a), cognitive function was impaired in both studies and was evaluated by
testing verbal recall and digit span memory among other measures. In Arizona residents
involved in a lawsuit (Kilburn and Warshaw, 1993), significant impairments in all three
cognitive measures were reported; verbal recall (p = 0.001), visual recall (p = 0.03) and digit
span test (p = 0.07), although a question exists whether the referent group was comparable to
exposed subjects and the study's lack of consideration of possible confounding exposures in
statistical analyses. Significant decreases in verbal recall ability was also reported in another
environmental exposure study where 236 residents near a microchip plant with TCE
concentration in well water ranging from 0.2-10,000 ppb (Kilburn, 2002a).
Cognitive impairments are assessed in the occupational exposure and case studies
(Rasmussen, 1993a, b; Troster and Ruff, 1990). In metal degreasers occupationally exposed to
TCE and CFC 113, significant cognitive performance decreases were noted in verbal recall
testing (p = 0.03) and verbal learning (p = 0.04; Rasmussen et al., 1993a). No significant effects
were found in the visual recall or digit span test for these workers. Troster and Ruff (1990)
reported decrements (no statistical analysis performed) in cognitive performance as measured in
verbal and visual recall tests that were conducted immediately after presentation (learning phase)
and one hour after original presentation (retention/memory phase) for two case studies.
Several controlled (chamber) exposure studies were conducted to cognitive ability during
TCE exposure and most did not find any significant decrements in the neurobehavioral
measurement. Only Salvini et al. (1971) found significant decrements in cognitive function. Six
males were exposed to 110 ppm (550 mg/m3) TCE for 4 hour intervals, twice per day.
Statistically significant results were observed for perception tests learning (p < 0.001), mental
fatigue (p < 0.01), subjects (p < 0.05); and CRT learning (p < 0.01), mental fatigue (p < 0.01),
6/22/2009
338
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 subjects (p < 0.05). Triebig et al. (1977a, b) exposed 7 total subjects (male and female) to 100
2 ppm TCE for 6 hours/day, 5 days/week and did not report any decreases in cognition but details
3 on the experimental procedures were not provided. Additionally, Gamberale et al. (1976) found
4 that subjects exposed to TCE as high as 194 ppm for 70 minutes did not exhibit any impairments
5 on a short term memory test in comparison to an air exposure.
6
Table 4.2-8 Summary of Human Cognition Effect Studies
Reference
Subjects
Exposure
Effect
Kilburn and
Warshaw, 1993
170 residents living in
Southwest Tucson with
TCE, other solvents, and
chromium in
groundwater.
Control: 68 residential
referents matched to
subjects from 2 previous
studies of waste oil and
oil refinery exposures.
>500 ppb of TCE in
well water before 1981
and 25 to 100 ppb
afterwards.
Exposure duration
ranged from 1 to 25
years
Decreased performance in the digit span
memory test and story recall ability.
Kilburn, 2002a
236 residents near a
microchip plant;
Controls: 67 local
referents from Phoenix,
AZ and 161 regional
referents from
Wickenburg, AZ
<0.2-10,000 ppb of
TCE, <0.2-260,000
ppb TCA, <0.2-6,900
ppb 1, 1-DCE,
<0.2-1,600 1, 2-DCE,
<0.2-23,000 ppb PCE,
<0.02-330 ppb VC in
well water.
Exposure duration
ranged from 2 to 37
years. Exposure
duration ranged from 2
to 37 years.
Cognitive effects decreased as measured
by lower scores on Culture Fair 2A,
vocabulary, grooved pegboard
(dominant hand), trail making test, and
verbal recall (i.e., memory).
6/22/2009
339
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Rasmussen,
1993a, b
96 Danish metal
degreasers. Age range:
19-68; No external
controls
Average exposure
duration: 7.1 yrs.);
range of full-time
degreasing: 1 month to
36 yrs.
1) Low exposure:
n= 19, average full-
time expo 0.5 yrs
2) Medium exposure:
n = 36, average full-
time exposure 2.1 yrs.
3) High exposure:
n = 41, average full-
time exposure 11 yrs.
TCA in high exposure
group = 7.7 mg/L
(max = 26.1 mg/L)
Cognitive impairment (psycho-organic
syndrome) prevalent in exposed
individuals. The incidence of this
syndrome was 10.5% in the low
exposure, 39.5% for medium exposure,
and 63.4% for high exposure. Age is a
confounder. Dose-response with 9 of 15
tests; Controlling for confounds,
significant relationship of exposure was
found with Acoustic-motor function
(p < 0.001), Paced Auditory Serial
Addition Test (p < 0.001), Rey Auditory
Verbal-Learning Test (p < 0.001),
vocabulary (p < 0.001) and visual
gestalts (p < 0.001); significant age
effects. Age is a confounder.
Troster and Ruff,
1990
2 occupationally TCE-
exposed workers;
Controls: 2 groups of
n = 30 matched controls;
(all age & education
matched
Exposure
concentration
unknown; Exposure
duration, 3-8 months.
Both TCE cases exhibited significant
deficits in verbal recall and visuospatial
learning.
Triebig, 1976
Controlled exposure
study 4 females, 3 males.
Controls: 4 females, 3
males
0, lOOppm
(550mg/m3), 6
hrs/day, 5 days.
There was no correlation seen between
exposed and unexposed subjects for any
measured psychological test results. No
methods description was provided.
Triebig, 1977a
7 men and 1 woman
occupationally exposed
with an age range from
23-38 years. No control
group.
50 ppm (260mg/m3).
Exposure duration not
reported.
The psychological tests showed no
statistically significant difference in the
results before or after the exposure-free
time period. No methods description
was provided.
6/22/2009
340
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Triebig, 1977b
Controlled exposure
study on 3 male and 4
female students.
Control: 3 male and 4
female students
0, 100 ppm
(550mg/m3), 6
hrs/day, 5 days
No significantly different changes were
obtained. No methods description was
provided.
Salvini et al., 1971
Controlled exposure
study 6 students, male.
Self used as control
TCE concentration
was llOppm for 4-
hour intervals, twice
per day. Oppm control
exposure for all as self
controls
Statistically significant results were
observed for perception tests learning
(p < 0.001) and CRT learning
(p<0.01).
Gamberale et al.,
1976
15 healthy men aged
20-31 yrs old.
Controls: Within
Subjects (15 self-
controls)
0 mg/m3, 540 mg/m3
(97ppm), 1,080 mg/m3
(194ppm), 70 minutes.
Repetition of the testing led to
a pronounced improvement in
performance as a result of the
training effect; No interaction effects
between exposure to TCE and training
Stewart et al.,
1970
130 (108 males, 22
females); Controls: 63
unexposed men
Trichloroacetic acid
(TCA) metabolite
levels in urine were
measured: 60.8% had
levels up to 20mg/L,
and 82.1% had levels
up to 60 mg/L.
No significant effect on cognitive tests
noted, but more effort required to
perform the test in exposed group.
Chalupa, 1960
Case study - Six
subjects. Average age
38.
No exposure data was
reported
80% of those with pathological EEG
displayed memory loss; 30% of those
with normal EEGs displayed memory
loss.
1
2 4.2.5.2 Cognitive Effects: Laboratory animal studies
3 Many reports have demonstrated significant differences in performance of learning tasks
4 such as the speed to complete the task. However, there is little evidence that learning and
5 memory function are themselves impaired by exposure. There are also limited data that suggest
6 alterations in the hippocampus of laboratory animals exposed to TCE. Given the important role
7 that this structure plays in memory formation, such data may be relevant to the question of
8 whether TCE impairs memory. The studies are briefly discussed below and details are provided
9 in Table 4.2-9.
6/22/2009
341
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Two studies (Kulig et al., 1987; Umezu et al., 1997) reported decreased performance in
operant-conditioning cognitive tasks for rodents. Kishi et al. (1993) acutely exposed Wistar rats
to TCE at concentrations of 250, 500, 1,000, 2,000 and 4,000 ppm for four hours. Rats exposed
to 250 ppm TCE and higher showed a significant decrease both in the total number of lever
presses and in avoidance responses compared with controls. The rats did not recover their pre-
exposure performance until about 2 hours after exposure. Likewise, Umezu et al. (1997)
reported a depressed rate of operant responding in male ICR strain mice (n = 6, exposed to all
TCE doses, see Table 4.2-9) in a conditioned avoidance task that reached significance with ip
injections of 1,000 mg/kg. Increased responding during the signaled avoidance period at lower
doses (250 and 500 mg/kg) suggests an impairment in ability to inhibit responding or failure to
attend to the signal.
Although cognitive impairments are noted, two additional studies indicate no change in
cognition with continuous TCE exposure or improvements in cognitive tasks. No decrements in
cognitive function as measured by the radial arm maze were observed in Mongolian gerbils
exposed continuously by inhalation to 320 ppm TCE for 9 months (Kjellstrand et al., 1980).
Improved performance was noted in a Morris swim test for weanling rats orally dosed with 5.5
mg/day for 4 weeks followed by 2 weeks of no exposure and an additional 2 weeks of 8.5
mg/day (Isaacson et al., 1990). This improved performance occurred despite a loss in
hippocampal myelination.
Table 4.2-9 Summary of Animal Cognition Effect Studies
Reference
Exposure route
Species/strain/
sex/number
Dose level/
Exposure duration
NOAEL;
LOAELa
Effects
Kjellstrand
etal., 1980
Inhalation
Gerbil,
Mongolian,
males and
females,
12/sex/dose
0, 320 ppm; 9 months,
continuous (24 hr/day)
except 1-2 hr/wk for
cage cleaning
NOAEL: 320
ppm
No significant effect
on spatial memory
(radial arm maze)
Isaacson et
al., 1990
Oral, drinking
water
Rat,
Sprague-Daw
ley, male
weanlings,
12/dose
1)0 mg/kg/day, 8 wks
2) 5.5 mg/day
(47mg/kg/dayb), 4
wks + 0 mg/kg/day, 4
NOAEL: 5.5
mg/day, 4
weeks—
spatial
learning
Decreased latency to
find platform in the
Morris water maze
(Group #3);
Hippocampal
demyelination
6/22/2009
342
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
wks
3) 5.5 mg/day, 4 wks
(47 mg/kg/dayb) + 0
mg/kg/day, 2 wks +
8.5 mg/day (24
mg/kg/dayb), 2 weeks
LOAEL: 5.5
mg/day—
hippocampal
demyelination
observed in all TCE
treated groups.
Kishi et al.,
1993
Inhalation
Rats, Wistar,
male, number
not specified
0, 250,500, 1,000,
2,000, 4,000 ppm, 4
hours
LOAEL: 250
ppm
Decreased lever
presses and
avoidance responses
in a shock avoidance
task
Umezu et
al., 1997
Intraperitoneal
Mouse, ICR,
male, 6
exposed to all
treatments
(repeated
exposure)
0, 125, 250, 500, 1,000
mg/kg, single dose and
evaluated 30 min post-
administration
NOAEL: 500
mg/kg
LOAEL:
l,000mg/kg
Decreased response
rate in an operant
response— condition
avoidance task.
Oshiro et
al., 2004
Inhalation
Rat, Long
Evans, male,
24
0, 1,600, 2,400 ppm;
6 hr/day, 5 days/wk,
4 weeks
NOAEL:
2,400 ppm
No change in
reaction time in
signal detection task
and when challenged
with amphetamine,
no change in
response from
control.
a NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level)
b mg/kg/day conversion estimated from average male Sprague-Dawley rat body weight from ages 21-49 days (118
g) for the 5.5 mg dosing period and ages 63-78 days (354 g) for the 8.5 mg dosing period.
1
2 4.2.5.3 Summary and Conclusions of Cognitive Function Studies
3 Human environmental and occupational exposure studies suggest impairments in
4 cognitive function. Kilburn and Warshaw (1993) and Kilburn (2002a) reported memory deficits
5 individuals. Significant impairments were found in visual and verbal recall and with the digit
6 span test. Similarly, in occupational exposure studies (Rasmussen et al., 1993a, b; Troster and
7 Ruff, 1990), short term memory tests indicated that immediate memory and learning were
8 impaired. In controlled exposure and/or chamber studies, two studies did not report any
9 cognitive impairment (Stewart et al., 1970; Gamberale et al., 1976) and one study (Salvini et al.,
10 1971) reported significant impairments in learning memory and complex choice reaction tasks.
6/22/2009
343
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
All of the controlled exposure studies were acute and/or short-term exposure studies and the
sensitivity of test procedures is unknown due to the lack of methodologic information provided
in the reports.
The animal studies measured cognitive function through spatial memory and operant
responding tasks. In the two studies where spatial memory was evaluated, there was either no
effect at 320 ppm TCE (Kjellstrand et al., 1980) or improved cognitive performance in weanling
rats at a dose of 5.5 mg/day for four weeks (Isaacson et al., 1990). Improved cognitive
performance was observed in weanling rats (Isaacson et al., 1990) and could be due to
continuing neurodevelopment as well as compensation from other possible areas in the brain
since there was a significant loss in hippocampal myelination. Significant decreases in operant
responding (avoidance/punished responding) during TCE exposure were reported in two studies
(Kishi et al., 1993; Umezu et al., 1997). When TCE exposure was discontinued operant
responding return to control levels and it is unclear if the significant effects are due to decreased
motor function or decreased cognitive ability.
4.2.6 Psychomotor Effects
There is considerable evidence in the literature for both animals and humans on
psychomotor testing although human and laboratory animal studies utilize very different
measures of motor behavior. Generally, the human literature employs a wide variety of
psychomotor tasks and assesses error rates and reaction time in the performance of the task. The
laboratory animal data, by contrast, tend to include unlearned naturalistic behaviors such as
locomotor activity, gait changes, and foot splay to assess neuromuscular ability.
4.2.6.1 Psychomotor effects: Human Studies
The effects of TCE exposure on psychomotor response have been studied primarily as a
change in reaction time (RT) with studies on motor dyscoordination resulting from TCE
exposure providing subjective reporting.
4.2.6.1.1 Reaction Time
Several studies have evaluated the effects of TCE on reaction time using simple and
choice reaction time tasks (SRT and CRT tasks). The studies are presented below and
summarized in more detail in Table 4.2-10.
6/22/2009
344
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Increases in reaction time were observed in environmental exposure studies by Kilburn
(2002a), Kilburn and Warshaw (1993), and Kilburn and Thornton (1996) as well as in an
occupational exposure study by Gun et al. (1978). All populations except that of Gun et al.
(1978) were exposed through groundwater contaminated as the result of environmental spills and
the exposure duration was for at least one year and exposure levels ranged from 0.2 to 10,000
ppb for the three studies. Kilburn and Warshaw (1993) reported that SRT significantly increased
from 281 ± 55 msec to 348 ± 96 msec in individuals (p < 0.0001). CRT of the exposed subjects
was 93 msec longer (p < 0.0001) than referents. Kilburn and Thornton (1996) evaluated SRT
and CRT function and also found similar increases in reaction time. The average SRT and CRT
for the combined control groups were 276 msec and 532 msec, respectively. These reaction
times increased in the TCE exposure group where the average SRT was 334 msec and CRT was
619 msec. Similarly, Kilburn (2002a) compared reaction times between 236 TCE-exposed
persons and the 161 unexposed regional controls. SRTs significantly increased from 283±63
msec in controls to 334 ±118 msec in TCE exposed individuals (p < 0.0001). Similarly, CRTs
also increased from 510 ± 87 msec to 619 ±153 msec with exposure to TCE (p < 0.0001).
No effect on SRT was reported in a geographical-based study by Reif et al. (2003). SRTs
were 301 msec for the lowest exposure group and 316 msec for the highest exposure group
(p = 0.42). When the SRT data was analyzed individuals that consumed at least on alcoholic
drink per month (n = 80), a significant increase (18 %,p< 0 .04) in SRT times were observed
between the lowest exposure and the highest exposure groups. In TCE exposed individuals who
did not consume alcohol (n = 55), SRTs decreased from 321 msec in the lowest exposed group to
296 msec in the highest exposed group, but this effect was not statistically significantly different.
A controlled exposure (chamber study) of 15 healthy men aged 20-31 yrs old, were exposed to
0, 540, and 1,080 mg/m3 TCE for 70 min or served as his own control, reported no statistically
significant differences with the SRT or CRT tasks. However, in the RT-Addition test the level of
performance varied between the different exposure conditions (F(2.24) = 4.35; p < 0.05) and
between successive measurement occasions (F(2.24) = 19.25;p < 0.001).
6/22/2009
345
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.2-10 Summary of Human Choice Reaction Time Studies
Reference
Subjects
Exposure
Effect
Kilburn, 2002a
236 residents near a
microchip plant in
Phoenix, AZ;
Controls:
161 regional referents
from Wickenburg, AZ
67 referents from
Phoenix, AZ not residing
near a plant
0.2-10,000 ppb of
TCE, chronic exposure
Simple and choice reaction times were
increased in the exposed group (p <
0.05).
Kilburn and
Warshaw, 1993
160 residents living in
Southwest Tucson with
TCE and other solvents
in groundwater.
Control: 68 residential
referents matched to
subjects from 2 previous
studies of waste oil and
oil refinery exposures.
>500 ppb of TCE in
well-water before
1981 and 25 to 100
ppb afterwards.
Exposure duration
ranged from 1 to 25
years.
Mean simple reaction time was 67
milliseconds (msec) longer than the
referent groupp < 0.0001).
Choice reaction time (CRT) of the
exposed subjects was between 93-100
msec longer in three different trials
(p < 0.0001) compared to referents.
Reif et al., 2003
143 residents of the
Rocky Mountain Arsenal
community of Denver
Referent group at lowest
concentration (<5 ppb).
High exposure group >15
ppb.
Medium exposure group >
5 ppb and <15 ppb
Low exposure referent
group < 5 ppb
Significant increase in reaction time as
measured by the simple reaction time
test (p < 0.04) in only among subjects
who reported alcohol use (defined as
having at least one drink per month).
Kilburn and
Thornton, 1996
Group A: Registered
voters from Arizona and
Louisiana with no
exposure to TCE:
n = 264, aged 18-83.
Group B volunteers from
California n = 29 (17
males & 12 females)
Group C: exposed to
TCE & other chemicals
for 5 years or more
n = 217
No exposure or
groundwater analyses
reported
Significant increase in simple and
choice reaction time in exposed group
compared to the unexposed populations.
6/22/2009
346
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Gamberale et al.,
1976
15 healthy men aged
20-31 yrs old.
Controls: Within
Subjects (15 self-
controls)
0 mg/m3, 540 mg/m3
(97ppm), 1,080 mg/m3
(194ppm), 70 minutes.
No change in CRT or SRT. Increase in
time required to perform the RT-
Addition Test (task for adding numbers)
(p < 0.05).
Gunetal., 1978
4 female workers from
one plant exposed to
TCE and 4 female
workers from another
plant exposed to TCE +
nonhalogenated
hydrocarbon solvent
Control: (n = 8) 4
unexposed female
workers from each plant
3 ppm-419 ppm,
duration not specified.
TCE-only exposure increased reaction
time in comparison to controls. In TCE
+ solvent group, ambient TCE was
lower and mean reaction time shortened
in session 2, then rose subsequently to
be greater than at the start.
4.2.6.1.2 Muscular Dyscoordination
Three studies examined motor dyscoordination effects from TCE exposure using
subjective and self-reported individual assessment. Rasmussen et al. (1993c) presented findings
on muscular dyscoordination for 96 metal degreasers exposed to either TCE or CFC 113. A
statistically significant increasing trend of dyscoordination with TCE exposure was observed
(p = 0.01) in multivariate regression analyses which adjusted for the effects of age, neurological
disease, arteriosclerotic disease, and alcohol abuse. Furthermore, a greater number of abnormal
coordination tests were observed in the higher exposure group compared to the low exposure
group (p = 0.003)
Gash et al. (2008) reported fine motor hand movement times in subjects who had filed
workman compensation claims were significantly slower (p < 0.0001) than age-matched non-
exposed controls. Exposures were based on self-reported information, and no information on the
control group is presented. Troster and Ruff (1990) reported a case study conducted on two
occupationally exposed workers to TCE. Mild deficits in motor speed were reported for both
cases. In the first case, manual dexterity was impaired in a male exposed to TCE (unknown
concentration) for eight months. In the second case study where a female was exposed to TCE
(low concentration; exact level not specified) for 3 months, there was weakness in the quadriceps
muscle as evaluated in a neurological exam and a decreased sensation to touch on one hand.
Both Gash et al. (2008) and Troster and Ruff (1990) provide very limited information given their
6/22/2009
347
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
deficiencies related to lack of exposure data, self-reported information, and limited reporting of
referents and statistical analysis.
4.2.6.2Psychomotor effects: Laboratory animal data
Several animal studies have demonstrated that TCE exposure produces changes in
psychomotor function. At high doses (>2000 mg/kg) TCE causes mice to lose their righting
reflex when the compound is injected intraperitoneally (Shih et al., 2001; Umezu et al., 1997).
At lower exposures (inhalation and oral), TCE produces alterations in neurobehavioral measures
including locomotor activity, gait, operant responding, and reactivity. The studies are described
in Sections 4.2.6.2.1-4.2.6.2.3 and summarized in Tables 4.2-11 and 4.2-12.
4.2.6.2.1 Loss of righting reflex.
Umezu et al. (1997) studied disruption of the righting reflex following acute injection
(i.p.) of 2,000, 4,000, and 5,000 mg/kg TCE in male ICR mice. TCE disrupted the righting
reflex at doses of 2,000 mg/kg and higher. At 2,000 mg/kg, loss of righting reflex (LORR) was
observed in only 2/10 animals injected. At 4,000 mg/kg, 9/10 animals experienced LORR and
100% of the animals experienced LORR at 5,000 mg/kg.
Shih et al. (2001) reported impaired righting reflexes at exposure doses of 5,000 mg/kg
(i.p.) in male Mfl mice. Mice pretreated with DMSO or disulfuram (CYP2E1 inhibitor) delayed
LORR in a dose related manner. By contrast, the alcohol dehydrogenase inhibitor, 4-
metylpyradine did not delay LORR that resulted from 5,000 mg/kg TCE. These data suggest
that the anesthetic properties of TCE involve its oxidation via CYP2E1 to an active metabolite.
4.2.6.2.2 Activity, sensory-motor and neuromuscular function.
Changes in sensory-motor and neuromuscular activity was reported in three studies
(Kishi et al., 1993; Moser et al., 1995; Moser et al., 2003). Kishi et al. (1993) exposed male
Wistar rats to 250, 500, 1,000, 2,000 and 4,000 ppm TCE for 4 hours. Rats exposed to 250 ppm
TCE showed a significant decrease both in the total number of lever presses and in avoidance
responses at 140 minutes of exposure compared with controls. Moser et al. (1995) evaluated the
effects of acute and short-term (14 day) administration of TCE in adult female Fischer 344 rats
(n = 8-10/dose) on activity level, neuromuscular function and sensorimotor function as part of a
larger functional observational battery (FOB) testing. The NOAEL levels identified by the
authors are 500 mg/kg (10% of the limit dose) for the acute treatment and 150 mg/kg (3% of the
limit dose) for the 14-day study. In the acute study, TCE produced the most significant effects in
6/22/2009
348
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
motor activity (activity domain), gait (neuromuscular domain), and click response (sensorimotor
domain). In the 14-day study, only the activity domain (rearing) and neuromuscular domain
(forelimb grip strength) were significantly different (p < 0.05) from control animals. In a
separate 10-day study (Moser et al., 2003), TCE administration significantly (p < 0.05) reduced
motor activity, tail pinch responsiveness, reactivity to handling, hind limb grip strength and body
weight. Significant increases (p < 0.05) in piloerection, gait scores, lethality, body weight loss,
and lacrimation was also reported in comparison to controls.
There are also two negative studies which used adequate numbers of subjects in their
experimental design but used lower doses than did Moser et al. (2003). Albee et al. (2006)
exposed male and female Fischer 344 rats (n = 10/sex) to TCE by inhalation at exposure doses of
250, 800, and 2,500 ppm, for 6 hr/day, 5 days/week, for 13 weeks. The FOB was performed
monthly although it is not certain how much time elapsed from the end of exposure until the
FOB test was conducted. No treatment related differences in grip strength or landing foot splay
were demonstrated in this study. Kulig et al. (1987) also failed to show significant effects of
TCE inhalation exposure on markers of motor behavior. Wistar rats (n = 8) exposed to 500,
1,000, and 1,500 ppm, for 16 hr/day, 5 days/week, for 18 weeks failed to show changes in
spontaneous activity, grip strength, or coordinated hind limb movement. Measurements were
made every three weeks during the exposure period and occurred between 45 and 180 minutes
following the previous TCE inhalation exposure.
Table 4.2-11 Summary of Animal Psychomotor Function and Reaction Time Studies
Reference
Exposure route
Species/strain/
sex/number
Dose level/
Exposure duration
NOAEL;
LOAEL a
Effects
Savolainen et
al., 1977
Inhalation
Rat, Sprague
Dawley, male,
10
0, 200 ppm; 6
hr/day, 4 days
LOAEL:
200 ppm
Increased frequency of
preening, rearing, and
ambulation. Increased
preening time.
Kishi et al.,
1993
Inhalation
Rats, Wistar,
male, number
not specified
0, 250,500,
1,000, 2,000,
4,000 ppm, 4
hours
LOAEL:
250 ppm
Decreased lever presses and
increased responding when
lever press coupled with a
10 s electric shock
(decreased avoidance
response).
Kulig et al.,
1987
Inhalation
Rat, Wistar,
male, 8/dose
0, 500, 1,000, 1,500
ppm; 16 hrs/day, 5
days/wk, 18 weeks
NOAEL:
1,500 ppm
No change in spontaneous
activity, grip strength or
hindlimb movement.
6/22/2009
349
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Moser et al.,
Oral
Rat, Fischer
0, 150, 500, 1,500,
NOAE1:
Decreased motor activity;
1995
344, female,
5,000 mg/kg, 1
500 mg/kg
Neuro-muscular and
8/dose
dose
10AE1:
sensorimotor impairment
1,500
mg/kg
0, 50, 150, 500,
NOAE1:
Increased rearing activity
1,500 mg/kg/day,
150
and decreased forelimb grip
14 days
mg/kg/day
strength.
10AE1:
500
mg/kg/day
Bushnell, 1997
Inhalation
Rat, long
0, 400, 800, 1,200,
NOAE1:
Decreased sensitivity and
Evans, male, 12
1,600, 2,000, 2,400
800 ppm
increased response time in
ppm, 1 hour/test
the signal detection task.
day, 4 consecutive
10AE1:
test days, 2 weeks
1,200 ppm
Shih et al,
Intraperitoneal
Mouse, MF1,
0, 5,000 mg/kg,
10AE1:
Impairment of righting
2001
male, 6
acute
5,000
reflex.
mg/kg
Umezu et al.,
Intraperitoneal
Mouse, ICR,
0, 2,000, 4,000,
10AE1:
loss of righting reflex,
1997
male, 10/group
5,000 mg/kg—loss
2,000
of righting reflex
mg/kg—
measure
loss of
righting
reflex
Mouse, ICR,
0, 62.5, 125,250,
NOAE1:
Decreased responses (lever
male,
500, 1,000 mg/kg,
500 mg/kg
presses) in an operant
6-10/group
single dose and
10AE1:
response task for food
evaluated 30 min
1,000
reward.
post-administration
mg/kg—
operant
Increased responding when
behavior
lever press coupled with a
20 V electric shock
NOAE1:
(punished responding).
125 mg/kg
10AE1:
250 mg/kg
—punished
responding
Bushnell and
Inhalation
Rat, long
0, 2,000, 2,400
10AE1:
Decreased performance on
Oshiro, 2000
Evans, male, 32
ppm; 70 min/day, 9
2,000 ppm
the signal detection task.
days
Increased response time and
decreased response rate.
6/22/2009
350
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Nunes et al.,
2001
Oral
Rat, Sprague
Dawley, male,
10/group
0, 2,000 mg/kg/day,
7 days
LOAEL:
2,000
mg/kg/day
Increased foot splay. No
change in any other
functional observational
battery (FOB) parameter
(e.g. piloerection, activity,
reactivity to handling)
Moser et al.,
2003
Oral
Rat, Fischer
344, female,
10/group
0, 40, 200, 800,
1,200 mg/kg/day,
10 days
Decreased motor activity;
Decreased sensitivity to tail
pinch; Increased
abnormality in gait;
Decreased grip strength;
Adverse changes in several
FOB parameters.
Albee et al.,
2006
Inhalation
Rat, Fischer
344, male
and female,
10/sex/group
0, 250, 800,
2,500 ppm; 6
hr/day, 5
days/wk, 13
wks.
NOAEL:
2,500
ppm
No change in any FOB
measured parameter.
a NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level)
4.2.6.2.3 Locomotor activity
The data, with regard to locomotor activity, are inconsistent. Several studies showed that
TCE exposure can decrease locomotor activity including Wolff and Siegmund (1978) where AB
mice (n= 18) were treated acutely with a dose of 182 mg/kg, ip at one of 4 time points during a
24 hour day. Moser et al. (1995, 2003) reported reduced locomotor activity in female Fischer
344 rats (n = 8-10) gavaged with TCE over an acute (LOAEL = 5,000mg/kg TCE) or subacute
period (LOAEL = 500 but no effect at 5,000 mg/kg). In the Moser et al. (2003), it appears that
200-mg/kg TCE yielded a significant reduction in locomotor activity and that the degree of
impairment at this dose represented a maximal effect on this measure. That is, higher doses of
TCE appear to have produced equivalent or slightly less of an effect on this behavior. While this
study identifies a LOAEL of 200-mg/kg TCE by gavage over a 10 day period, this is a much
more lower dose effect than that reported in Moser et al. (1995). Both studies (Moser et al.,
1995, 2003) demonstrate a depression in motor activity that occurs acutely following TCE
administration. Kulig et al. (1987) demonstrated that rats had increased response latency to a
two choice visual discrimination following 1,000 and 1,500 ppm TCE exposures for 18 weeks.
However, no significant changes in grip strength, hindlimb movement, or any other motor
activity measurements were noted.
6/22/2009
351
-------
1
2
3
4
5
6
7
8
9
10
11
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
There are also a few studies (Fredriksson et al., 1993; Waseem et al., 2001) generally
conducted using lower exposure doses that failed to demonstrate impairment of motor activity or
ability following TCE exposure. Waseem et al. (2001) failed to demonstrate changes in
locomotor activity in male Wistar rats (n = 8) dosed with TCE (350, 700, and 1,400 ppm) in
drinking water for 90 days. Wistar rats (n = 8) exposed to 500, 1,000, and 1,500 ppm for 16
hr/day, 5 days/week, for 18 weeks failed to show changes in spontaneous activity. No changes
in locomotor activity were observed for 17-day-old male NMRI mice that were dosed postnatally
with 50 or 290 mg/kg-day from day 10 to 16 (Fredriksson et al., 1993). However, rearing
activity was significantly decreased in the NMRI mice at day 60.
Table 4.2-12 Summary of Animal Locomotor Activity Studies.
Reference
Exposure route
Species/strain/
sex/number
Dose level/
Exposure duration
NOAEL;
LOAELa
Effects
Wolff and
Intraperitoneal
Mouse, AB,
0, 182 mg/kg,
LOAEL:
Decreased spontaneous
Siegmund,
male, 18
tested 30 minutes
182 mg/kg
motor activity.
1978
after injection
Kulig et al.,
Inhalation
Rat, Wistar,
0, 500, 1,000, 1,500
NOAEL:
No change in spontaneous
1987
male, 8/dose
ppm; 16 hrs/day, 5
days/wk, 18 weeks
500 ppm
LOAEL:
1,000 ppm
activity, grip strength or
hindlimb movement.
Increased latency time in the
two-choice visual
discrimination task
(cognitive disruption and/or
motor activity related effect)
Moser et al.,
Oral
Rat, Fischer
0, 150, 500, 1,500,
NOAEL:
Decreased motor activity;
1995
344, female,
5,000 mg/kg, 1
500 mg/kg
Neuro-muscular and
8/dose
dose
LOAEL:
1,500
mg/kg
sensorimotor impairment
0, 50, 150, 500,
NOAEL:
Increased rearing activity
1,500 mg/kg/day,
150
14 days
mg/kg/day
LOAEL:
500
mg/kg/day
Waseem et al.,
Oral
Rat, Wistar,
0, 350, 700, 1,400
NOAEL:
No significant effect on
2001
male, 8/group
ppm in drinking
water for 90 days
1,400 ppm
spontaneous locomotor
activity
6/22/2009
352
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Inhalation
Rat, Wistar,
male, 8/group
0, 376 ppmforup
to 180 days; 4
hr/day, 5 days/wk
LOAEL:
376 ppm
Changes in locomotor
activity and vary by
timepoint when measured
over the 180 day period.
Moser et al.,
2003
Oral
Rat, Fischer
344, female,
10/group
0, 40, 200, 800,
1,200 mg/kg/day,
10 days
Decreased motor activity;
Decreased sensitivity;
Increased abnormality in
gait; Adverse changes in
several FOB parameters.
a NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level)
4.2.6.3 Summary and Conclusions for Psychomotor Effects
In human studies, psychomotor effects such as reaction time and muscular
dyscoordination have been examined following TCE exposure. In the reaction time studies,
statistically significant increases in choice reaction time (CRT) and simple reaction time (SRT)
were reported in the Kilburn studies (Kilburn, 2002a; Kilburn and Warshaw, 1993; Kilburn and
Thornton, 1996). All of these studies were geographically based and it was suggested that the
results were used for litigation and the differences between exposed and referent groups on other
factors influencing reaction speed time may introduce a bias to the findings. Additionally, in
these studies exposure to TCE and other chemicals occurred through drinking water for at least
one year and TCE concentrations in well water ranged from 0.2 ppb to 10,000 ppb. Reif et al.
(2003) whose exposure assessment approach included exposure modeling of water distribution
system to estimate TCE concentrations in tap water at census track of residence found that
residents with drinking water containing TCE (up to >15 ppb—the highest level not specified)
and other chemicals did not significantly increase CRTs or SRTs. Inhalation studies also
demonstrated increased reaction times. An acute exposure chamber study (Gamberale et al.,
1976) tested for CRT, SRT, and RT-addition following a 70-minute exposure to TCE. A
concentration-dependent significant decrease in performance was observed with the RT-addition
test and not for CRT or SRT tasks. An occupational exposure study on 8 female workers
exposed to TCE (Gun et al., 1978) also reported increased reaction time in the females exposed
to TCE-only. Muscular dyscoordination for humans following TCE exposure has been reported
in a few studies as a subjective observation. The studies indicated that exposure resulted in
decreased motor speed and dexterity (Troster and Ruff, 1990; Rasmussen et al., 1993c) and self-
reported faster asymptomatic fine motor hand movements (Gash et al., 2008).
Animal studies evaluated psychomotor function by examining locomotor activity, operant
responding, changes in gait, loss of righting reflex, and general motor behavior (see Tables 4-
6/22/2009
353
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
2.11 and 4-2.12 for references). Overall, the studies demonstrated that TCE causes loss of
righting reflex at injection doses of 2,000 mg/kg or higher (Umezu et al., 1997; Shih et al.,
2001). Regarding general psychomotor testing, significant decreases in lever presses and
avoidance were observed at inhalation exposures as low as 250 ppm for 4 hours (LOAEL; Kishi
et al., 1993). Following subchronic inhalation exposures, no significant changes in psychomotor
activity were noted at up to 2,500 ppm for 13 weeks (Albee et al., 2006) or at 1,500 ppm for 18
weeks (Kulig et al., 1987). In the oral administration studies (Moser et al., 1995, 2003),
psychomotor effects were evaluated using an FOB. More psychomotor domains were
significantly affected by TCE treatment in the acute study in comparison to the 14-day study, but
a lower NOAEL (150 mg/kg/day) was reported for the 14-day study in comparison to the acute
study (500 mg/kg; Moser et al., 1995). Upon closer examination of the data, a biphasic effect in
one measure of the FOB (rearing) was resulting in the lower NOAEL for the 14-day study and
doses that were higher and lower than the NOAEL did not produce a statistically significant
increase in the number of rears. Therefore, it can be surmised that acute exposure to TCE results
in significant changes in psychomotor function. However, there may be some tolerance to these
psychomotor changes in increased exposure duration to TCE as evidenced by the results noted in
the short-term and subchronic exposure studies.
4.2.7 Mood Effects and Sleep Disorders
4.2.7.1 Effects on Mood: Human Studies
Reports of mood disturbance (depression, anxiety) resulting from TCE exposure are
numerous in the human literature. These symptoms are subjective and difficult to quantify.
Studies by Gash et al. (2008), Kilburn and Warshaw (1993), Kilburn (2002a, 2002b), McCunney
et al. (1988), Mitchell et al. (1969), Rasmussen and Sabroe (1986), and Troster and Ruff (1990)
reported mood disturbances in humans. Reif et al. (2003) and Triebig (1976, 1977) reported no
effect on mood following TCE exposures.
4.2.7.2 Effects on Mood: Laboratory animal findings
It is difficult to obtain comparable data of emotionality in laboratory studies. However,
Moser et al. (2003) and Albee et al. (2006) both report increases in handling reactivity among
rats exposed to TCE. In the Moser study, female Fischer 344 rats received TCE by oral gavage
for periods of 10 days at doses of 0, 40, 200, 800, and 1,200 while Albee et al. (2006) exposed
6/22/2009
354
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Fischer 344 rats to TCE by inhalation at exposure doses of 250, 800, and 2,500 ppm for 6 hr/day,
5 days/week, for 13 weeks.
4.2.13 Sleep Disturbances
Arito et al. (1994) exposed male Wistar rats to 50-, 100-, and 300-ppm TCE for 8
hour/day, 5 days/week, for 6 weeks and measured electroencephalographic (EEG) responses.
EEG responses were used as a measure to determine the number of awake (wakefulness hours)
and sleep hours. Exposure to all the TCE levels significantly decreased amount of time spent in
wakefulness (W) during the exposure period. Some carry over was observed in the 22 hr post
exposure period with significant decreases in wakefulness seen at 100 ppm TCE. Significant
changes in W-sleep elicited by the long-term exposure appeared at lower exposure levels. These
data seem to identify a low dose effect of TCE and established a LOAEL of 50 ppm for sleep
changes.
4.2.8 Developmental neurotoxicity
4.2.8.1 Human Studies
In humans, CNS birth defects were observed in a few studies (ATSDR, 2001; Bove,
1996; Bove et al., 1995; Lagakos et al., 1986). Postnatally, observed adverse effects in humans
include delayed newborn reflexes following exposure to TCE during childbirth (Beppu, 1968),
impaired learning or memory (Bernad et al., 1987, abstract; White et al., 1997); aggressive
behavior (Bernad et al., 1987, abstract); hearing impairment (Burg and Gist, 1999); speech
impairment (Burg and Gist, 1999; White et al., 1997); encephalopathy (White et al., 1997);
impaired executive and motor function (White et al., 1997); attention deficit (Bernad et al., 1987,
abstract; White et al., 1997), and autism spectrum disorder (Windham et al., 2006). The human
developmental neurotoxicity studies are discussed in more detail in Section 4.7.2.1.2.
Table 4.2-13 Summary of human developmental neurotoxicity associated with
TCE exposures
Finding
Species
Citations
CNS defects, neural tube defects
Human
ATSDR, 2001
Bove, 1996; Bove et al., 1995
Lagakos et al., 1986
Delayed newborn reflexes
Human
Beppu, 1968
Impaired learning or memory
Human
Bernad et al., 1987, abstract
White etal., 1997
6/22/2009
355
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Aggressive behavior
Human
Bernad et al., 1987, abstract
Hearing impairment
Human
Burg and Gist, 1999
Speech impairment
Human
Burg and Gist, 1999
White etal., 1997
Encephalopathy
Human
White etal., 1997
Impaired executive function
Human
White etal., 1997
Impaired motor function
Human
White etal., 1997
Attention deficit
Human
White et al., 1997
Human
Bernad et al., 1987, abstract
Autism spectrum disorder (ASD)
Human
Windham et al., 2006
4.2.8.2Animal Studies
There are a few studies demonstrating developmental neurotoxicity following
trichloroethylene exposure (range of exposures) to experimental animals. These studies
collectively suggest that developmental neurotoxicity result from TCE exposure, however some
types of effects such as learning and memory measures have not been evaluated. Most of the
studies demonstrate either spontaneous motor activity changes (Taylor et al., 1985) or
neurochemical changes such as decreased glucose uptake and changes in the specific gravity of
the cortex and cerebellum (Westergren et al., 1984; Noland-Grebec et al., 1986; Isaacson and
Taylor, 1989). In addition, in most of these studies there is no assessment of the exposure to
TCE or metabolites in the pups/offspring. Details of the studies are presented below and
summarized in Table 4.2-14.
Taylor et al. (1985) administered TCE to female Sprague-Dawley rats in their drinking
water from 14 days before breeding throughout gestation and until pups were weaned at 21 days.
Measured TCE concentrations in the dams ranged from 312-646 mg/L, 625-1,102 mg/L, and
1,250-1,991 mg/L in the low, mid, and high dose groups as measured from the drinking water.
Pups were evaluated for exploratory activity at 28, 60, or 90 days. No significant differences
were noted between control and treated pups at 28 days. At 60 days, all TCE-treated animals
had significantly increased exploratory activity in comparison to age-matched controls, but only
the high group had increased activity at 90 days. A significant increase in spontaneous motor
activity (as measured by a wheel-running task) was noted in only the high dose TCE
(1,250-1,991 mg/L) group during the onset of the darkness period. This study demonstrated that
both spontaneous and open field activities are significantly affected by developmental TCE
exposure.
6/22/2009
356
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Spontaneous behavioral changes were also investigated in another study by Fredriksson
et al. (1993). Male and female NMRI pups (mice) were orally administered 50 or 290 mg/kg/day
for 7 days starting at postnatal day 10. Spontaneous motor activity was investigated in male
mice at ages 17 and 60 days. TCE-treated animals tested at day 17 did not demonstrate changes
in any spontaneous activity measurements in comparison to control animals. Both doses of TCE
(50 and 290 mg/kg/day) significantly decreased rearing in 60 day-old male mice.
Westergren et al. (1984) examined the brain specific gravity of litters from mice exposed
to TCE. NMRI mice (male and female) were exposed to 150 ppm TCE (806.1 mg/m3) for 30
days prior to mating. Exposure in males continued until the end of mating and females were
exposed until the litters were born. Brains were removed from the offspring at either postnatal
day 1,10, 20-22, or 29-31. At postnatal days 1 and 10, significant decreases were noted in the
specific gravity of the cortex. Significant decreases in the specific gravity of the cerebellum
were observed at postnatal day 10 (decrease from 1.0429 ± 0.00046 to 1.0405 ± 0.00030) and
20-22 (decrease from 1.0496 ± 0.00014 to 1.0487 ± 0.00060). Cerebellum measurements were
not reported for postnatal day 29-31 animals. Neurobehavioral assessments were not conducted
in this study. Additionally, decreased brain specific gravity is suggestive of either decreased
brain weight or increased brain volume (probably from edema) or a combination of the two
factors and is highly suggestive of an adverse neurological effect. The effects of TCE on the
cortical specific gravity were not persistent since cortices from postnatal day 29-31 animals did
not exhibit any significant changes. It is unclear if the effects on the cerebellum were persistent
since results were not reported for the postnatal day 29-31 animals. However, the magnitude of
the change in the specific gravity of the cerebellum is decreased from postnatal day 10 to
postnatal day 20-22 suggesting that the effect may be reversible given a longer recovery period
from TCE.
The effect of TCE on glucose uptake in the brain was evaluated in rat pups exposed to
TCE during gestation and through weaning. The primary source of energy utilized in the CNS is
glucose. Changes in glucose uptake in the brain are a good indicator for neuronal activity
modification. Noland-Grebec et al. (1986) administered 312 mg/L TCE through drinking water
to female Sprague-Dawley rats from 2 weeks before breeding and up until pups reached 21 days
of age. To measure glucose uptake, 2-deoxyglucose was administered intraperitoneally to male
pups at either postnatal day 7, 11, 16, or 21. Significant decreases in glucose uptake were noted
in whole brain and cerebellum at all postnatal days tested. Significant decreases in glucose
uptake were also observed in the hippocampus except for animals tested at postnatal day 21. The
observed decrease in glucose uptake suggests decreased neuronal activity.
6/22/2009
357
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Female Sprague-Dawley rats (70 days old) were administered TCE in drinking water at a
level of either 4.0 or 8.1 mg/day for 14 days prior to mating and continuing up through lactation
(Isaacson and Taylor, 1989). Only the male pups were evaluated in the studies. At postnatal day
21, brains were removed from the pups, sectioned, and stained to evaluate the changes in myelin.
There was a significant decrease (40% decrease) in myelinated fibers in the CA1 region of the
hippocampus of the male pups. This effect appeared to be limited to the CA1 region of the
hippocampus since other areas such as the optic tract, fornix, and cerebral peduncles did not have
decreases in myelinated fibers.
Neurological changes were found in pups exposed to TCE in a study conducted by NTP
in Fischer 344 rats (George et al., 1986). TCE was administered to rats at dietary levels of 0,
0.15, 0.30, or 0.60%. No intake calculations were presented for the rat study and therefore a
dose rate is unavailable for this study. Open field testing revealed a significant (p < 0.05) dose-
related trend toward an increase in the time required for male and female F1 weanling pups
(PND 21) to cross the first grid in the testing device, suggesting an effect on the ability to react to
a novel environment.
Blossom et al. (2008) treated male and female MRL +/+ mice with 0 or 0.1 mg/mL TCE
in the drinking water. Treatment was initiated at the time of mating, and continued in the
females (8/group) throughout gestation and lactation. Behavioral testing consisted of righting
reflex on PNDs 6, 8, and 10; bar-holding ability on PNDs 15 and 17; and negative geotaxis on
PNDs 15 and 17. Nest building was assessed and scored on PND 35, the ability of the mice to
detect and distinguish social odors was examined with an olfactory habituation/dishabituation
method at PND 29, and a resident intruder test was performed at PND 40 to evaluate social
behaviors. Righting reflex, bar holding, and negative geotaxis were not impaired by treatment.
There was a significant association between impaired nest quality and TCE exposure in tests of
nest-building behavior; however, TCE exposure did not have an effect on the ability of the mice
to detect social and non-social odors using habituation and dishabituation methods. Resident
intruder testing identified significantly more aggressive activities (i.e., wrestling and biting) in
TCE-exposed juvenile male mice as compared to controls, and the cerebellar tissue from the
male TCE-treated mice had significantly lower GSH levels and GSH:GSSG ratios, indicating
increased oxidative stress and impaired thiol status, which have been previously reported to be
associated with aggressive behaviors (Franco et al., 2006). Histopathological examination of the
brain did not identify alterations indicative of neuronal damage or inflammation.
6/22/2009
358
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.2-14. Summary of mammalian in vivo developmental neurotoxicity studies— oral
exposures
Reference
Species/strain/
sex/number
Dose level/
Exposure duration
NOAEL; LOAELa
Effects
Fredriksson
etal., 1993
Mouse, NMRI, male
pups, 12 pups from 3-4
different litters/group
0, 50, or 290 mg/kg-
day
PND 10-16
LOAEL: 50 mg/kg-
day
Rearing activity sig. 1 at both
dose levels on PND 60
George et
al., 1986
Rat, F334, male and
female, 20 pairs/
treatment group,
40 controls/sex
0,0.15,0.30, or 0.60%
microencapsulated
TCE in diet.
Breeders exposed 1
wk pre-mating, then
for 13 wk; pregnant
: s throughout
pregnancy (i.e., 18 wk
total)
LOAEL: 0.15%
Open field testing in pups: a
sig. dose-related trend toward
T time required for male and
female pups to cross the first
grid in the test device
Isaacson &
Taylor,
1989
Rat, Sprague-Dawley,
females, 6 dams/group
0, 312, or 625 mg/L.
(0, 4.0, or 8.1 mg/day)
b
Dams (and pups)
exposed from 14 days
prior to mating until
end of lactation.
LOAEL: 312 mg/L
Sig. X myelinated fibers
in the stratum
lacunosum-moleculare of
pups. Reduction in
myelin in the CA1 region
of the hippocampus.
Noland-
Gerbec et
al., 1986
Rat, Sprague-Dawley,
females, 9-11 dams/
group
0, 312 mg/L
(Avg. total intake of
dams: 825 mg TCE
over 61 days.)
Dams (and pups)
exposed from 14 days
prior to mating until
end of lactation.
LOAEL: 312 mg/L
Sig. i uptake of 3H-2-DG in
whole brains and cerebella
(no effect in hippocampus)
of exposed pups at 7, 11, and
16 days, but returned to
control levels by 21 days.
6/22/2009
359
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Taylor et
al., 1985
Rat, Sprague-
Dawley, females,
no. dams/group not
reported
0, 312, 625, and
1,250 mg/L in
drinking water
Dams (and pups)
exposed from 14 days
prior to mating until
end of lactation.
LOAEL: 312
mg/L
Exploratory behavior sig. t in
60- and 90-day old male rats
at all treatment levels.
Locomotor activity (measured
through the wheel-running
tasks) was higher in rats from
dams exposed to 1,250 mg/L
TCE.
Blossom et
al., 2008
Mouse, MRL +/+,
dams and both sexes
offspring, 8
litters/group; 3-8
pups/group
Drinking water, From
GD 0 to PND 42; 0 or
0 1 mg/mL; maternal
dose = 25.7 mg/kg-
day; offspring PND
24-42 dose = 31.0
mg/kg-day
LOAEL: 31 mg/kg-
day for offspring
Righting reflex, bar holding,
and negative geotaxis were not
impaired. Significant
association between impaired
nest quality and TCE
exposure.
Lower GSH levels and
GSH:GSSG ratios with TCE
exposure.
aNOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level), and LOEL
(Lowest Observed Effect Level) are based upon reported study findings.
bDose conversions provided by study author(s).
4.2.8.3 Summary and Conclusions for the Developmental Neurotoxicity Studies
Gestational exposure to TCE in humans has resulted in several developmental
abnormalities. These changes include neuroanatomical changes such as neural tube defects
(ATSDR, 2001; Bove et al., 1995, 1996; Lagakos et al., 1986) and encephalopathy (White et al.,
1997). Clinical neurological changes such as impaired cognition (Bernad et al., 1987; White et
al., 1997), aggressive behavior (Bernad et al., 1987), and speech and hearing impairment (Burg
and Gist, 1999; White et al., 1997) are also observed when TCE exposure occurs in utero.
In animal studies, anatomical and clinical developmental neurotoxicity is also observed.
Following inhalation exposures of 150 ppm to mice during mating and gestation, the specific
gravity of offspring brains was significantly decreased at postnatal time points through the age of
weaning; this effect did not persist to 1 month of age (Westergren et al., 1984). In studies
reported by Taylor et al. (1985), Isaacson and Taylor (1989), and Noland-Gerbec et al. (1986),
312 mg/L exposures in drinking water that were initiated 2 weeks prior to mating and continued
to the end of lactation resulted in (a) significant increase in exploratory behavior at postnatal
days 60 and 90, (b) reductions in myelination in the CA1 hippocampal region of offspring at
6/22/2009
360
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
weaning, and (c) significantly decreased uptake of 2-deoxyglucose in the rat brain at postnatal
day 21. Gestational exposures to mice (Fredriksson et al., 1993) resulted in significantly
decreased rearing activity on postnatal day 60, and dietary exposures during the course of a
continuous breeding study in rats (George et al., 1986) found a significant trend toward increased
time to cross the first grid in open field testing. In a study by Blossom et al. (2008), male mice
exposed gestationally to TCE exhibited lower GSH levels and lower GSH:GSSG ratios which is
also observed in mice that have more aggressive behaviors (Franco et al., 2006).
4.2.9 Mechanistic studies of TCE neurotoxicity
4.2.9.1 Dopamine neuron disruption
There are very recent laboratory animal findings resulting from short-term TCE
exposures that demonstrate vulnerability of dopamine neurons in the brain to this chlorinated
hydrocarbon. The key limitation of these laboratory animal studies is that only 1 dosing regimen
was included in each study. Moreover, there has been no systematic body of data to show that
other chlorinated hydrocarbons such as tetrachloroethylene or aromatic solvents similarly target
this cell type. Confidence in the limited data regarding dopamine neuron death and in vivo TCE
exposure would be greatly enhanced by identifying a dose-response relationship. If indeed TCE
can target dopamine neurons it would be anticipated that human exposure to this agent would
result in elevated rates of parkinsonism. There are no systematic studies of this potential
relationship in humans although one limited report attempted to address this possibility.
Difficulties in subject recruitment into that study limit the weight that can be given to the results.
Endogenously formed chlorinated tetrahydro-beta-carbolines (TaClo) have been
suggested to contribute to the development of Parkinson-like symptoms (Bringmann et al., 1992,
1995; Reiderer et al., 2002; Kochen et al., 2003). TaClo can be formed endogenously from
metabolites of TCE such as trichloroacetaldehyde. TaClo has been characterized as a potent
neurotoxicant to the dopaminergic system. Some research groups have hypothesized that
Parkinson-like symptoms resulting from TCE exposure may occur through the formation of
TaClo, but not enough evidence is available to determine if this mechanism occurs.
6/22/2009
361
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.2.9.1.1 Dopamine Neuron Disruption: Human Studies
There are no human studies that present evidence of this effect. Nagaya et al. (1990)
examined serum dopamine P-hydroxylase activity without differences observed in mean
activities between control and exposed subjects. In the study, 84 male workers exposed to TCE
were compared to 83 male age-matched controls. The workers had constantly used TCE in their
jobs and their length of employment ranged from 0.1 to 34 years.
4.2.9.1.2 Dopamine Neuron Disruption: Animal Studies
There are limited data from mice and rats that suggest the potential for TCE to disrupt
dopamine neurons in the basal ganglia (Table 4.2-15). Gash et al. (2008) showed that TCE
gavage in Fischer 344 rats (n = 9) at an exposure level of 1,000 mg/kg/day, 5 days/week, for 6
weeks yielded degeneration of dopamine neurons in the substantia nigra and alterations in
dopamine turnover as reflected in a shift in dopamine metabolite to parent compound ratios.
Guehl et al. (1999) reported similar findings in OF1 mice (n = 10) that were injected ip with 400
mg/kg/day TCE 5 days/week for 4 weeks. Each of these studies evaluated only a single dose
level of TCE so that establishing a dose response relationship is not possible. Consequently,
these data are of limited utility in risk assessment because they do not establish the potency of
TCE to damage DA neurons. They are important, however, in identifying a potential permanent
impairment that might occur following TCE exposure at relatively high exposure doses. They
also identify a potential mechanism by which TCE could produce CNS injury.
Table 4.2-15 Summary of Animal Dopamine Neuronal Studies
Reference
Exposure route
Species/strain/
sex/number
Dose level/
Exposure
duration
NOAEL;
LOAELa
Effects
Guehl et
al., 1999
Intraperitoneal
Administration
Mouse, OF1, male,
10
0, 400 mg/kg; 5
days/wk, 4
weeks
LOAEL:
400 mg/kg
Significant dopaminergic
neuronal death in substantia
nigra.
Gash et al.,
2008
Oral gavage
Rat, Fischer 344,
male, 9/group
0, 1,000 mg/kg;
5 days/wk, 6
weeks
LOAEL:
1,000 mg/kg
Degeneration of dopamine-
containing neurons in substantia
nigra.
Change in dopamine
metabolism
a NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level)
6/22/2009
362
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.2.9.1.3 Summary and Conclusions of Dopamine Neuron Studies
Only two animal studies have reported changes in dopamine neuron effects from TCE
exposure (Gash et al., 2008; Guehl et al., 1999). Both studies demonstrated toxicity to
dopaminergic neurons in the substantia nigra in rats or mice. LOAELs of 400 mg/kg (mice;
Guehl et al., 1999) and 1,000 mg/kg (rats; Gash et al., 2008) were reported for this effect.
Dopaminergic neuronal degeneration following TCE exposure has not been studied in humans.
However, there were no changes in serum dopamine P-hydroxylase activity in TCE-exposed and
control individuals (Nagaya et al., 1990). Loss of dopaminergic neurons in the substantia nigra
also occurs in patients with Parkinson's disease and the substantia nigra is an important region in
helping to control movements. As a result, loss of dopaminergic neurons in the substantia nigra
may be one of the potential mechanisms involved in the clinical psychomotor effects that are
observed following TCE exposure.
4.2.9.2Neurochemical and Molecular Changes.
There is limited data obtained only from laboratory animals that TCE exposure may have
consequences on GABAergic and glutamatergic neurons (Briving et al., 1986; Shih et al., 2001,
see Table 4.2-16). However, the data obtained are limited with respect to brain region examined,
persistence of effect, and whether there might be functional consequences to these changes. The
data of Briving et al. (1986) demonstrating changes in cerebellar high affinity uptake for GAB A
and glutamate following chronic low level (50 and 150 ppm) TCE exposure do not appear to be
reflected in the only other brain region evaluated (hippocampus). However, glutamate levels
were increased in the hippocampus. The data of Shih et al. (2001) is indirect in that it shows an
altered response to GABAergic antagonist drugs in mice treated by acute injection with 250, 500,
1,000, and 2,000 mg/kg TCE. However, this data does show some dose dependency with
significant findings observed with TCE exposure as low as 250 mg/kg.
The development and physiology of the hippocampus has also been evaluated in two
different studies (Isaacson and Taylor, 1989; Ohta et al., 2001). Isaacson and Taylor (1989)
found a 40 percent decrease in myelinated fibers from hippocampi dissected from neonatal
Sprague-Dawley rats (n = 2-3) that were exposed to TCE (4 and 8.1 mg/day) in utero and during
the preweaning period. Ohta et al. (2001) injected male ddY mice with 300 mg/kg TCE and
found a significant reduction in response to titanic stimuli in excised hippocampal slices. Both
of these studies demonstrated that there is some interaction with TCE and the hippocampal area
in the brain.
6/22/2009
363
-------
1
2
3
4
5
6
7
8
9
10
11
12
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Impairment of sciatic nerve regeneration was demonstrated in mice and rats exposed to
TCE (Kjellstrand et al., 1987). Under heavy anesthesia, the sciatic nerve of the animals was
artificially crushed to create a lesion. Prior to the lesion, some animals were pre-exposed to TCE
for 20 days and then for an additional four days after the lesion. Another set of animals were
only exposed to TCE for four days following the sciatic nerve lesion. For mice, regeneration of
the sciatic nerve in comparison to air-exposed animals was 20 and 33% shorter in groups
exposed to 150-ppm and 300-ppm TCE for 4 days, respectively. This effect did not significantly
increase in mice pre-exposed to TCE for 20 days and the regeneration was 30% shorter in the
150-ppm group and 22% shorter in the 300-ppm group. Comparatively, a 10% reduction in
sciatic nerve regeneration length was observed in rats exposed to TCE for 20 days prior to the
lesion plus the four days after the sciatic nerve lesion.
Table 4.2-16 Summary of neurophysiological, neurochemical, and neuropathological
effects with TCE exposure
Reference
Exposure
Species/strain/
Dose level/
NOAEL;
Effects
route
sex/number
Exposure duration
LOAEL a
Neurophysiological Studies
Shih et al.,
Intraperitoneal
Mouse, Mfl,
0, 250 500, 1,000,
...
Increased threshold
2001
male, 6/group
2,000 mg/kg, 15
for seizure
minutes; followed by
appearance with
tail infusion of PTZ
TCE pretreatment
(5 mg/mL), picrotoxin
for all convulsants.
(0.8 mg/mL),
Effects strongest on
bicuculline (0.06
the GABAa
mg/mL), strychnine
antagonists, PTZ,
(0.05 mg/mL), 4-AP
picrotoxin, and
(2 mg/mL), or
bicuculline
NMDA (8 mg/mL)
suggesting GABAa
receptor
involvement.
NMDA and glycine
Rc involvement also
suggested.
Ohta et al.,
Intraperitoneal
Mouse, ddY,
0, 300, 1,000 mg/kg,
LOAEL: 300
Decreased response
2001
male, 5/group
sacrificed 24 hours
mg/kg
(LTP response) to
after injection
tetanic stimulation in
the hippocampus.
Neurochemical Studies
6/22/2009
364
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Briving et al.,
1986
Inhalation
Gerbils,
Mongolian,
male and
female, 6/group
0, 50, 150 ppm,
continuous, 24 hr/day,
12 months
NOAEL: 50 ppm;
LOAEL: 150 ppm
for glutamate
levels in
hippocampus
Increased glutamate
levels in the
hippocampus.
Increased glutamate
and GABA uptake
in the cerebellar
vermis.
NOAEL: 150
ppm for
glutamate and
GABA uptake in
hippocampus
LOAEL: 50 ppm
for glutamate and
GABA uptake in
cerebellar vermis
Subramoniam
etal., 1989
Oral
Rat, Wistar,
female,
0, 1,000 mg/kg, 2 or
20 hours
0, 1,000 mg/kg/day, 5
days/week, 1 year
PI and PIP2
decreased by 24 and
17% at 2 hr;
PI and PIP2
increased by 22 and
38% at 20 hrs.
PI, PIP, and PIP2
reduced by 52,23,
and 45% in 1 year
study.
Haglid et al.,
1981
Inhalation
Gerbil,
Mongolian,
male and
female,
6-7/group
0, 60, 320 ppm, 24
hr/day, 7 days/week,
3 months
LOAEL: 60 ppm,
brain protein
changes
NOAEL: 60 ppm;
LOAEL: 320
ppm, brain DNA
changes
1) Decreases in total
brain soluble protein
whereas increase in
S100 protein.
2) Elevated DNA in
cerebellar vermis
and sensory motor
cortex
Neuropathologies Studies
Kjellstrand et
al., 1987
Inhalation
Mouse, NMRI,
male
0, 150, 300 ppm, 24
hr/day, 4 or 24 days
LOAEL: 150
ppm, 4 and 24
days
Sciatic nerve
regeneration was
inhibited in both
mice and rats.
Rat, Sprague
Dawley, female
0, 300 ppm, 24
hr/day, 4 or 24 days
NOAEL: 300
ppm, 4 days
LOAEL: 300
ppm, 24 days
6/22/2009
365
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Isaacson &
Oral
Rat, Sprague-
0, 312, or 625 mg/L.
LOAEL: 312
Sig. 1 myelinated
Taylor, 1989
Dawley,
(0,4.0, or 8.1
mg/L
fibers in the stratum
females, 6
mg/day)
lacunosum-
dams/group
Dams (and pups)
exposed from 14 days
prior to mating until
end of lactation.
moleculare of pups.
Reduction in myelin
in the hippocampus.
a NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level)
There are also a few in vitro studies (summarized in Table 4.2.-17) that have
demonstrated that TCE exposure alters the function of inhibitory ion channels such as GABAa
and glycine receptors (Krasowski and Harrison, 2000; Beckstead et al., 2000), and serotonin
receptors (Lopreato et al., 2003). Krasowski and Harrison (2000) and Beckstead et al. (2000)
were able to demonstrate that human GABAa and glycine receptors could be potentiated by TCE
when a receptor agonist was co-applied. Krasowski and Harrison (2000) conducted an additional
experiment in order to determine if TCE was interacting with the receptor or perturbating the
cellular membrane (bilipid layer). Specific amino acids on the GABAa and glycine receptors
were mutated and in the presence of a receptor agonist (GABA for GABAa and glycine for
glycine receptors) and in these mutated receptors TCE-mediated potentiation was significantly
decreased or abolished suggesting that there was an interaction between TCE and these
receptors. Lopreato et al. (2003) conducted a similar study with the 5HT3A serotonin receptor
and found that when TCE was co-applied with serotonin, there was a potentiation in receptor
response. Additionally, TCE has been demonstrated to alter the function of voltage sensitive
calcium channels (VSCCs) by inhibiting the calcium mediated-current at a holding potential of -
70 mV and shifting the activation of the channels to a more hyperpolarizing potential (Shafer et
al., 2005).
Table 4.2-17 Summary of in vitro ion channel effects with TCE exposure
Reference
Cellular
System
Neuronal Channel/
Receptor
Concentrations
Effects
In Vitro Studies
Shafer et al.,
2005
PC12 cells
Voltage Sensitive
Calcium Channels
(VSCC)
0, 500, 1,000,
1,500, 2,000 \M
Shift of VSCC activation to a more
hyperpolarizing potential.
Inhibition of VSCCs at a holding
potential of -70 mV.
6/22/2009
366
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Beckstead et
al., 2000
Xenopus
oocytes
Human
recombinant
Glycine receptor
al,
GAB Aa receptors,
aipi, aip2y2L
0, 390 (iM
50% potentiation of the GABAa
receptors; 100% potentiation of the
glycine receptor
Lopreato et al.,
2003
Xenopus
oocytes
Human recombinant
serotonin 3 A
receptor
0, 390 (iM
Potentiation of serotonin receptor
function.
Krasowski and
Harrison, 2000
Human
embryonic
kidney 293
cells
Human recombinant
Glycine receptor al,
GAB Aa receptors
a2pi
Not provided
Potentiation of glycine receptor function
with an EC50 of 0.65 ± 0.05 mM.
Potentiation of GABAa receptor
function with an EC50 of 0.85 ± 0.2 mM
4.2.10 Potential Mechanisms for TCE-mediated Neurotoxicity
The mechanisms of TCE neurotoxicity have not been established despite a significant
level of research on the outcomes of TCE exposure. Results from several mechanistic studies
can be used to help elucidate the mechanism(s) involved in TCE-mediated neurological effects.
The disruption of the trigeminal nerve appears to be a highly idiosyncratic outcome of
TCE exposure. There are limited data to suggest that it might entail a demyelination
phenomenon, but similar demyelination does not appear to occur in other nerve tracts. In this
regard, then, TCE is unlike a variety of hydrocarbons that have more global demyelinating
action. There are some data from central nervous system that focus on shifts in lipid profiles as
well as data showing loss of myelinated fibers in the hippocampus. However, the changes in
lipid profiles are both quite small and, also, inconsistent. And the limited data from
hippocampus are not sufficient to conclude that TCE has significant demyelinating effects in this
key brain region. Indeed, the bulk of the evidence from studies of learning and memory function
(which would be tied to hippocampal function) suggests no clear impairments due to TCE.
Some researchers (Albee et al., 1997, 2006; Barret et al., 1991, 1992; ; Laureno, 1988,
1993) have indicated that changes in trigeminal nerve function may be due to dichloroacetylene
which is formed under nonbiological conditions of high alkalinity or temperature during
volatilization of TCE. In experimental settings, trigeminal nerve function (Albee et al., 1997)
and trigeminal nerve morphology (Barret et al., 1991, 1992) was found to be more altered
6/22/2009
367
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
following a low exposure to dichloroacetylene in comparison to the higher TCE exposure.
Barret et al. (1991, 1992) also demonstrated that TCE administration results in morphological
changes in the trigeminal nerve. Thus, dichloroacetylene may contribute to trigeminal nerve
impairment may be plausible following an inhalation exposure under conditions favoring its
formation. Examples of such conditions include passing through a carbon dioxide scrubber
containing alkaline materials, application to remove a wax coating from a concrete-lined stone
floor, or mixture with alkaline solutions or caustic (Saunders 1967; Greim et al. 1984; Bingham
et al. 2001). However, dichloroacetylene exposures have not been identified or measured in
human epidemiologic studies with TCE exposure, and thus do not appear to be common to
occupational or residential settings (Lash and Green, 1993). Moreover, changes in trigeminal
nerve function have also been consistently reported in humans exposed to TCE following an oral
exposure (Kilburn, 2002a; across many human studies of occupational and drinking water
exposures under conditions with highly varying potentials for dichloroacetylene formation
(Barret et al, 1982, 1984, 1987; Feldman et al., 1988). As a result, the mechanism(s) for
trigeminal nerve function impairment following TCE exposure is unknown., 1992; Kilburn and
Warshaw, 1993; Kilburn, 2002a; Mihri et al., 2004; Ruitjen et al., 1991). The varying
dichloroacetylene exposure potential across these studies suggests TCE exposure, which is
common to all of them, as the most likely etiologic agent for the observed effects.
The clearest consequences of TCE are permanent impairment of hearing in animal
models and disruption of trigeminal nerve function in humans with animal models showing
comparable changes following administration of a TCE metabolite. With regard to hearing loss,
the effect of TCE has much in common with the effects of several aromatic hydrocarbons
including ethylbenzene, toluene, and ^-xylene. Many studies have attempted to determine how
these solvents damage the cochlea. Of the hypotheses that have been advanced, there is little
evidence to suggest oxidative stress, changes in membrane fluidity, or impairment of central
efferent nerves whose endings innervate receptor cells in the cochlea. Rather, for reasons that
are still uncertain these solvents seem to preferentially target supporting cells in the cochlea
whose death then alters key structural elements of the cochlea resulting ultimately in hair cell
displacement and death. Recently, potential modes of action resulting in ototoxicity have been
speculated to be due to blockade of neuronal nicotinic receptors present on the auditory cells
(Campo et al., 2007) and potentially changes in calcium transmission (Campo et al., 2008) from
toluene exposure. Although these findings were reported following an acute toluene exposure, it
is speculated that this mechanism may be a viable mechanism for TCE -mediated ototoxicity.
A few studies have tried to relate TCE exposure with selective impairments of dopamine
neurons. Two studies (Gash et al., 2008; Guehl et al., 1999) demonstrated dopaminergic
6/22/2009
368
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
neuronal death and/or degeneration following an acute TCE administration. However, the only
human TCE exposure study examining dopamine neuronal activity found no changes in serum
dopamine P-hydroxylase activity in comparison to non-exposed individuals (Nagaya et al.,
1990). It is thought that tetrahydro-beta-carbolines (TaClo), which can be formed from TCE
metabolites such as trichloroacetaldehyde, may be the potent neurotoxicant that selectively
targets the dopaminergic system. More studies are needed to confirm the dopamine neuronal
function disruption and if this disruption is mediated through TaClo.
There is good evidence that TCE and certain metabolites such as choral hydrate have
CNS depressant properties and may account for some of the behavioral effects (such as
vestibular effects, psychomotor activity changes, central visual changes, sleep and mood
changes) that have been observed with TCE. Specifically, in vitro studies have demonstrated
that TCE exposure results in changes in neuronal receptor function for the GABAa, glycine, and
serotonin receptors (Krasowski and Harrison, 2000; Beckstead et al., 2000; Lopreato et al.,
2003). All of these inhibitory receptors that are present in the CNS are potentiated when
receptor-specific agonist and TCE are applied. These results are similar to other anesthetics and
suggest that some of the behavioral functions are mediated by modifications in ion channel
function. However, it is quite uncertain whether there are persistent consequences to such high
dose TCE exposure. Additionally, with respect to the GABAergic system, acute administration
of TCE increased the seizure threshold appearance and this effect was the strongest with
convulsants that were GAB A receptor antagonists (Shih et al., 2001). Therefore, this result
suggests that TCE interacts with the GABA receptor and that was also verified in vitro
(Krasowski and Harrison, 2000; Beckstead et al., 2000).
Also, TCE exposure has been linked to decreased sensitivity to titanic stimulation in the
hippocampus (Ohta et al., 2001) as well as significant reduction in myelin in the hippocampus in
a developmental exposure (Isaacson and Taylor, 1990). These effects are notable since the
hippocampus is highly involved in memory and learning functions. Changes in the hippocampal
physiology may correlate with the cognitive changes that were reported following TCE
exposure.
4.2.11 Overall Summary and Conclusions—Weight of Evidence
Both human and animal studies have associated TCE exposure with effects on several
neurological domains. The strongest neurological evidence of hazard in humans is for changes
6/22/2009
369
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
in trigeminal nerve function or morphology and impairment of vestibular function. Fewer and
more limited evidence exists in humans on delayed motor function, and changes in auditory,
visual, and cognitive function or performance. Acute and subchronic animal studies show
morphological changes in the trigeminal nerve, disruption of the peripheral auditory system
leading to permanent function impairments and histopathology, changes in visual evoked
responses to patterns or flash stimulus, and neurochemical and molecular changes. Additional
acute studies reported structural or functional changes in hippocampus, such as decreased
myelination or decreased excitability of hippocampal CA1 neurons, although the relationship of
these effects to overall cognitive function is not established. Some evidence exists for motor-
related changes in rats/mice exposed acutely/subchronically to TCE, but these effects have not
been reported consistently across all studies.
Epidemiologic evidence supports a relationship between TCE exposure and trigeminal
nerve function changes, with multiple studies in different populations reporting abnormalities in
trigeminal nerve function in association with TCE exposure (Barret et al., 1982, 1984, 1987;
Feldman et al., 1988, 1992; Kilburn and Warshaw, 1993; Ruitjen et al., 2001; Kilburn, 2002a;
Mhiri et al., 2004). Of these, two well conducted occupational cohort studies, each including
more than 100 TCE-exposed workers without apparent confounding from multiple solvent
exposures, additionally reported statistically significant dose-response trends based on ambient
TCE concentrations, duration of exposure, and/or urinary concentrations of the TCE metabolite
TCA (Barret et al., 1984; Barret et al., 1987). Limited additional support is provided by a
positive relationship between prevalence of abnormal trigeminal nerve or sensory function and
cumulative exposure to TCE (most subjects) or CFC-113 (<25% of subjects) (Rasmussen et al.,
1993c). Test for linear trend in this study was not statistically significant and may reflect
exposure misclassification since some subjects included in this study did not have TCE exposure.
The lack of association between TCE exposure and overall nerve function in three small studies
(trigeminal: El-Ghawabi et al., 1973; ulnar and medial: Triebig et al., 1982, 1983) does not
provide substantial evidence against a causal relationship between TCE exposure and trigeminal
nerve impairment because of limitations in statistical power, the possibility of exposure
misclassification, and differences in measurement methods. Laboratory animal studies have also
shown TCE-induced changes in the trigeminal nerve. Although one study reported no significant
changes in trigeminal somatosensory evoked potential in rats exposed to TCE for 13 weeks
(Albee et al., 2006), there is evidence of morphological changes in the trigeminal nerve
following short-term exposures in rats (Barret et al., 1991, 1992).
Human chamber, occupational, geographic based/drinking water, and laboratory animal
studies clearly established TCE exposure causes transient impairment of vestibular function.
Subjective symptoms such as headaches, dizziness, and nausea resulting from occupational
6/22/2009
370
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(Granjean et al., 1955; Liu et al., 1988; Rasmussen and Sabroe, 1986; Smith et al., 1970),
environmental (Hirsch et al., 1996), or chamber exposures (Stewart et al., 1970; Smith et al.,
1970) have been reported extensively. A few laboratory animal studies have investigated
vestibular function, either by promoting nystagmus or by evaluating balance (Niklasson et al.,
1993; Tham et al., 1979; Tham et al., 1984; Umezu et al., 1997).
In addition, mood disturbances have been reported in a number of studies, although these
effects also tend to be subjective and difficult to quantify (Gash et al., 2007; Kilburn and
Warshaw, 1993; Kilburn, 2002a, 2002b; McCunney et al., 1988; Mitchell et al., 1969;
Rasmussen and Sabroe, 1986; Troster and Ruff, 1990), and a few studies have reported no
effects from TCE on mood (Reif et al., 2003; Triebig et al., 1976, 1977a). Few comparable
mood studies are available in laboratory animals, although both Moser et al. (2003) and Albee et
al. (2006) report increases in handling reactivity among rats exposed to TCE. Finally,
significantly increased number of sleep hours was reported by Arito et al. (1994) in rats exposed
via inhalation to 50-300 ppm TCE for 8 hr/d for 6 weeks.
Four epidemiologic studies of chronic exposure to TCE observed disruption of auditory
function. One large occupational cohort study showed a statistically significant difference in
auditory function with cumulative exposure to TCE or CFC-113 as compared to control groups
after adjustment for possible confounders, as well as a positive relationship between auditory
function and increasing cumulative exposure (Rasmussen et al., 1993b). Of the three studies
based on populations from ATSDR's TCE Subregistry from the National Exposure Registry,
more limited than Rasmussen et al. (1993b) due to inferior exposure assessment, Burg et al.
(1995) and Burg and Gist (1999) reported a higher prevalence of self-reported hearing
impairments. The third study reported that auditory screening revealed abnormal middle ear
function in children less than 10 years of age, although a dose-response relationship could not be
established and other tests did not reveal differences in auditory function (ATSDR, 2003a).
Further evidence for these effects is provided by numerous laboratory animal studies
demonstrating that high dose subacute and subchronic TCE exposure in rats disrupts the auditory
system leading to permanent functional impairments and histopathology.
Studies in humans exposed under a variety of conditions, both acutely and chronically,
report impaired visual functions such as color discrimination, visuospatial learning tasks, and
visual depth perception in subjects with TCE exposure. Abnormalities in visual depth perception
were observed with a high acute exposure to TCE under controlled conditions (Vernon and
Ferguson, 1969). Studies of lower TCE exposure concentrations also observed visuofunction
effects. One occupational study (Rasmussen et al., 1993b) reported a statistically significant
positive relationship between cumulative exposure to TCE or CFC-113 and visual gestalts
learning and retention among Danish degreasers. Two studies of populations living in a
6/22/2009
371
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
community with drinking water containing TCE and other solvents furthermore suggested
changes in visual function (Kilburn et al., 2002a; Reif et al., 2003). These studies used more
direct measures of visual function as compared to Rasmussen et al. (1993b), but their exposure
assessment is more limited because TCE exposure is not assigned to individual subjects (Kilburn
et al., 2002a), or because there are questions regarding control selection (Kilburn et al., 2002a)
and exposure to several solvents (Kilburn et al., 2002a; Reif et al., 2003).
Additional evidence of effects of TCE exposure on visual function is provided by a
number of laboratory animal studies demonstrating that acute or subchronic TCE exposure
causes changes in visual evoked responses to patterns or flash stimulus (Boyes et al., 2003, 2005;
Blain et al., 1994). Animal studies have also reported that the degree of some effects is
correlated with simultaneous brain TCE concentrations (Boyes et al., 2003, 2005) and that, after
a recovery period, visual effects return to control levels (Blain et al., 1994; Rebert et al., 1991).
Overall, the human and laboratory animal data together suggest that TCE exposure can cause
impairment of visual function, and some animal studies suggest that some of these effects may
be reversible with termination of exposure.
Studies of human subjects exposed to TCE either acutely in chamber studies or
chronically in occupational settings have observed deficits in cognition. Five chamber studies
reported statistically significant deficits in cognitive performance measures or outcome measures
suggestive of cognitive effects (Stewart et al., 1970; Gamberale et al., 1976; Triebig et al., 1976,
1977a; Gamberale et al., 1977). Danish degreasers with high cumulative exposure to TCE or
CFC-113 had a high risk [OR = 13.7, 95% CI; 2.0-92.0] for psychoorganic syndrome
characterized by cognitive impairment, personality changes, and reduced motivation, vigilance,
and initiative compared to workers with low cumulative exposure. Studies of populations living
in a community with contaminated groundwater also reported cognitive impairments (Kilburn
and Warshaw, 1993; Kilburn, 2002a), although these studies carry less weight in the analysis
because TCE exposure is not assigned to individual subjects and their methodological design is
weaker.
Laboratory studies provide some additional evidence for the potential for TCE to affect
cognition, though the predominant effect reported has been changes in the time needed to
complete a task, rather than impairment of actual learning and memory function (Kulig et al.,
1987; Kishi et al., 1993; Umezu et al., 1997). In addition, in laboratory animals, it can be
difficult to distinguish cognitive changes from motor-related changes. However, several studies
have reported structural or functional changes in the hippocampus, such as decreased
myelination (Issacson et al., 1990; Isaacson and Taylor, 1989) or decreased excitability of
hippocampal CA1 neurons (Ohta et al., 2001), although the relationship of these effects to
overall cognitive function is not established.
6/22/2009
372
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Two studies of TCE exposure, one chamber study of acute exposure duration and one
occupational study of chronic duration, reported changes in psychomotor responses. The
chamber study of Gamberale et al. (1976) reported a dose-related decrease in performance in a
choice reaction time test in healthy volunteers exposed to 100 and 200 ppm TCE for 70 minutes
as compared to the same subjects without exposure. Rasmussen et al. (1993c) reported a
statistically significant association with cumulative exposure to TCE or CFC-113 and
dyscoordination trend among Danish degreasers. Observations in a third study (Gun et al., 1978)
are difficult to judge given the author's lack of statistical treatment of data. In addition, Gash et
al. (2007) reported that 14 out of 30 TCE-exposed workers exhibited significantly slower fine
motor hand movements as measured through a movement analysis panel test. Studies of
population living in communities with TCE and other solvents detected in groundwater supplies
reported significant delays in simple and choice reaction times in individuals exposed to TCE in
contaminated groundwater as compared to referent groups (Kilburn, 2002a; Kilburn and
Warshaw, 1993; Kilburn and Thornton, 1996). Observations in these studies are more uncertain
given questions of the representativeness of the referent population, lack of exposure assessment
to individual study subjects, and inability to control for possible confounders including alcohol
consumption and motivation. Finally, in a presentation of 2 case reports, decrements in motor
skills as measured by the grooved pegboard and finger tapping tests were observed (Troster and
Ruff, 1990).
Laboratory animal studies of acute or sub-chronic exposure to TCE observed
psychomotor effects, such as loss of righting reflex (Umezu et al., 1997; Shih et al., 2001) and
decrements in activity, sensory-motor function, and neuromuscular function (Kishi et al., 1993;
Moser et al., 1995; Moser et al., 2003). However, two studies also noted an absence of
significant changes in some measures of psychomotor function (Kulig et al., 1987; Albee et al.,
2006). In addition, less consistent results have been reported with respect to locomotor activity
in rodents. Some studies have reported increased locomotor activity after an acute i.p. dosage
(Wolff and Siegmund, 1978) or decreased activity after acute or short term oral gavage dosing
(Moser et al., 1995, 2003). No change in activity was observed following exposure through
drinking water (Waseem et al., 2001), inhalation (Kulig et al., 1987) or orally during the
neurodevelopment period (Fredriksson et al., 1993).
Several neurochemical and molecular changes have been reported in laboratory
investigations of TCE toxicity. Kjellstrand et al. (1987) reported inhibition of sciatic nerve
regeneration in mice and rats exposed continuously to 150 ppm TCE via inhalation for 24 days.
Two studies have reported changes in GABAergic and glutamatergic neurons in terms of GABA
or glutamate uptake (Briving et al., 1986) or response to GABAergic antagonistic drugs (Shih et
al., 2001) as a result of TCE exposure, with the Briving et al. (1986) conducted at 50 ppm for 12
6/22/2009
373
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 months. Although the functional consequences of these changes is unclear, Tham et al. (1979,
2 1984) described central vestibular system impairments as a result of TCE exposure that may be
3 related to altered GABAergic function. In addition, several in vitro studies have demonstrated
4 that TCE exposure alters the function of inhibitory ion channels such as receptors for GAB Aa
5 glycine, and serotonin (Krasowski and Harrison, 2000; Beckstead et al., 2000; Lopreato et al.,
6 2003) or of voltage-sensitive calcium channels (Shafer et al., 2005).
6/22/2009
374
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 References
2
3 Albee, RR; Nitschke, KD; Mattsson, JL; et al. (1997) Dichloroacetylene: Effects on the rat
4 trigeminal nerve somatosensory evoked potential. Neurotoxicology and Teratology
5 19(l):27-37.
6
7 Albee, RR; Spencer, PJ; Johnson, KA; et al. (2006) Lack of trigeminal nerve toxicity in rats
8 exposed to trichloroethylene vapor for 13 weeks. Int J Toxicol 25(6):531-540.
9
10 Arito, H; Takahashi, M; Ishikawa, T. (1994) Effect of subchronic inhalation exposure to low-
11 level trichloroethylene on heart rate and wakefulness-sleep in freely moving rats. Jpn J Ind
12 Health (Sangyo Igaku) 36(1): 1-8.
13
14 ATSDR (Agency for Toxic Substances and Disease Registry). (2003) Impact of
15 trichloroethylene exposure on oral motor, speech, and hearing in children. U.S. Department of
16 Health and Human Services, Atlanta, GA.
17
18 Barret, L; Arsac, P; Vincent, M; et al. (1982) Evoked trigeminal nerve potential in chronic
19 trichloroethylene intoxication. J Toxicol Clin Toxicol 19(4):419-423.
20
21 Barret, L; Faure, J; Guilland, B; et al. (1984) Trichloroethylene occupational exposure: elements
22 for better prevention. Int Arch Occup Environ Health 53(4):283-289.
23
24 Barret, L; Garrel, S; Danel, V; et al. (1987) Chronic trichloroethylene intoxication: a new
25 approach by trigeminal-evoked potentials? Arch Environ Health 42(5):297-302.
26
27 Barret, L; Torch, S; Usson, Y Barret, L; Torch, S; Usson, Y; et al. (1991) A morphometric
28 evaluation of the effects of trichloroethylene and dichloroacetylene on the rat mental nerve.
29 Preliminary results. Neurosci Lett 131(2): 141-144.
6/22/2009
375
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Barret, L; Torch, S; Leray, CL; Sarlieve, L; Saxod, R (1992) Morphometric and Biochemical
Studies in Trigeminal Nerves of Rat after Trichlorethylene or Dichloracetylene Oral
Administration. Neurotoxicology 13:601-614.
Beckstead MJ, Weiner JL, Eger EI 2nd, Gong DH, Mihic SJ. (2000). Glycine and gamma-
aminobutyric acid(A) receptor function is enhanced by inhaled drugs of abuse. Mol Pharmacol.
57:1199-205.
Blain, L; Lachapelle, P; Molotchnikoff, S. (1994) Electroretinal responses are modified by
chronic exposure to trichloroethylene. Neurotoxicology 15(3):627—631.
Boyes, WK; Bushnell, PJ; Crofton, KM; et al. (2000) Neurotoxic and pharmacokinetic responses
to trichloroethylene as a function of exposure scenario. Environ Health Perspect 108(Suppl
2):317-322.
Boyes, WK; Bercegeay, M; Ali, JS; et al. (2003) Dose-based duration adjustments for the effects
of inhaled trichloroethylene on rat visual function. Toxicol Sci 76(1): 121-130.
Boyes, WK; Bercegeay, M; Krantz, T; et al. (2005) Momentary brain concentration of
trichloroethylene predicts the effects on rat visual function. Toxicol Sci 87(1): 187-196.
Briving, C; Jacobson, I; Hamberger, A; et al. (1986) Chronic effects of perchloroethylene and
trichloroethylene on the gerbil brain amino acids and glutathione. Neurotoxicology
7(1): 101-108.
Burg, JR; Gist, G; Allred, SL; et al. (1995) The national exposure registry - morbidity analyses
of noncancer outcomes from the trichloroethylene subregistry baseline data. Int J Occ Med
Toxicol 4(2):237-257.;
Burg, JR; Gist, GL. (1999) Health effects of environmental contaminant exposure: an intrafile
comparison of the Trichloroethylene Subregistry. Arch Environ Health 54(4) :231-241.
6/22/2009
376
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Buxton, P.H., and Hayward, M. (1967). Polyneuritis cranialis associated with industrial
2 trichloroethylene poisoning. J. Neurol. Neurosurg. Psych. 30, 511.
3 Chalupa, B., J. Synkova, and M. Sevcik 1960. The Assessment of Electroencephalographic
4 changes in memory disturbances in acute intoxications with industrial poisons, Brit. J. Ind. Med.
5 17,238.
6 Crofton, KM Lassiter, TL and Rebert, CS. Solvent-induced ototoxicity in rats: an atypical
7 selective mid-frequency hearing deficit, Hear. Res., 80 (1994, 25-30.
8
9 Crofton, KM; Zhao, X. (1997) The ototoxicity of trichloroethylene: extrapolation and relevance
10 of high-concentration, short-duration animal exposure data. Fundam Appl Toxicol
11 38(1): 101—106.
12
13 El Ghawabi, SM; Mansoor, MB; El Gamel, MS; et al. (1973) Chronic trichloroethylene
14 exposure. JEgyptMed Assoc 56(11-12):715-724.
15 Fechter, LD, Liu, Y., Herr, DW, Crofton, KM. Trichloroethylene ototoxicity: Evidence for a
16 cochlear origin., Toxicol. Sci. 42, 28-35, 1998.
17
18 Feldman, RG; Chirico-Post, J; Proctor, SP. (1988) Blink reflex latency after exposure to
19 trichloroethylene in well water. Arch Environ Health 43(2): 143-148.
20 Feldman, RG; Niles, C; Proctor, SP; et al. (1992) Blink reflex measurement of effects of
21 trichloroethylene exposure on the trigeminal nerve. Muscle Nerve 15(4):490-495.
22 Fredriksson. A: Danielsson. BR: Eriksson. P. (1993) Altered behavior in adult mice orally
23 exposed to tri- and tetrachloroethylene as neonates. Tox Lett 66:13-19.
24
25 Gamberale, F; Annwall, G; Olson, BA. (1976) Exposure to trichloroethylene III. Psychological
26 functions. Scand J Work Environ Health 2(4):220-224.
27
6/22/2009
377
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Gash, DM; Rutland, K; Hudson, NL; et al. (2008) Trichloroethylene: Parkinsonism and complex
2 1 mitochondrial neurotoxicity. Ann Neurol 63(2): 184-92.
3 Grandjean, E; Munchinger, R; Turrian, V; et al. (1955) Investigations into the effects of exposure
4 to trichloroethylene in mechanical engineering. Br J Ind Med 12(2): 131-142.
5 Guehl, D; Bezard, E; Dovero, S; et al. (1999) Trichloroethylene and parkinsonism: a human and
6 experimental observation. Eur J Neurol 6(5):609-611.
7
8 Gun, RT; Grygorcewicz, C; Nettelbeck, TJ. (1978) Choice reaction time in workers using
9 trichloroethylene. Med J Aust 1 (10):535—536.
10 Hirsch, AR; Rankin, KM; Panelli, PP. (1996) Trichloroethylene exposure and headache.
11 Headache Quarterly 7(2): 126-13 8.
12 Isaacson, LG; Taylor, DH. (1989) Maternal exposure to 1,1,2-trichloroethylene affects myelin in
13 the hippocampal formation of the developing rat. Brain Res 488(l-2):403-407.
14
15 Isaacson, LG; Spohler, SA; Taylor, DH. (1990) Trichloroethylene affects learning and decreases
16 myelin in the rat hippocampus. Neurotoxicol Teratol 12(4):375—381.
17
18 Jaspers, RMA; Muijser, H; Lammers, JH; et al. (1993) Mid-frequency hearing loss and reduction
19 of acoustic startle responding in rats following trichloroethylene exposure. Neurotoxicol Teratol
20 15(6):407-412.
21
22 Kilburn KH; Warshaw R; Thorton JC. (1987) Formaldehyde impairs memory, equilibrium, and
23 dexterity in histology technicians: effects which persist for days after exposure. Arch Environ
24 health 42:117-120.
25
26 Kilburn, KH; Warshaw, RH. (1992). Prospective study of neurobehavioral effects of
27 formaldehyde and solvents on histology technicians. Environ Res 58:134-146.
28
6/22/2009
378
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Kilburn, KH; Warshaw, RH. (1993) Effects on neurobehavioral performance of chronic exposure
2 to chemically contaminated well water. Toxicol Ind Health 9(3):391-404.
3 Kilburn KH; Warshaw, R; (1995a) Neurotoxic effects from residential exposure to chemicals
4 from an oil reprocessing facility and superfund site. Neurotoxicol Teratol 17:89-102.
5 Kilburn KH; Warshaw, RH. (1995b) Hydrogen sulfide and reduced-sulfur gases adversely affect
6 neurophysiological functions. Toxicol Ind Health 11:185-197.
7 Kilburn, KH; Thornton, JC. (1996) Prediction equations for simple and visual two-choice
8 reactions times in environmental neurotoxicology. Arch Environ Health 51(6):439-444.
9 Kilburn, KH. (2002a) Is neurotoxicity associated with environmental trichloroethylene (TCE)?
10 Arch Environ Health 57(2): 113—120.
11 Kilburn, KH. (2002b): Do duration of exposure, proximity to electronic manufacturing plants,
12 and involvement in a lawsuit affect chlorinated solvent toxicity? Arch Environ Health
13 57(2):121-126.
14 Kishi, R; Harabuchi, I; Ikeda, T; et al. (1993) Acute effects of trichloroethylene on blood
15 concentrations and performance decrements in rats and their relevance to humans. Br J Ind Med
16 50(5):470-480.
17
18 Kjellstrand, P; Lanke, J; Bjerkemo, M; et al. (1980) Irreversible effects of trichloroethylene
19 exposure on the central nervous system. Scand J Work Environ Health 6(l):40-47.
20
21 Kjellstrand, P; Kanje, M; Bjerkemo, M. (1987) Regeneration of the sciatic nerve in mice and rats
22 exposed to trichloroethylene. Toxicol Lett 38(1—2): 187—191.
23
24 Konietzko, H; Elster, I; Sayer, H; et al. (1975) [EEG changes under standardized
25 trichloroethylene load (author's transl)]. Int Arch Occup Environ Hlth 35(3-4): 257-264.
26 Originally published: Zentralnervose schaden durch trichlorathylen. Staub, Reinhaltung Der
27 Luft 3 5(6):240-241.
28
6/22/2009
379
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Krasowski, MD, Harrison, NL. 2000. The actions of ether, alcohol and alkane general
2 anaesthetics on GABAA and glycine receptors and the effects of TM2 and TM3 mutations. Br J
3 Pharmacol 129(4): 731 -43.
4
5 Kulig, BM. (1987) The effects of chronic trichloroethylene exposure on neurobehavioral
6 functioning in the rat. Neurotoxicol Teratol 9(2): 171-178.
7
8 Kylin, B; Axell, K; Samuel, HE; et al. (1967) Effect of inhaled trichloroethylene on the ens as
9 measured by optokinetic nystagmus. Arch Environ Hlth 15:48-52.
10
11 Landrigan, PJ; Kominsky, JR; Stein, GF; et al. (1987) Common-source community and industrial
12 exposure to trichloroethylene. Arch Environ Health 42(6):327-332.
13 LaurenoR. 1988. Letter to Editor on Noseworth and Rice (1988), Trichloroethylene poisoning
14 mimicking multiple sclerosis. Can J Neurol Sci 15:354.
15 Laureno R. 1993. Letter to Editor on Feldman et al. (1992), Blink reflex measurement of effects
16 of trichloroethylene. Muscle Nerve 16:217.
17 Lash TL, Green LC. 1993. Letter to Editor on Feldman et al. (1992), Blink reflex measurement
18 of effects of trichloroethylene. Muscle Nerve 16:217-218.
19 Lopreato, GF; Phelan, R; Borghese, CM; et al. (2003) Inhaled drugs of abuse enhance serotonin-
20 3 receptor function. Drug Alcohol Depend 70(1): 11-5.
21
22 Liu, YT; Jin, C; Chen, Z; et al. (1988) Increased subjective symptom prevalence among workers
23 exposed to trichloroethylene at sub-OEL levels. Tohoku J Exp Med 155(2): 183-195.
24
25 McCunney, R.J., Diverse manifestations of trichloroethylene (1988). Br J Ind Med 45:122-126.
26
6/22/2009
380
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Mhiri, C; Choyakh, F; Ben, HM; et al. (2004) Trigeminal somatosensory evoked potentials in
2 trichloroethylene-exposed workers. Neurosciences 9(2): 102-107.
3
4 Mitchell, A B S; Parsons-Smith, BG. (1969) Trichloroethylene neuropathy. Brit Med. J
5 1:422-423.
6
7 Moser, Cheek & MacPhail. A multidisciplinary approach to toxicological screening III.
8 Neurobehavioral toxicity. J Toxicol. Environ. Hlth., 1995, 45, 173-210.
9
10 Moser, VC; MacPhail, RC; Gennings, C. (2003) Neurobehavioral evaluations of mixtures of
11 trichloroethylene, heptachlor, and di(2-ethylhexyl)phthlate in a full-factorial design. Toxicology
12 188(2—3): 125—137.
13
14 Muijser, H; Lammers, J; Kulig, BM. (2000) Effects of exposure to trichloroethylene and noise on
15 hearing in rats. Noise and Health 6:57-66.
16
17 Nagaya, T; Ishikawa, N; Hata, H. (1990) No change in serum dopamine-beta-hydroxylase
18 activity in workers exposed to trichloroethylene. Tox Lett 54(2/3):221-227.
19
20 Niklasson, Tham, Larsby, & Eriksson, Effects of toluene, styrene, trichloroethylene and
21 trichloroethane on the vestibule and opto-oculo motor system in rats, Neurotoxicol. Teratol.,
22 1993, 15, 327-334.
23
24 Noland-Gerbec, Pfohl, Taylor and Bull. 2-deopxyglucose uptake in the developing rat brain upon
25 pre and postnatal exposure to trichloroethylene, Neurotoxicology, 7„ 157-164, 1986.
26
27 Ohta, M; Saito, T; Saito, K; et al. (2001) Effect of trichloroethylene on spatiotemporal pattern of
28 LTP in mouse hippocampal slices. Int J Neurosci 111(3-4):257-271.
6/22/2009
381
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Rasmussen, K; Sabroe, S. (1986) Neuropsychological symptoms among metal workers exposed
2 to halogenated hydrocarbons. Scand J Soc Med 14(3): 161-168.
3
4 Rasmussen, K; Jeppesen, HJ; Sabroe, S. (1993a) Solvent-Induced Chronic Toxic
5 Encephalopathy. American Journal of Industrial Medicine 23:779-792.
6
7 Rasmussen, K; Jeppesen, HJ; Sabroe, S. (1993b) Psychometric tests for assessment of brain
8 function after solvent exposure. Am J Ind Med 24(5):553-565.):
9
10 Rasmussen, K; Arlien-Soborg, P; Sabroe, S. (1993c) Clinical neurological findings among metal
11 degreasers exposed to chlorinated solvents. Acta Neurol Scand 87(3):200-204.
12
13 Rebert CS, Day VL, Matteucci MJ, Pryor GT (1991) Sensory-evoked potentials in rats
14 chronically exposed to trichloroethylene: predominant auditory dysfunction. Neurotoxicol
15 TeratoL 13(1):83—90
16
17 Rebert CS, Boyes WK, Pryor GT, Svensgaard DJ, Kassay KM, Gordon GR, Shinsky N (1993).
18 Combined effects of solvents on the rat's auditory system: styrene and trichloroethylene. Int J
19 Psvchophvsiol. 14(1):49—59
20
21 Rebert CS, Schwartz RW, Svendsgaard DJ, Pryor GT, Boyes WK (1995). Combined effects of
22 paired solvents on the rat's auditory system. Toxicology. 105(2-3):345-54.
23
24 Reif, JS; Burch, JB; Nuckols, JR; et al. (2003) Neurobehavioral effects of exposure to
25 trichloroethylene through a municipal water supply. Environ Res 93(3):248-258.
26
27 Ruijten, M; Verberk, MM; Salle, HJ. (1991) Nerve function in workers with long term exposure
28 to trichloroethene. Br J Ind Med 48(2):87-92.
6/22/2009
382
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
2 Salvini, M., Binasch, S., and Riva, M. (1971). Evaluation of the psychophysiological functions in
3 humans exposed to trichloroethylene. Br J Ind Med, 28:293-295.
4 Saunders, RA. 1967. A new hazard in closed environmental atmospheres. Arch Environ Health.
5 14:380-384.
6 Shafer, TJ; Bushnell, PJ; Benignus, VA; et al. (2005) Perturbation of voltage-sensitive Ca2+
7 channel function by volatile organic solvents. J Pharmacol Exp Ther 315(3): 1109-1118.
8
9 Shih, CL, Chen, HH, Chiu, TH. Acute exposure to trichloroethylene differentially alters the
10 susceptibility to chemoconvulsants in mice. Toxicology, 162, 2001, 35-42.
11 Smith, GF. (1970) The investigation of the mental effects of trichloroethylene. Ergonomics
12 13(5):580-586.
13
14 Stewart, RD; Dodd, HC; Gay, HH; et al. (1970) Experimental human exposure to
15 trichloroethylene. Arch Environ Health 20(1):64-71.
16
17 Tham, R; Larsby, B; Odkvist, LM; et al. (1979) The influence of trichloroethylene and related
18 drugs on the vestibular system. Acta Pharmacol Toxicol (Copenh) 44(5):336-342.
19
20 Tham, R; BUNNFORS, I; Eriksson, B; et al. (1984) Vestibulo ocular disturbances in rats
21 exposed to organic solvents. Acta Pharmacol Toxicol 54(l):58-63.
22
23 Triebig, G; Essing, HG; Schaller, KH; et al. (1976) Biochemical and psychological examinations
24 of trichloroethylene exposed volunteers. Zentralbl Bakteriol [Orig B] 163(5-6):383-416.
25 (translated from German)
26
6/22/2009
383
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Triebig, G., S. Lehrl, W. Kinzel, H. Erzigkeit, J. V. Galster and K. H. Schaller (1977a). Psycho-
2 Pathometric Results of Follow-up studies of Persons Exposed to Trichloroethylene. Zbl. Bakt.
3 Hyg., I. Abt. Orig. B 164, 314-327 (1977). (translated from German)
4 Triebig, G; Schaller, KH; Erzigkeit, H; et al. (1977b). Biochemical investigations and
5 psychological studies of persons chronically exposed to trichloroethylene with regard to non-
6 exposure intervals. Int Arch Occup Environ Health 38(3): 149-162. (translated from German)
7
8 Triebig, G; Trautner, P; Weltle, D; Saure, E; Valentin, H. (1982). Investigations on the
9 Neurotoxicity of processing materials III. Measurement of Motor and Sensory Nerve
10 Conduction Velocity in Persons with Job-Related Trichloroethylene Exposure. Int Arch Occup
11 Environ Health (1982) 51: 25-34 (translated from German)
12
13 Triebig, G., W. Bestler, P. Baumeister, H. Valentin (1983). Investigations on the neurotoxicity
14 of chemical substances at the workplace, Part IV: Determination of the motor and sensory nerve
15 conduction velocity in persons occupationally exposed to a mixture of organic solvents. Int.
16 Arch. Occup. Environ. Health, 52, 139-150. (translated from German)
17
18 Troster, AI; Ruff, RM. (1990) Neuropsychological sequelae of exposure to the chlorinated
19 hydrocarbon solvents trichloroethylene and trichloroethane. Archives of Clinical
20 Neuropsychology 5(1):31-47.
21
22 Umezu, T; Yonemoto, J; Soma, Y; et al. (1997) Behavioral effects of trichloroethylene and
23 tetrachloroethylene in mice. Pharmacol Biochem Behav 58(3):665-671.
24
25 Vernon RJ, Ferguson RK. (1969) Effects of trichloroethylene on visual-motor performance.
26 Arch Environ Health 18(6):894-900.
27
6/22/2009
384
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Waseem, M; Ali, M; Dogra, S; et al. (2001) Toxicity of trichloroethylene following inhalation
2 and drinking contaminated water. J Appl Toxicol 21(6):441-444.
3
4 Windemuller, FJ; Ettema, JH. (1978) Effects of combined exposure to trichloroethylene and
5 alcohol on mental capacity. Int Arch Occup Environ Health 41(2):77-85.
6
7 Winneke, G. Acute behavioral effects of exposure to some organic solvents-
8 psychophysiological aspects (1987), Acta Neurol Scand, 32; 117-129.
9
10 Wolff & Siegmund, the effect of trichloroethylene on the spontaneous locomotor activity of mice
11 and of tetrachloroethane on the mortality of mice as a function of the time of day. Biol. Zbl, 97,
12 345-351, 1978.
13
14 Yamamura, K; Ikeda, I; Sadamoto, T; et al. (1983) Effects of Trichloroethylene Exposure on
15 Hearing. An Investigation of Cochlear Microphonics and Action Potential of the Guinea Pig.
16 European Journal of Applied Physiology and Occupational Physiology 52:47-50.
17
6/22/2009
385
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.3 KIDNEY TOXICITY AND CANCER
4.3.1 Human studies of kidney
4.3.1.1 Nonspecific Markers of Nephrotoxicity
Investigations of nephrotoxicity in human populations show that highly exposed workers
exhibit evidence of damage to the proximal tubule (NRC, 2006). The magnitude of exposure
needed to produce kidney damage is not clear. Observation of elevated excretion of urinary
proteins in the four studies (Briining et al., 1999a, b; Bolt et al., 2004; Green et al., 2004)
indicates the occurrence of a toxic insult among TCE-exposed subjects compared to unexposed
controls. Two studies are of subjects with previously diagnosed kidney cancer (Briining et al.,
1999a; Bolt et al., 2004), subjects in Briining et al. (1999b) and Green et al. (2004) are disease
free. Urinary proteins are considered nonspecific markers of nephrotoxicity and include: al-
Microglobulin, albumin, and NAG (Price et al., 1999, 1996; Lybarger et al., 1999). Four studies
measure al-microglobulin with elevated excretion observed in the German studies (Briining et
al., 1999a, b; Bolt et al., 2004) but not Green et al. (2004). However, Green et al. (2004) found
statistically significant group mean differences in NAG, another nonspecific marker of tubular
toxicity, in disease free subjects. Observations in Green et al. (2004) provide evidence of tubular
damage among workers exposed to trichloroethylene at 32 ppm (mean) [range, 0.5-252 ppm].
Elevated excretion of NAG as a nonspecific marker of tubular damage has also been observed
with acute TCE poisoning (Carrieri et al., 2007). These and other studies relevant to evaluating
TCE nephrotoxicity are discussed in more detail below.
Biological monitoring of persons who previously experienced "high" exposures to
trichloroethylene (100-500 ppm) in the workplace show altered kidney function evidenced by
urinary excretion of proteins suggestive of renal tubule damage. Similar results were observed in
the only study available of subjects with TCE exposure at current occupational limits (NRC,
2006). Table 4.3.1 provides details and results from these studies. Briining et al. (1999a) report
a higher prevalence of elevated proteinuria suggestive of slight to severe tubular damage and an
elevated excretion of al-microglobin, another urinary biomarker of renal tubular function, was
observed in 41 renal cell carcinoma cases with prior trichloroethylene exposure and with pending
workman's compensation claims compared with the non-exposed renal cell cancer patients and
to hospitalized surgical patients. The lack of statistical treatment of proportions and control for
possible confounding from difference in renal cancer stage and blood pressure between
trichloroethylene exposure and non-exposure cases are uncertainties. Similarly, severe tubular
proteinuria is seen in 14 of 39 workers (35%) exposed to trichloroethylene in the electrical
6/22/2009
386
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
department, fitters shop and through general degreasing operations of felts and sieves in a
cardboard manufacturing factory (Briming et al., 1999b). No subjects of 46 non-exposed males
office and administrative workers from the same factory demonstrate severe tubular proteinuria,
although slight tubular proteinuria is seen in 20% of exposed workers and in 2% of nonexposed
workers (Briining et al., 1999b). Exposed subjects also had statistically significantly elevated
levels of a 1-microglobulin compared to unexposed controls. Furthermore, subjects with tubular
damage as indicated by urinary protein patterns had higher GSTa concentrations than non-
exposed subjects (p < 0.001). Both sex and use of spot or 24-hour urine samples are shown to
influence al-microglobulin (Andersson et al., 2008); however, these factors are not considered to
greatly influence observations given only males were subjects and al-microglobulin levels in
spot urine sample are adjusted for creatinine concentration.
Bolt et al. (2004) measured a 1-microglobulin excretion in living subjects from the renal
cell carcinoma case-control study by Briining et al. (2003). Some subjects in this study were
highly exposed. Of the 134 with renal cell cancer, 19 reported past exposures that led to narcotic
effects and 18 of the 401 controls, experienced similar effects (odds ratio [OR] = 3.71, 95%
confidence interval [CI] 1.80-7.54) (Briining et al., 2003). Bolt et al. (2004) found that al-
microglobulin excretion increased in exposed renal cancer patients compared with non-exposed
patients controls. A lower proportion of exposed cancer patients had normal al-microglobulin
excretion, less than 5 mg/L, the detection level for the assay and the level considered by these
investigators as associated with no clinical or subclinical tubule damage, and a higher proportion
of high values, defined as >45 mg/L, compared to cases who did not report TCE occupational
exposure and to non-exposed controls. The lack of statistical treatment of proportions and
unadjusted urinary values for creatinine are uncertainties. On the other hand, reduced clearance
attributable to renal cancer does not explain the lower percentage of normal values among
exposed cases given findings of similar prevalence of normal excretion among unexposed renal
cell cases and controls.
In their study of 70 current employees (58 males, 12 females) of an electronic factory
with trichloroethylene exposure and 54 (50 males, 4 females) age-matched subjects drawn from
hospital or administrative staff, Green et al. (2004) found that urinary excretion of albumin, total
N-acetyl-P-D-glucosaminidase (NAG) and formate were increased in the exposed group
compared with the unexposed group4. No differences between exposed and unexposed subjects
were observed in other urinary proteins, including al-microglobulin, p2-microglobulin, and
4 Elevation of NAG in urine is a sign of proteinuria, and proteinuria is both a sign and a cause of kidney malfunction
(Zandi-Nejad et al., 2004). For a urine sample, 10-17 mg of albumin per g of creatinine is considered to be
suspected albuminuria in males (15-25 in females) (De Jong and Brenner 2004).
6/22/2009
387
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
GSTa. Green et al. (2004) stated that NAG is not an indicator of nephropathy, or damage, but
rather is an indicator of functional change in the kidney. Green et al. (2004) further concluded
that increased urinary albumin or NAG was not related to trichloroethylene exposure; analyses to
examine the exposure-response relationship found neither NAG or albumin concentration
correlated to urinary-TCA (U-TCA) or employment duration (years). NRC (2006) did not
consider U-TCA as sufficiently reliable to use as a quantitative measure of TCE exposure,
concluding that the data reported by Green et al. (2004) were inadequate to establish exposure-
response information because the relationship between U-TCA and ambient TCE intensity is
highly variable and nonlinear, and conclusions about the absence of association between TCE
and nephrotoxicity can not be made based on U-TCA. Moreover, use of employment duration
does not consider exposure intensity differences between subjects with the same employment
duration, and bias introduced through misclassification of exposure may explain the Green et al.
(2004) findings.
Selden et al. (1993) in their study of 29 metal workers (no controls) reported a correlation
between NAG and U-TCA (r = 0.48, p < 0.01) but not with other exposure metrics of recent or
long-term exposure. Personal monitoring of worker breath indicated median and mean time-
weighted-average TCE exposures of 3 and 5 ppm, respectively. Individual NAG concentrations
were within normal reference values. Rasmussen et al. (1993), also, reported a positive
relationship (p = 0.05) between increasing urinary NAG concentration (adjusted for creatinine
clearance) and increasing duration in their study of 95 metal degreasers (no controls) exposed to
either TCE (70 subjects) or CFC 113(25 subjects). Multivariate regression analyses which
adjusted for age were suggestive of an association between NAG and exposure duration
(p = 0.011). Mean urinary NAG concentration was higher among subjects with annual exposure
of >30 hours/week, defined as peak exposure, compared to subjects with annual exposure of less
than <30 hours/week (72.4 + 44.1 ug/g creatinine compared to 45.9 + 30.0 ug/g creatinine,
p<0.01).
Nagaya et al. (1989) did not observe statistically significant group differences in urinary
P2-microglobulin and total protein in spot urine specimens of male degreasers and their controls,
nor were these proteins correlated with urinary total trichloro-compounds (U-TTC). The paper
lacks details on subject selection, whether urine collection was at start of work week or after
sufficient exposure, and presentation of ^-values and correlation coefficients. The presentation
of urinary protein concentrations stratified by broad age groups is less statistically powerful than
examination of this confounder using logistic regression. Furthermore, although valid for
pharmacokinetic studies, examination of renal function using urinary TTC as a surrogate for
TCE exposure is uncertain, as discussed above for Green et al. (2004).
6/22/2009
388
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.3.1.2 End-stage Renal Disease
End-stage renal disease is associated with hydrocarbon exposure, a group which includes
trichloroethylene, 1,1, 1-trichloroethane, and JP4 (jet propellant 4), in the one study examining
this endpoint (Radican et al., 2006). Table 4.3.1 provides details and results from Radican et al.
(2006). This study assessed end-stage renal disease in a cohort of aircraft maintenance workers
at Hill Air Force Base (Blair et al., 1998) with strong exposure assessment to trichloroethylene
(NRC, 2006). Other occupational studies do not examine end-stage renal disease specifically,
instead reporting relative risks associated with deaths due to nephritis and nephrosis (Boice et al.,
1999, 2006; ATSDR, 2004), all genitourinary system deaths (Garabrant et al., 1988; Costa et al.,
1989; Ritz, 1999), or providing no information on renal disease mortality in the published paper
(Blair et al., 1998; Morgen et al., 1998; Chang et al., 2003).
4.3.2 Human studies of kidney cancer
Cancer of the kidney and renal pelvis is the 6th leading cause of cancer in the United
States with an estimated 54,390 (33,130 men and 21,260 women) newly diagnosed cases and
13,010 deaths (Jemal et al., 2008; Ries et al., 2008). Age-adjusted incidence rates based on cases
diagnosed in 2001-2005 from 17 SEER geographic areas are 18.3 per 100,000 for men and 9.2
per 100,000 for women. Age-adjusted mortality rates are much lower; 6.0 per 100,000 for men
and 2.7 for women.
Cohort, case-control, and geographical studies have examined trichloroethylene and
kidney cancer, defined either as cancer of kidney and renal pelvis in cohort and geographic based
studies or as renal cell carcinoma, the most common type of kidney cancer, in case-control
studies. Appendix C identifies these study's design and exposure assessment characteristics.
Observations in these studies are presented below in Table 4.3.3. Rate ratios for incidence
studies in Table 4.3.3 are, generally, larger than for mortality studies.
Additionally, a large body of evidence exists on kidney cancer risk and either job or
industry titles where trichloroethylene usage has been documented. TCE has been used as a
degreasing solvent in a number of jobs, task, and industries, some of which include metal,
electronic, paper and printing, leather manufacturing and aerospace/aircraft manufacturing or
maintenance industries and job title of degreaser, metal workers, electrical worker, and machinist
(IARC, 1995; Bakke et al., 2007). NRC (2006) identifies characteristics for kidney cancer case-
control studies that assess job title or occupation in their Table 3-8. Relative risks and 95%
confidence intervals reported in these studies are found in Table 4.3.4 below.
6/22/2009
389
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.3.2.1 Studies of Job Titles and Occupations with Historical TCE Usage
Elevated risks are observed in many of the cohort or case-control studies between kidney
cancer and industries or job titles with historical use of trichloroethylene (Partenen et al., 1991;
McCredie and Stewart, 1993; Schlehofer et al., 1995; Mandel et al., 1995; Pesch et al., 2000a;
Parent et al., 2000; Mattioli et al., 2002; Briining et al., 2003; Zhang et al., 2004; Charbotel et al.,
2006; Wilson et al., 2008). Overall, these studies, although indicating association with metal
work exposures and kidney cancer, are insensitive for identifying a TCE hazard. The use of job
title or industry as a surrogate for exposure to a chemical is subject to substantial
misclassification that will attenuate rate ratios due to exposure variation and differences among
individuals with the same job title. Several small case-control studies (Jensen et al., 1988;
Harrington et al., 1989; Sharpe et al., 1989; Auperin et al., 1994; Vamvakas et al., 1998; Parent
et al., 2000) have insufficient statistical power to detect modest associations due to their small
size and potential exposure misclassification (NRC, 2006). For these reasons, statistical
variation in the risk estimate is large and observation of statistically significantly elevated risks
associated with metal work in many of these studies is noteworthy. Some studies also examined
broad chemical grouping such as degreasing solvents or chlorinated solvents. Observations in
studies that assessed degreasing agents or chlorinated solvents reported statistically significant
elevated kidney cancer risk (Asal et al., 1998; Harrington et al., 1989; McCredie and Stewart,
1993; Mellemgaard et al., 1994, Schlehofer et al., 1995; Pesch et al., 2000a; Briining et al.,
2003). Observations of association with degreasing agents together with job title or occupations
where TCE has been used historically provide a signal and suggest an etiologic agent common to
degreasing activities.
4.3.2.2 Cohort and Case-Controls Studies of TCE Exposure
Cohort and case-controls studies that include job-exposure matrices for assigning TCE
exposure potential to individual study subjects show associations with kidney cancer, specifically
renal cell carcinoma, and trichloroethylene exposure. Support for this conclusion derives from
findings of increased risks in cohort studies (Henschler et al., 1995; Raaschou-Nielsen et al.,
2003; Zhao et al., 2005) and in case-control studies from the Arnsberg region of Germany
(Vamvakas et al., 1998; Pesch et al., 2000a; Briining et al., 2003), the Arve Valley region in
France (Charbotel et al., 2006, 2009), and the United States (Sinks et al., 1992; Dosemeci et al.,
1999).
A consideration of a study's statistical power and exposure assessment approach is
necessary to interpret observations in Table 4.3.3. Most cohort studies are underpowered to
detect a doubling of kidney cancer risks including the essentially null studies by Greenland et al.
6/22/2009
390
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(1994), Axelson et al. (1994 [incidence]), Anttila et al. (1995 [incidence]), Blair et al. (1998
[incidence and mortality]), Morgan et al. (1998), Boice et al. (1999) and Hansen et al. (2001).
Only the exposure duration-response analysis of Raaschou-Nielsen et al. (2003) had over 80%
statistical power to detect a doubling of kidney cancer risk (NRC, 2006), and they observed a
statistically significant association between kidney cancer and >5 year employment duration.
Rate ratios estimated in the mortality cohort studies of kidney cancer (e.g., Garabrant et al.,
1988; Sinks et al., 1992; Axelson et al., 1994; Greenland et al., 1994; Blair et al., 1998; Morgan
et al., 1998; Ritz, 1999; Boice et al., 1999, 2006) are likely underestimated to some extent
because of nondifferential misclassification of outcome in these studies, although the magnitude
is difficult to predict (NRC, 2006). Cohort studies with more uncertain exposure assessment
approaches, e.g., studies of all subjects working at a factory (Garabrant et al., 1998; Costa et al.,
1989; Sung et al., 2007; Chang et al., 2003, 2005; Clapp and Hoffmann, 2008), do not show
association but are quite limited given their lack of attribution of higher or lower exposure
potentials; risks are likely diluted due to their inclusion of no or low exposed subjects.
Two studies were carried out in geographic areas with a high frequency and a high degree
of TCE exposure and were designed with a priori hypotheses to test for the effects of TCE
exposure on renal cell cancer risk (Briining et al., 2003; Charbotel et al., 2006, 2009) and for this
reason their observations have important bearing to the epidemiologic evidence evaluation. Both
studies found a 2-fold elevated risk with any TCE exposure after adjustment for several possible
confounding factors including smoking (2.47, 95% CI: 1.36, 4.49) for self-assessed exposure to
TCE (Briining et al., 2003); high cumulative TCE exposure (2.16, 95% CI: 1.02, 4.60) with a
positive and statistically significant trend test, p = 0.04, (Charbotel et al., 2006). Furthermore,
renal cell carcinoma risk in Charbotel et al. (2005) increased to over 3-fold (95% CI: 1.19, 8.38)
in statistical analyses which considered a 10 year exposure lag period. An exposure lag period is
often adopted in analysis of cancer epidemiology to reduce exposure measurement biases
(Salvan et al., 1995). Most exposed cases in this study were exposed to TCE below any current
occupational standard (26 of 37 cases [70%]) had held a job with a highest time weighted
average (TWA) <50 ppm] (Charbotel et al., 2009). A subsequent analysis of Charbotel et al.
(2009) using an exposure surrogate defined as the highest TWA for any job held, an inferior
surrogate given TCE exposures in other jobs were not considered, reported an almost 3-fold
elevated risk (2.80, 95% CI: 1.12, 7.03) adjusted for age, sex, body mass index (BMI), and
smoking with exposure to TCE in any job to >50 ppm TWA (Charbotel et al., 2009).
Zhao et al. (2005) compared test-stand workers at a California aerospace company to
non-exposed workers from the same company as the internal referent population, and found a
monotonic increase in incidence of kidney cancer by increasing cumulative TCE exposure. In
addition, a 5-fold increased incidence was associated with high cumulative TCE exposure. This
6/22/2009
391
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
relationship for high cumulative TCE exposure, lagged 20 years, was accentuated with
adjustment for other occupational exposures (RR = 7.40, 95% CI: 0.47, 116), although the
confidence intervals were increased. An increased confidence interval with adjustments is not
unusual in occupational studies, as exposure is usually highly correlated with them, so that
adjustments often inflate standard error without removing any bias (NRC, 2006). Observed risks
were lower for kidney cancer mortality and are likely underestimated because of nondifferential
misclassification of outcome. Boice et al. (2006), another study of workers at this company and
which overlaps with Zhao et al. (2005), found a 2-fold increase in kidney cancer mortality (SMR
= 2.22, 95% CI: 0.89, 4.57). This study examined mortality in a cohort whose definition date
differs slightly from Zhao et al. (2005) and used a qualitative approach for TCE exposure
assessment.
Zhao et al. (2005) and Charbotel et al. (2006) are two of the few studies to conduct a
detailed assessment of exposure that allowed for the development of a job-exposure matrix that
provided rank-ordered levels of exposure to TCE and other chemicals. Their inclusion of rank-
ordered exposure levels is a strength compared to more inferior exposure assessment approaches
in some other studies such as duration of exposure or a grouping of all exposed subjects.
The finding in Raaschou-Nielsen et al. (2003) of an elevated renal cell carcinoma risk
with longer employment duration is noteworthy given this study's use of a relatively insensitive
exposure assessment approach. One strength of this study is the presentation of incidence ratios
for a subcohort of higher exposed subjects, those with at least 1-year duration of employment
and first employment before 1980, as a sensitivity analysis for assessing the effect of possible
exposure misclassification bias. Renal cell carcinoma risk was higher in this subcohort
compared to the larger cohort and indicated some potential for misclassification bias in the
grouped analysis. For both the cohort and subcohort analyses, risk appeared to increase with
increasing employment duration, although formal statistical tests for trend are not presented in
the published paper.
4.3.2.2.1 Discussion of Controversies on Studies in the Arnsberg Region of Germany
Two previous studies of workers in this region, a case-control study of Vamvakas et al.
(1998) and Henschler et al. (1995), a study prompted by a kidney cancer case cluster, observed
strong associations between kidney cancer and TCE exposure. A fuller discussion of the studies
from the Arnsberg region and their contribution to the overall weight of evidence on cancer
hazard is warranted in this evaluation given the considerable controversy (Bloemen and
Tomenson, 1995; Swaen, 1995; McLaughlin and Blot, 1997; Green and Lash, 1999; Cherrie et
al., 2001; Mandel, 2001) surrounding Henschler et al. (1995) and Vamvakas et al. (1998).
6/22/2009
392
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Criticisms of Henschler et al. (1995) and Vamvakas et al. (1998) relate, in part, to
possible selection biases that would lead to inflating observed associations and limited inferences
of risk to the target population. Specifically, these include (1) the inclusion of kidney cancer
cases first identified from a cluster and the omission of subjects lost to follow-up from Henschler
et al. (1995); (2) use of a Danish population as referent, which may introduce bias due to
differences in coding cause of death and background cancer rate differences (Henschler et al.,
1995); 3) follow-up of some subjects outside the stated follow-up period (Henschler et al., 1995);
4) differences between hospitals in the identification of cases and controls in Vamvakas et al.
(1998); 5) lack of temporality between case and control interviews (Vamvakas et al., 1998); 6)
lack of blinded interviews (Vamvakas et al., 1998); 7) age differences in Vamvakas et al. (1998)
cases and controls that may lead to a different TCE exposure potential;, 8) inherent deficiencies
in Vamvakas et al. (1998) as reflected by its inability to identify other known kidney cancer risk
factors; and, 9) exposure uncertainty, particularly unclear intensity of TCE exposure. Overall,
NRC (2006) noted that some of the points above may have contributed to an underestimation of
the true exposure distribution of the target population (points 5, 6, and 7), other points would
underestimate risk (points 3), and that these effects could not have explained the entire excess
risk observed in these studies (points 1, 2, 4). The NRC (2006) furthermore disagreed with the
exposure uncertainty criticism (point 9), and concluded TCE exposures, although of unknown
intensity, were substantial and, clearly showed graded differences on several scales in Vamvakas
et al. (1998) consistent with this study's semi-quantitative exposure assessment.
Briining et al. (2003) was carried out in a broader region in southern Germany, which
included the Arnsberg region and a different set of cases and control identified from a later time
period than Vamvakas et al. (1998). The TCE exposure range in this study was similar to that in
Vamvakas et al. (1998), although at a lower exposure prevalence because of the larger and more
heterogeneous ascertainment area for cases and controls. For "ever exposed" to TCE, Briining et
al. (2003) observed a risk ratio of 2.47 (95% CI: 1.36, 4.49) and a 4-fold increase in risk (95%
CI; 1.80, 7.54) among subjects with any occurrence of narcotic symptom and a 6-fold increase in
risk (95%) CI: 1.46, 23.99) for subjects who had daily occurrences of narcotic symptoms; risks
which are lower than observed in Vamvakas et al. (1998). The lower rate ratio in Briining et al.
(2003) might indicate bias in the Vamvakas et al. study or statistical variation between studies
related to the broader base population included in Briining et al. (2003).
Observational studies such as epidemiologic studies are subject to biases and
confounding which can be minimized but never completely eliminated through a study's design
and statistical analysis methods. While Briining et al. (2003) overcomes many of the
deficiencies of Henschler et al. (1995) and Vamvakas et al. (1998), nonetheless, possible biases
and measurement errors could be introduced through their use of prevalent cases and residual
6/22/2009
393
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
noncases, use of controls from surgical and geriatric clinics, non-blinding of interviewers, a 2-
year difference between cases and controls in median age, use or proxy or next-of-kin interviews,
and self-reported occupational history.
The impact of any one of the above points could either inflate or depress observed
associations. Biases related to a longer period for case compared to control ascertainment could
go in either direction. Next-of-kin interviewers for deceased cases, all controls being alive at the
time of interview, would be expected to underestimate risk if exposures were not fully reported
and thus misclassified. On the other hand, the control subjects who were enrolled when the
interviews were conducted might not represent the true exposure distribution of the target
population through time and would lead to overestimate of risk. Selection of controls from
clinics is not expected to greatly influence observed associations since these clinics specialized
in the type of care they provided (NRC, 2006). Briining et al. (2003) is not the only kidney case-
control study where interviewers were not blinded; in fact, only the study of Charbotel et al.
(2006) included blinding of interviewers. Blinding of interviewers is preferred to reduce
possible bias. Briining et al.'s use of frequency matching using 5-year age groupings is common
in epidemiologic studies and any biases introduced by age difference between cases and controls
is expected to be minimal because the median age difference was 3 years.
Despite these issues, the three studies of the Arnsberg region, with very high apparent
exposure and different base populations showed a significant elevation of risk and all have
bearing on kidney cancer hazard evaluations. The emphasis provided by each study for
identifying a kidney cancer hazard depends on its strengths and weaknesses. Briining et al.
(2003) overcomes many of the deficiencies in Henschler et al. (1995) and Vamvakas et al.
(1998). The finding of a statistically significantly approximately 3-fold elevated odds ratio with
occupational TCE exposure in Briining et al. (2003) strengthens the signal previously reported by
Henschler et al. (1995) and Vamvakas et al. (1998). A previous study of cardboard workers in
the United States (Sink et al., 1992), a study like Henschler et al. (1995) which was prompted by
a reported cancer cluster, had observed association with kidney cancer incidence, particularly
with work in the finishing department where TCE use was documented. Henschler et al. (1995),
Vamvakas et al. (1998) and Sinks et al. (1992) are less likely to provide a precise estimate of the
magnitude of the association given greater uncertainty in these studies compared to Briining et
al. (2003). For this reason, Briining et al. (2003) is preferred for meta-analysis treatment since it
is considered to better reflect risk in the target population than the two other studies. Another
study (Charbotel et al., 2006) of similar exposure conditions of a different base population and of
different case and control ascertainment methods as the Arnsberg region studies has become
available since the Arnsberg studies. This study shows a statistically significant elevation of risk
and high cumulative TCE exposure in addition to a positive trend with rank-order exposure
6/22/2009
394
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
levels. Charbotel et al. (2006) adds evidence to observations from earlier studies on high TCE
exposures in Southern Germany and suggests that peak exposure may add to risk associated with
cumulative TCE exposure.
4.3.2.3 Examination of Possible Confounding Factors
Examination of potential confounding factors is an important consideration in the
evaluation of observations in the epidemiologic studies on TCE and kidney cancer. A known
risk factor for kidney cancer is cigarette smoking. Obesity, diabetes, hypertension and
antihypertensive medications, and analgesics are linked to kidney cancer, but causality has not
been established (Moore et al., 2005; McLaughlin et al., 2006). On the other hand, fruit and
vegetable consumption is considered protective of kidney cancer risk (McLaughlin et al., 2006).
Studies by Asal et al. (1988), Partanen et al. (1991), McCredie and Stewart (1993), Auperin et al.
(1994), Chow et al. (1994), Mellemgaard et al. (1994), Mandel et al. (1995), Vamvakas et al.
(1998), Dosemeci et al. (1999), Pesch et al. (2000a), Briining et al. (2003), and Charbotel et al.
(2006) controlled for smoking and all studies except Pesch et al. (2000a) controlled for body
mass index (BMI). Vamvakas et al. (1998) and Dosemeci et al. (1999) controlled for
hypertension and or diuretic intake in the statistical analysis. Because it is unlikely that exposure
to trichloroethylene is associated with smoking, body mass index, hypertension, or diuretic
intake, these possible confounders do not significantly affect the estimates of risk (NRC, 2006).
Direct examination of possible confounders is less common in cohort studies than in
case-control studies where information is obtained from study subjects or their proxies. Use of
internal controls, such as for Zhao et al. (2005), in general minimizes effects of potential
confounding due to smoking or socioeconomic status since exposed and referent subjects are
drawn from the same target population. Effect of smoking as a possible confounder may be
assessed indirectly through (1) examination of risk ratios for other smoking-related sites and (2)
examination of the expected contribution by these three factors to cancer risks. Lung cancer risk
in Zhao et al. (2005) was not elevated compared to referent subjects and this observation
suggests smoking patterns were similar between groups. Smoking was more prevalent in the
Raaschou-Nielsen et al. (2003) cohort than the background population as suggested by the
elevated risks for lung and other smoking-related sites; however, Raaschou-Nielsen et al. (2003)
do not consider smoking to fully explain the 20 and 40 percent excesses in renal cell carcinoma
risk in the cohort and subcohort. A high percentage of smokers in the cohort would be needed to
account for the magnitude of renal cell carcinoma excess. Specifically, Raaschou-Nielsen et al.
(2003) noted "a high smoking rate would be expected to generate a much higher excess risk of
lung cancer than was observed in this study."
6/22/2009
395
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
The magnitude of confounding bias related to cigarette smoking in occupationally
employed populations to the observed lung, bladder and stomach cancer risk is minimal; less
than 20% for lung cancer and less than 10% for bladder and stomach cancers (Siemiatycki et al.,
1988; Blair et al., 2007). For lung cancer and metalwork specifically, smoking adjusted lung
cancer risks were approximately 10% lower after adjustment for smoking (Blair et al., 2007).
Thus, difference in cigarette smoking between exposed and referent subjects is not sufficient to
fully explain observed excess kidney cancer risks associated with TCE, particularly, high TCE
exposure. Information on possible confounding due to BMI (obesity) and to diabetes is lacking
in cohort studies; however, any uncertainties are likely small given the generally healthy nature
of an employed population and its favorable access to medical care.
Mineral oils such as cutting fluids or hydrazine common to some job titles with potential
TCE exposures (such as machinists, metal workers, and test stand mechanics) were included as
covariates in statistical analyses of Zhao et al. (2005), Boice et al. (2006) and Charbotel et al.
(2006, 2009). A TCE effect on kidney cancer incidence was still evident although effect
estimates were often imprecise due to lowered statistical power (Zhao et al., 2005; Charbotel et
al., 2006, 2009). Observed associations were similar in analyses including chemical co-
exposures in both Zhao et al. (2005) and Charbotel et al. (2006, 2009) compared to chemical co-
exposure unadjusted risks. The association or odds ratio (OR) between high TCE score and
kidney cancer incidence in Zhao et al. (2005) was 7.71 (95% CI: 0.65, 91.4) after adjustment for
other carcinogens including hydrazine and cutting oils, compared to analyses unadjusted for
chemical co-exposures (4.90, 95% CI: 1.23, 19.6).
In Charbotel et al. (2006), exposure to TCE was strongly associated with exposure to
cutting fluids and petroleum oils (22 of the 37 TCE-exposed cases were exposed to both).
Statistical modeling of all factors significant at 10% threshold showed the OR for cutting fluids
to be almost equal to 1, whereas the OR for the highest level of TCE exposure was close to two
(Charbotel et al., 2006). Moreover, when exposure to cutting oils was divided into three levels, a
decrease in OR with level of exposure was found. In conditional logistic regression adjusted for
cutting oil exposure, the relative risk (OR) was similar to relative risks from unadjusted for
cutting fluid exposures [high cumulative TCE exposure: 1.96 (95% CI: 0.71-5.37) compared to
2.16 (95% CI: 1.02-4.60); high cumulative and peak: 2.63 (95% CI: 0.79-8.83) compared to
2.73 (95% CI: 1.06-7.07) (Charbotel, 2006). Charbotel et al. (2009) further examined TCE
exposure defined as the highest TWA in any job held, inferior to cumulative exposure given its
lack of consideration of TCE exposure potential in other jobs, either as exposure to TCE alone,
cutting fluids alone, or to both after adjusting for smoking, body mass index, age, sex, and
exposure to other oils [TCE alone: 1.62 (95% CI: 0.75, 3.44); Cutting fluids alone: 2.39 (95%
CI: 0.52, 11.03); TCE >50 ppm TWA + cutting fluids: 2.70 (95% CI: 1.02, 7.17). There were
6/22/2009
396
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
few cases exposed to cutting fluids alone (n = 3) or to TCE alone (n = 15), all of whom had TCE
exposure (in the highest exposed job held) of <35 ppm TWA, and the subgroup analyses were of
limited statistical power. A finding of higher risk for both cutting oil and TCE exposure >50
ppm compared to cutting oil alone supports a TCE effect for kidney cancer. Adjustment for
cutting oil exposures, furthermore, did not greatly affect the magnitude of TCE effect measures
in the many analyses presented by Charbotel et al. (2006, 2009) suggesting cutting fluid
exposure as not greatly confounding TCE effect measures.
Boice et al. (2006) was unable to directly examine hydrazine exposure on TCE effect
measures because of a lack of model convergence in statistical analyses. Three of 7 TCE-
exposed kidney cancer cases were identified with hydrazine exposure of 1.5 years or less and the
absence of exposure to the other 4 cases suggested confounding related to hydrazine was
unlikely to greatly modify observed association between TCE and kidney cancer.
4.3.2.4 Susceptible Populations — Kidney Cancer and TCE Exposure
Two studies of kidney cancer cases from the Arnsberg region in Germany have examined
the influence of polymorphisms of the glutathione-S-transferase metabolic pathway on renal cell
carcinoma risk and TCE exposure (Briining et al., 1997b; Wiesenhiitter et al., 2007). In their
study of 45 TCE-exposed male and female renal cell carcinoma cases pending legal
compensation and 48 unmatched male TCE-exposed controls, Briining et al. (1997b) observed a
higher prevalence of exposed cases homozygous and heterozygous for GST-MI positive, 60%,
than the prevalence for this genotype among exposed controls, 35%. The frequency of GST-MI
positive was lower among this control series than the frequency found in other European
population studies, 50% (Briining et al., 1997b). The prevalence of the GST-T1 positive
genotype was 93% among exposed cases and 77% among exposed controls. The prevalence of
GST-T1 positive genotype in the European population is 75% (Briining et al., 1997b).
Wiesenhiitter et al. (2007) compares the frequency of genetic polymorphism among
subjects from the renal cancer case-control study of Briining et al. (2003) and to the frequencies
of genetic polymorphisms in the areas of Dormund and Lutherstadt Wittenberg, Germany.
Wiesenhiitter et al. (2007) identified the genetic frequencies of GST-MI and GST-T1
phenotypes for 98 of the original 134 cases (73%) and 324 of the 401 controls (81%). The
prevalence of GST-MI positive genotype was 48% among all renal cell carcinoma cases, 40%
among TCE-exposed cases, and 52% among all controls. The prevalence of GST-T1 positive
genotypes was 81% among all cases and 81% among all controls. The prevalence of GST-T1
positive genotypes reported in this paper for all TCE-exposed cases was 20%. The numbers of
exposed (n = 4) and unexposed (n = 15) GST-T1 positive cases does not sum to the 79 cases with
the GST-T1 positive genotype identified in the Table's first row; EPA staff has written Professor
6/22/2009
397
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Bolt requesting clarification of the data in Table 1 of Wiesenhiitter et al. (2007) (personal
communication from Cheryl Siegel Scott to Professor Herman Bolt, email dated August 05,
2008) [no reply received as of January, 2009 to request], Wiesenhiitter et al. (2007) noted
background frequencies in the German population in the expanded control group were 50% for
GST-MI positive and 81% for GST-T1 positive genotypes.
Observations in Briining et al. (1997b) and Wiesenhiitter et al. (2007) must be interpreted
cautiously. Few details are provided in these studies on selection criteria and not all subjects
from the Briining et al. (2003) case-control study are included. For GST-MI positive, the higher
prevalence among exposed cases in Briining et al. (1997b) compared Wiesenhiitter et al. (2007)
and the lower prevalence among controls compared to background frequency in the European
population may reflect possible selection biases. On the other hand, the broader base population
included in Briining et al. (2003) may explain the observed lower frequency of GST-MI positive
cases in Wiesenhiitter et al. (2007). Moreover, Wiesenhiitter et al. (2007) does not report
genotype frequencies for controls by exposure status and this information is essential to an
examination of whether renal cell carcinoma risk and TCE exposure may be modified by
polymorphism status.
Of the three larger (in terms of number of cases) studies that did provide results
separately by sex, Dosemeci et al. (1999) suggest that there may be a sex difference for TCE
exposure and renal cell carcinoma (OR = 1.04, [95% CI: 0.6, 1.7]) in males and 1.96 (95% CI:
1.0, 4.0 in females), while Raaschou-Nielsen (2003) report the same SIR (1.2) for both sexes and
crude ORs calculated from data from the Pesch et al. (2000a) study (provided in a personal
communication from Beate Pesch, Forschungsinstitut fiir Arbeitsmedizin (BGFA), to Cheryl
Scott, U.S. EPA, 21 February 2008) are 1.28 for males and 1.23 for females. Whether the
Dosemeci et al. (1999) observations are due to susceptibility differences or to exposure
differences between males and females cannot be evaluated. Blair et al. (1998) and Hansen et al.
(2001) also present some results by sex, but these two studies have too few cases to be
informative about a sex difference for kidney cancer.
4.3.2.5 Meta-analysis for Kidney Cancer
Meta-analysis (detailed methodology in Appendix C) was adopted as a tool for
examining the body of epidemiologic evidence on kidney cancer and TCE exposure and to
identify possible sources of heterogeneity. The meta-analyses of the overall effect of TCE
exposure on kidney cancer suggest a small, statistically significant increase in risk that was
stronger in a meta-analysis of the highest exposure group. There was no observable
heterogeneity across the studies for any of the meta-analyses and no indication of publication
6/22/2009
398
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
bias. Thus, these findings of increased risks of kidney cancer associated with TCE exposure are
robust.
The meta-analysis of kidney cancer examines 14 cohort and case-control studies
identified through a systematic review and evaluation of the epidemiologic literature on TCE
exposure (Siemiatycki et al., 1991; Parent et al., 2000; Axelson et al., 1994; Anttila et al., 1995;
Blair et al., 1998; Morgan et al., 1998; Boice et al., 1999; Dosemeci et al., 1999; Greenland et al.
1994; Pesch et al., 2000a; Hansen et al., 2001; Briining et al., 2003; Raaschou-Nielsen et al.,
2003; Zhao et al., 2005; Charbotel et al., 2006). Details of the systematic review and meta-
analysis of the TCE studies are fully discussed in Appendix B and C.
The pooled estimate from the primary random effects meta-analysis of the 14 studies was
1.26 (95% CI: 1.11, 1.42). The analysis was dominated by two (contributing almost 70% of the
weight) or three (almost 80% of the weight) large studies (Dosemeci et al., 1999; Pesch et al.,
2000a; Raaschou-Nielsen et al., 2003). Figure 4.3.1 arrays individual studies by their weight.
No single study was overly influential; removal of individual studies resulted in pooled RR
(RRp) estimates that were all statistically significant and that ranged from 1.22 (with the removal
of Briining et al. [2003]) to 1.28 (with the removal of Raaschou-Nielsen et al. [2003]). Similarly,
the overall RRp estimate was not highly sensitive to alternate RR estimate selections nor was
heterogeneity or publication bias apparent. Subgroup analyses were done examining the cohort
and case-control studies separately with the random effects model; the resulting RRp estimates
were 1.16 (95% CI 0.96, 1.41) for the cohort studies and 1.41 (1.08, 1.83) for the case-control
studies. There was heterogeneity in the case-control subgroup, but it was not statistically
significant (p = 0.17).
Nine studies reported risks for higher exposure groups (Siemiatycki et al., 1991; Parent et
al., 2000; Blair et al., 1998; Morgan et al., 1998; Boice et al., 1999; Dosemeci et al., 1999; Pesch
et al., 2000a; Briining et al., 2003; Raaschou-Nielsen et al., 2003; Zhao et al., 2005; Charbotel et
al., 2006). Different exposure metrics were used in the various studies, and the purpose of
combining results across the different highest exposure groups was not to estimate an RRp
associated with some level of exposure. Instead, the focus on the highest exposure category was
meant to result in an estimate less affected by exposure misclassification. In other words, it is
more likely to represent a greater differential TCE exposure compared to people in the referent
group than the exposure differential for the overall (typically any versus none) exposure
comparison. Thus, if TCE exposure increases the risk of kidney cancer, the effects should be
more apparent in the highest exposure groups.
The RRp estimate from the random effects meta-analysis of the studies with results
presented for higher exposure groups was 1.61 (95% CI 1.27, 2.03), higher than the RRp from
the overall kidney cancer meta-analysis. As with the overall analyses, the meta-analyses of the
6/22/2009
399
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
highest-exposure groups were dominated by Pesch et al. (2000a) and Raaschou-Nielsen et al.
(2003), which provided about 70% of the weight. Axelson et al. (1994), Anttila et al. (1995) and
Hansen et al. (2001) do not report risk ratios for kidney cancer by higher exposure and a
sensitivity analysis was carried out to address reporting bias. The RRp estimate from the
primary random effects meta-analysis with null RR estimates (i.e., RR = 1.0) included for
Axelson et al. (1994), Anttila et al. (1995) and Hansen et al. (2001) to address reporting bias
associated with ever exposed was 1.55 (95% CI: 1.24, 1.94). Figure 4.3.2 arrays individual
studies by their weight. The inclusion of these 3 additional studies contributed less than 8% of
the total weight. No single study was overly influential; removal of individual studies resulted in
RRp estimates that were all statistically significant and that ranged from 1.46 (with the removal
of Raaschou-Nielsen et al. [2003]) to 1.61 (with the removal of Pesch et al. [2000a]). Similarly,
the RRp estimate was not highly sensitive to alternate RR estimate selections and heterogeneity
observed across the studies for any of the meta-analyses conducted with the highest-exposure
groups.
NRC (2006) deliberations on trichloroethylene commented on two prominent evaluations
of the then-current TCE epidemiologic literature using meta-analysis techniques, Wartenberg et
al. (2000) and Kelsh et al. (2005), submitted by Exponent-Health Sciences to NRC during their
deliberations. Wartenberg et al. (2000) reported a RRp of 1.7 (95% CI: 1.1, 2.7) for kidney
cancer incidence in the TCE subcohorts (Axelson et al., 1994; Anttila et al., 1995; Blair et al.,
1998; Henschler et al., 1995). For kidney cancer mortality in TCE subcohorts (Henschler et al.,
1995; Blair et al., 1998; Boice et al., 1999; Morgan et al., 1998; Ritz 1999), Wartenberg et al.
(2000) reported a RRp of 1.2 (95% CI: 0.8, 1.7). Kelsh et al. (2005) examined a slightly
different grouping of cohort studies as did Wartenberg et al. (2000), presenting a pooled relative
risk estimate for kidney cancer incidence and mortality combined. The RRp for kidney cancer in
cohort studies (Axelson et al., 1994; Anttila et al., 1995; Blair et al., 1998; Morgan et al., 1998;
Boice et al., 1999; Hansen et al., 2001; Raaschou-Nielsen et al., 2003) was 1.29 (95% CI:
1.06-1.57) with no evidence of heterogeneity. Kelsh et al. (2005), also, presented separately a
pooled relative risk for renal cancer case-control studies and TCE. For case-control studies
(Siemiatycki et al., 1991; Greenland et al., 1994; Vamvakas et al., 1998; Dosemeci et al., 1999;
Pesch et al., 2000a; Briining et al., 2003), the RRp for renal cell carcinoma was 1.7 (95% CI: 1.0,
2.7) (interpolated from Figure 26 of NRC presentation) with evidence of heterogeneity, and RRp
of 1.2 (95%) CI: 0.9, 1.4) (interpolated from Figure 26 of NRC presentation) and no evidence of
heterogeneity in a sensitivity analysis removing Vamvakas et al. (1998) and Briining et al.
(2003), two studies Kelsh et al. (2005) considered as "outliers."
The present analysis was conducted according to NRC (2006) suggestions for
transparency, systematic review criteria, and examination of both cohort and case-control
6/22/2009
400
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
studies. The present analysis includes the recently published study of Charbotel et al. (2006) and
an analysis that examines both the TCE subcohort and case-control studies together. As
discussed above, the pooled estimate from the primary random effects meta-analysis of the 14
studies was 1.26 (95% CI: 1.11, 1.42). Additionally, EPA examined kidney cancer risk for
higher exposure group. The RRp estimate from the random effects meta-analysis of the studies
with results presented for higher exposure groups was 1.61 (95% CI 1.27, 2.03), higher than the
RRp from the overall kidney cancer meta-analysis, and 1.55 (95% CI: 1.24, 1.94) in the meta-
analysis with null RR estimates (i.e., RR = 1.0) to address possible reporting bias for three
studies.
6/22/2009
401
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.3.1. Summary of human kidney toxicity studies
Subjects
Effect
Exposure
Reference
206 subjects-
Increased (32-microglobulin
TCE exposure was through
Nagaya et al.,
104 male workers exposed to
and total protein in spot urine
degreasing activities in metal
1989
TCE; 102 male controls
specimen.
parts factory or
(source not identified)
semiconductor industry
(^-microglobulin:
Exposed, 129.0 + 113.3 mg/g
U-total trichlorocompounds:
creatinine (Cr)
Exposed, 83.4 mg/g Cr
Controls, 113.6 + 110.6 mg/g
(range, 2-66.2 mg/g Cr
Cr
Controls, N.D. 5
Total protein:
8.4 + 7.9 years mean
Exposed, 83.4+113.2 mg/g
employment duration
creatinine (Cr)
Controls, 54.0 + 18.6 mg/g Cr
29 metal workers
NAG in morning urine
Breathing zone monitoring, 3
Seldenetal., 1993
specimen, 0.17 + 0.11
ppm (median) and 5 ppm
U/mmol Cr
(mean)
191 subjects-
Increased urinary proteins
All exposed RCC cases
Briining et al.,
41 renal cell carcinoma cases
patterns, al-microglobulin,
exposed to 'high" and "very
1999a
pending cases involving
and total protein in spot urine
high" TCE intensity
compensation with TCE
specimen.
exposure;
18 year mean exposure
50 unexposed renal cell
Slight/severe tubular damage:
duration
carcinoma cases from same
TCE RCC cases, 93%
area as TCE-exposed cases;
Non-exposed RCC cases, 46%
100 non-diseased control and
Surgical controls, 11%
hospitalized surgical patients
al-microglobulin (mg/g
creatinine):
Exposed RCC cases, 24.6 +
[SD6] 13.9
Unexposed RCC cases, 11.3 +
[SD] 9.8
Surgical controls, 5.5 + [SD]
6.8
5 N.D. = not detectable
6 SD = Standard deviation
6/22/2009
402
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
85 male workers employed in
cardboard manufacturing
factory (39 TCE exposed, 46)
non-exposed office and
administrative controls)
Increased urinary protein
patterns and excretion of
proteins in spot urine
specimen.
Slight/severe tubular damage:
TCE exposed, 67%
Non-exposed, RCC cases, 9%
p< 0.001
al-microglobulin (mg/g
creatinine):
Exposed, 16.2 + [SD] 10.3
Unexposed, 7.8 + [SD] 6.9
p< 0.001
GSTa (|ig/g creatinine):
Exposed 6.0 + [SD] 3.3
Unexposed, 2.0 + [SD] 0.57
p< 0.001
No group differences in total
protein or GSTpi
"High" TCE exposure to
workers in the fitters shop
and electrical department
"Very high" TCE exposure
to workers through general
degreasing operations in
carton machinery section
Briining et al.,
1999b
99 renal cell carcinoma cases
and 298 hospital controls
(from Briining et al. [2003]
and alive at the time of
interview)
Increased excretion of al-
microglobulin in spot urine
specimen.
Proportion of subjects with
al-microglobulin <5.0
mg/L:
Exposed cases, 51%
Unexposed cases, 15%
Exposed controls, 55%
Unexposed controls, 55%
All exposed RCC cases
exposed to 'high" and "very
high" TCE intensity
Bolt et al., 2004
124 subjects (70 workers
Analysis of urinary proteins in
Mean U-TCA of exposed
Green et al., 2004
7 For a urine sample, 10-17 mg of albumin per g of creatinine is considered to be suspected albuminuria in males
(15-25 in females) (De Jong and Brenner 2004).
6/22/2009
403
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
currently exposed to TCE and
spot urine sample obtained 4
workers was 64 + [SD] 102
54 hospital and administrative
days after exposure.
(Range, 1-505)
staff controls)
Mean U-TCOH of exposed
Increased excretion of
workers was 122 + [SD] 119
albumin, NAG, and formate in
(Range, 1-639)
spot urine specimen.
Albumin (mg/g creatinine)7:
Mean TCE concentration to
Exposed, 9.71 + [SD] 11.6
exposed subjects was
Unexposed, 5.50 + [SD] 4.27
estimated as 32 ppm (range,
p < 0.05
0.5-252 ppm) and was
estimated by applying the
Total NAG (U/g creatinine):
German occupational
Exposed, 5.27+[SD] 3.78
exposure limit (maximale
Unexposed, 2.41 + [SD] 1.91
arbeitsplatz konzentration,
p<0.01
MAK) standard to U-TCA
and assuming that the linear
Format (mg/g creatinine):
relationship holds for
Exposed, 9.45 + [SD] 4.78
exposures above 100 ppm.
Unexposed, 5.55 + [SD] 3.00
p<0.01
86% of subjects with
exposure to <50 ppm TCE
No group mean differences in
GSTa, retinol binding protein,
al-microglobulin, (32-
microglobulin, total protein,
and methylmalonic acid.
101 cases or deaths from end-
TCE exposure:
Cumulative TCE exposure
Radican et al.,
stage renal disease (ESDR)
Cox Proportional Hazard
(intensity x duration)
2006
among male and female
Analysis:
identified using 3 categories,
subjects in Hill Air Force
Ever exposed to TCE8,
<5 unit-year, 5-25 unit year,
Base aircraft maintenance
1.86 (1.02,3.39)
>25 unit-year per job
worker cohort of Blair et al.
exposure matrix of Stewart et
(1998)
Logistic regression:5
al. (1991)
No chemical exposure
(referent group): 1.0
<5 unit-year, 1.73 0.86, 3.48)
8 Hazard ratio and 95% confidence interval
6/22/2009
404
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
5-25 unit-year, 1.65 (0.82,
3.35)
>25 unit-year, 1.65 (0.82,
3.35)
Monotonic trend test, p > 0.05
Indirect low-intermittent TCE
exposure, 2.47 (1.17, 5.19)
Indirect peak/infrequent TCE
exposure 3.55 (1.25, 10.74)
Direct TCE exposure, "not
statistically significant" but
hazard ratio and confidence
intervals were not presented in
paper
6/22/2009
405
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.3.2. Summary of human studies on somatic mutations of the VHL gene1
Briining et al., 1997a
Brauch et al., 1999
Schraml et al., 1999
Brauch etal., 2004
Charbotel et al., 2007
TCE exposure
status
Exposed
Exposed
Unexposed
Exposed
Unexposed
Exposed
Unexposed
Exposed
Unexposed
Number of
subjects/
Number with
mutations (%)
23/23 (100%)
44/33
(75%)
73/42
(58%)
9/3 (33%)
113/38
(34%)
17/14
(82%)
21/2 (10%)
25/2
(9%)
23/2 (8%)
Renal Cell
Carcinoma
subtype
Unknown
Unknown
Clear cell 9
(75%)
Papillary 2
(18%)
Oncocytomas
1 (8%)
Unknown
Clear cell 37 (%)
Oncocytic adenoma 1
(%) bilateral
metachronous 1 (%)
Clear cell 51 (75%)
Papillary 10 (10-15%)
Chromophobe 4 (5%)
Oncocytomas 4 (5%)
Tissue Type
Analyzed
paraffin
Paraffin, fresh
(lymphocyte)
paraffin
paraffin
Paraffin, frozen tissues,
Bouin's fixative
Assay
2 2
SSCP , sequencing
SSCP, sequencing,
restriction enzyme
digestion
CGH, sequencing
Sequencing
Sequencing
Number of
mutations
23
50
42
4
50
24
2
2
2
Type of
mutation
Missense
1
27
NA
1
Unknown
17
2
1
1
Nonmissense3
3
23
NA
3
Unknown
7
0
1
1
6/22/2009
406
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Adapted from NRC (2006) with addition of Schraml et al. (1999) and Charbotel et al. (2007).
2 By single stand conformation polymorphism (SSCP). Four (4) sequences confirmed by comparative genomic hybridization.
3 Includes insertions, frameshifts and deletions
6/22/2009
407
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.3.3. Summary of human studies on TCE exposure and kidney cancer
Exposure Group
Cohort Studies - Incidence
Aerospace workers (Rocketdyne)
Any exposure to TCE
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
Relative Risk
(95% CI)
Not reported
1.001
1.87 (0.56, 6.20)
4.90 (1.23, 19.6)
p = 0.023
No. obs.
events
Reference
Zhao et al., 2005
TCE, 20 years exposure lag
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
1.001 6
1.19 (0.22,6.40) 7
7.40(0.47,116) 3
p = 0.120
All employees at electronics factory (Taiwan)
Males
Females
Females
Danish blue-collar worker w/TCE exposure
Any exposure, all subjects
Any exposure, males
Any exposure, females
Exposure Lag Time
1.06 (0.45, 2.08);
1.09 (0.56, 1.91)-
1.10 (0.62, 1.82):
1.2 (0.98, 1.46)
1.2 (0.97, 1.48)
1.2 (0.55,2.11)
20 years 1.3 (0.86,1.88)
Employment duration
<1 year 0.8(0.5,1.4)
1-4.9 years 1.2(0.8,1.7)
>5 years 1.6(1.1,2.3)
Subcohort w/higher exposure
Any TCE exposure 1.4 (1.0, 1.8)
Employment duration
1-4.9 years 1.1 (0.7, 1.7)'
>5 years 1.7 (1.1, 2.4)"
12
15
103
93
10
28
16
28
32
53
23
30
Chang et al., 2005
Sung et al., 2008
Raaschou-Nielsen et al..
2003
Biologically monitored Danish workers
Any TCE exposure, males
Any TCE exposure, females
1.1 (0.3,2.8)
0.9 (0.2, 2.6)
2.4 (0.03, 14)
Hansen etal., 2001
6/22/2009
408
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Cumulative exp (Ikeda)
Mean concentration (Ikeda)
Employment duration
<17 ppm-yr
>17 ppm-yr
<4 ppm
4+ ppm
< 6.25 yr
>6.25
Not reported
Not reported
Not reported
Aircraft maintenance workers from Hill Air Force Base
TCE subcohort Not reported
Males, Cumulative exp
Blair etal., 1998
0
1.01
< 5 ppm-yr
1.4 (0.4,4.7)
9
5-25 ppm-yr
1.3 (0.3, 4.7)
5
>25 ppm-yr
0.4 (0.1,2.3
2
Females, Cumulative exp
0
1.01
< 5 ppm-yr
0
5-25 ppm-yr
0
>25 ppm-yr
3.6 (0.5,25.6)
2
Biologically-monitored Finnish workers
Anttila et al., 1995
All subjects
0.87 (0.32, 1.89)
6
Mean air-TCE (Ikeda extrapolation)
<6 ppm
Not reported
6+ ppm
Not reported
Cardboard manufacturing workers in Arnsberg, Germany
Henschleretal., 1995
Exposed workers
7.97 (2.59, 8.59)5
5
Biologically-monitored Swedish workers
Axelson et al., 1994
Any TCE exposure, males
1.16 (0.42,2.52)
6
Any TCE exposure, females
Not reported
Cardboard manufacturing workers, Atlanta area, GA
Sinks et al., 1992
All subjects
3.7(1.4,8.1)
6
All departments
co (3.0, go)6
5
Finishing department
16.6 (1.7, 453.1)6
3
Cohort Studies-Mortality
Computer manufacturing workers (IBM), NY
6/22/2009
409
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Aerospace workers (Rocketdyne)
Any TCE (utility/eng flush)
Any exposure to TCE
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
Males 1.64 (0.45, 4.21)
Females
2.22 (0.89, 4.57)
Not reported
1.001
1.43 (0.49,4.16)
2.13 (0.50, 8.32)
p = 0.31
4
0
7
Clapp and Hoffman,
2008
Boice et al., 2006
Zhao et al., 2005
TCE, 20 years exposure lag
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
1.001 10
1.69 (0.29, 9.70) 6
1.82 (0.09, 38.6) 1
p = 0.635
View-Master employees
Males
Females
US Uranium-processing workers (Fernald)
Any TCE exposure
ATSDR, 2004
2.76 (0.34, 9.96)
6.21 (2.68, 12.23)7
Not reported
Ritz, 1999
Mod TCE exposure, >2 years duration4
Aerospace workers (Lockheed)
Routine Exp
0.99 (0.40, 2.04)
7
Routine-Intermittent1
Not presented
11
Duration of exposure
0 years
1.0
22
< 1 year
0.97 (0.37, 2.50)
6
1-4 years
0.19 (0.02, 1.42)
1
> 5 years
0.69 (0.22,2.12)
4
Boice et al., 1999
p for trend
Aerospace workers (Hughes) Morgan etal., 1998
TCE Subcohort 1.32 (0.57,2.60) 8
Low Intensity (<50 ppm)5 0.47 (0.01,2.62) 1
High Intensity (>50 ppm)5 1.78 (0.72, 3.66) 7
6/22/2009 410
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE Subcohort (Cox Analysis)
Never exposed 1.001 24
Everexposed 1.14 (0.51, 2.58)8 8
Peak
No/Low 1.001 24
Med/Hi 1.89 (0.85, 4.23)8 8
Cumulative
Referent 1.001 24
Low 0.31 (0.04, 2.36)8 1
High 1.59 (0.68, 3.71)8 7
Aircraft maintenance workers (Hill AFB, Utah)
Blair etal., 1998
TCE subcohort
1.6 (0.5,5.1)'
15
Males, Cumulative exp
0
1.01
< 5 ppm-yr
2.0 (0.5, 7.6)
8
5-25 ppm-yr
0.4 (0.1,4.0)
1
>25 ppm-yr
1.2 (0.3,4.8)
4
Females, Cumulative exp
0
1.01
< 5 ppm-yr
0
5-25 ppm-yr
9.8 (0.6, 157)
1
>25 ppm-yr
3.5 (0.2, 56.4)
1
TCE subcohort
1.18 (0.47, 2.94)9
18
Males, Cumulative exp
1.24 (0.41, 3.71)9
16
0
1.01
< 5 ppm-yr
1.87 (0.59, 5.97 9
10
5-25 ppm-yr
0.31 (0.03, 2.75)9
1
>25 ppm-yr
1.16 (0.31, 4.32)9
5
Females, Cumulative exp
0.93 (0.15, 5.76)9
2
0
1.01
< 5 ppm-yr
0
5-25 ppm-yr
2.86 (0.27, 29.85)9
1
>25 ppm-yr
0.97 (0.10, 9.50)9
1
Cardboard manufacturing workers in Arnsberg, Germany
TCE exposed workers
3.28 (0.40, 11.84)
2
Unexposed workers
(0.00, 5.00)
0
Deaths reported to among GE pension fund (Pittsfield, MA)
0.99 (0.30, 3.32)6
12
Radican et al„ 2008
Henschleretal., 1995
Cardboard manufacturing workers, Atlanta area, GA
Greenland et al., 1994
Sinks et al., 1992
6/22/2009
411
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1.4 (0.0, 7.7)
Aircraft manufacturing plant employees (Italy)
All subjects
Not reported
Aircraft manufacturing plant employees (San Diego, CA)
All subjects
0.93 (0.48, 1.64)
12
Case-control Studies
Population of Arve Valley, France
Any TCE exposure
1.64 (0.95, 2.84)
37
Cumulative TCE exposure
Referent/non-exposed
1.001
49
Low
1.62 (0.75, 3.47)
12
Medium
1.15 (0.47,2.77)
9
High
2.16(1.02, 4.60)10
16
Test for trend
p = 0.04
Cumulative TCE exposure + peak
Referent/non-exposed
1.001
49
Low/med, no peaks
1.35 (0.69, 2.63)
18
Low/med + peaks
1.61 (0.36. 7.30)
3
High, no peaks
1.76 (0.65, 4.73)
8
High + peaks
2.73 (1.06, 7.07)10
8
Cumulative TCE exposure, 10 year lag
Referent/non-exposed
1.001
49
Low/med, no peaks
1.44 (0.69, 2.80)
19
Low/med + peaks
1.38 (0.32, 6.02)
3
High, no peaks
1.50 (0.53,4.21)
7
High + peaks
3.15 (1.19, 8.38)
8
Time-weighted-average TCE exposure11
Referent/non-exposed
1.001
46
Any TCE without cutting fluid
1.62 (0.76, 3.44)
15
Any cutting fluid without TCE
2.39 (0.52, 11.03)
3
<50 ppm TCE + cutting fluid
1.14(0.49, 2,66)
12
50+ ppm TCE + cutting fluid
2.70(1.02,7.17)
10
Population of Arnsberg Region, Germany
Longest job held-TCE/PERC (CAREX)
1.80 (1.01,3.20)
117
Self-assessed exposure to TCE
2.47(1.36,4.49)
25
Duration of self-assessed TCE exposure
Costa et al., 1989
Garabrant et al., 1988
Charbotel et al., 2005,
2006, 2009
Briining et al., 2003
6/22/2009
412
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
0 1.001 109
<10 yrs 3.78 (1.54,9.28) 11
10-20 years 1.80 (0.67,4.79) 7
>20 years 2.69 (0.84,8.66) 8
Population in 5 German Regions
Any TCE Exposure
Pesch et al., 2000a
Not reported
Males Not reported
Females Not reported
TCE exposure (Job Task Exposure Matrix)
Males
Medium 1.3 (1.0, 1.8)
High 1.1 (0.8, 1.5)
Substantial 1.3(0.8,2.1)
Females
Medium 1.3 (0.7, 2.6)
High 0.8 (0.4, 1.9)
Substantial 1.8 (0.6, 5.0)
Population of Minnesota
Ever exposed to TCE, NCI JEM
Males 1.04 (0.6, 1.7)
Females 1.96 (1.0, 4.0)
Males + Females 1.30 (0.9, 1.9)
68
59
22
11
7
5
33
22
55
Dosemeci et al., 1999
Population of Arnsberg Region, Germany
Self-assessed exposure to TCE
10.80 (3.36, 34.75) 19
Vamvakas et al., 1998
Population of Montreal, Canada
Any TCE exposure
Substantial TCE exposure
0.8 (0.4, 2.0)
0.8 (0.2, 2.6)1
Siemiatycki et al., 1991
Geographic Based Studies
Residents in two study areas in Endicott, NY
1.90 (1.06,3.13)
15
ATSDR, 2006, 2008
Residents of 13 census tracts inRedlands, CA
0.80 (0.54, 1.12) 54
Morgan and Cassidy,
2002
Finnish residents
Residents of Hausjarvi
Residents of Huttula
Not reported
Not reported
Vartiainenetal., 1993
1 Internal referents, workers not exposed to TCE
6/22/2009
413
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
2 Relative risks for TCE exposure after adjustment for 1st employment, socioeconomic status, age at event,
and all other carcinogens, including hydrazine
3 Chang et al. (2005) - urinary organs combined
4 SIR for renal cell carcinoma
5 Henschler et al. (1995) Expected number of incident cases calculated using incidence rates from the
Danish Cancer Registry
6 Odds ratio from nested case-control analysis
7 Proportional mortality ratio
8 Risk ratio from Cox Proportional Hazard Analysis, stratified by age, sex and decade (Environmental
Health Strategies, 1997)
9 In Radican et al. (2008) estimated relative risks from Cox proportional hazard models were adjusted for
age and sex.
10 Analyses adjusted for age, sex, smoking and body mass index. The odds ratio, adjusted for age, sex,
smoking, body mass index and exposure to cutting fluids and other petroleum oils, for high cumulative
TCE exposure was 1.96 (95% CI: 0.71, 5.37) and for high cumulative + peak TCE exposure was 2.63 (95%
CI: 0.79, 8.83).
11 The exposure surrogate is calculated for one occupational period only and is not the average exposure
concentration over the entire employment period.
12 90% Confidence Interval
13 99% Confidence Interval
6/22/2009
414
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.3.4. Summary of case-control studies on kidney cancer and occupation or
job title
Case Ascertainment Area/Exposure Group
Swedish Cancer Registry Cases
Machine/electronics industry
Shop and construction metal work
Machine assembly
Metal plating work
Shop and construction metal work
Relative Risk
(95% CI)
1.30 (1.08, 1.55)4 [M]
1.75 (1.04, 2.76)4 [F]
1.19 (1.00, 1.40)4 [M]
1.62 (0.94, 2.59)4 [M]
2.70 (0.73, 6.92)4 [M]
1.66 (0.71, 3.26)4 [F]
No. exposed
cases
120
18
143
Reference
Wilson et al„ 2008
Arve Valley, France
Charbotel et al„ 2006
Metal industry
1.02 (0.59, 1.76)
28
Metal workers, job title
1.00 (0.56, 1.77)
25
Metal industry, screw-cutting workshops
1.39 (0.75, 2.58)
22
Machinery, electrical and transportation
equipment manufacture
1.19(0.61,2.33)
15
Iowa Cancer Registry Cases
Assemblers
2.5 (0.8, 7.6)
5
>10 years employment
4.2(1.2, 15.3)
4
Arnsberg Region, Germany
Iron/steel
1.15 (0.29,4.54)
3
Occupations with contact to metals
1.53 (0.97, 2.43)
46
Longest job held
1.14(0.66, 1.96)
24
Metal greasing/degreasing
5.57 (2.33, 13.32)
15
Degreasing agents
Low exposure
2.11 (0.86, 5.18)
9
High exposure
1.01 (0.40,2.54)
7
Zhang et al., 2004
Briining et al., 2003
Bologna, Italy
Metal workers
Printers
Solvents
2.21 (0.99, 5.37) 37
1.55 (0.17,13.46) 7
0.79 (0.31, 1.98)[M] 17
1.47(0.12, 17.46) [F] 3
Mattioli et al., 2002
Montreal, Canada
Metal fabricating and machining industry
1.0 (0.6, 1.8)
14
Parent et al., 2000
6/22/2009
415
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Metal processors
1.2 (0.4, 3.4)
4
Printing and publishing industry
1.1 (0.4, 3.0)
4
Printers
3.0(1.2,7.5)
6
Aircraft mechanics
2.8(1.0, 8.4)
4
5 Regions in Germany
Electrical and electronic equipment assembler
3.2 (1.0, 10.3) [M]
5
2.7(1.3, 5.8) [F}
11
Printers
3.5 (1.1, 11.2)[M]
5
2.1 (0.4, 11.7) [F]
2
Metal cleaning/degreasing, job task
1.3 (0.7, 2.3) [M]
15
1.5 (0.3, 7.7) [F]
2
New Zealand Cancer Registry
Toolmakers and blacksmiths
1.48 (0.72, 3.03)
No
Printers
0.67 (0.25, 1.83)
Minnesota Cancer Surveillance System
Iron or steel
1.6(1.2,2.2)
8
Rhein-Neckar-Odenwald Area, Germany
Metal
Industry
1.63 (1.07, 2.48)
71
Occupation
1.38 (0.89,2.12)
Electronic
Industry
0.51 (0.26, 1.01)
14
Occupation
0.57 (0.25, 1.33)
9
Chlorinated solvents
2.52 (1.23,5.16)
27
Metal and metal compounds
1.47 (0.94, 2.30)
62
Danish Cancer Registry
Iron and steel
1.4 (0.8, 2.4) [M]
31
1.0(0.1, 3.2) [F]
1
Solvents
1.5 (0.9, 2.4)[M]
50
6.4 (1.8, 23) [F]
16
France
Machine fitters, assemblers, and precision
instrument makers
0.7 (0.3, 1.9)
16
New South Wales, Australia
Iron and steel
1.18 (0.75, 1.85)1
52
Printing or graphics
2.39 (1.26, 4.52)2 19
1.18(0.87, 2.08)1 29
Pesch et al., 2000a
Delahunt et al., 1995
Mandel et al., 1995
Schlehofer et al., 1995
Mellemgaard et al., 1994
Auperin et al., 1994
McCredie and Stewart, 1993
6/22/2009
416
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Machinist or tool maker
Solvents
0.82 (0.32, 2.11)2 6
1.15 (0.72, 1.86)1 48
1.83 (0.92, 3.61)2 16
1.54 (1.11, 2.14)1 109
1.40 (0.82, 2.40)2 24
Finnish Cancer Registry
Iron and metalware work
1.87 (0.94, 3.76)
22
Machinists
2.33 (0.83,6.51)
10
Paper and pulp; printing/publishing
2.20(1.02,4.72) [M]
18
5.95 (1.21,29.2) [F]
7
Nonchlorinated solvents
3.46 (0.91, 13.2) [M]
9
Partenen et al., 1991
West Midlands UK Cancer Registry
Organic solvents
Everexposed 1.30 (0.31,8.50)
Intermediate exposure 1.54 (0.69, 4.10)
Harrington et al., 1989
Montreal, Canada
Organic solvents
1.68 (0.83,2.22) 33
Degreasing solvents 3.42(0.92,12.66) 10
Sharpe et al., 1989
Oklahoma
Metal degreasing
Machining
Painter, paint manufacture
1.7 (0.7, 3.8) [M] 19
1.7 (0.7, 4.3) [M] 13
1.3 (0.7, 2.6) [M] 22
Asal et al., 1988
Missouri Cancer Registry
Machinists
2.2 (0.5, 10.3)
Brownson, 1988
Danish Cancer Registry
Iron and metal, blacksmith
Painter, paint manufacture
1.4 (0.7, 2.9);
1.8 (0.7, 4.6)
17
10
Jensen etal., 1988
Renal cell carcinoma, McCredie and Stewart (1993)
Renal pelvis, McCredie and Stewart (1993)
Renal pelvis and ureter, Jensen et al. (1988)
Renal pelvis, Wilson et al. (2008)
6/22/2009
417
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE and kidney cancer
Study name
Statistics for each study
Risk ratio and 95% CI
Risk
Lower
Upper
ratio
limit
limit
p-Value
Anttila 1995
0.870
0.391
1.937
0.733
Axelson 1994
1.160
0.521
2.582
0.716
Blair 1998
1.600
0.501
5.110
0.428
Boice 1999
0.990
0.472
2.077
0.979
Greenland 1994
0.990
0.298
3.293
0.987
Hansen 2001
1.100
0.413
2.931
0.849
Morgan 1998 unpub RR
1.143
0.507
2.576
0.747
Raaschou-Nielsen 2003 RCC
1.200
0.950
1.516
0.126
Zhao 2005 moil 20 y lag
1.720
0.377
7.853
0.484
bruning 2003
2.470
1.359
4.488
0.003
charbotel 2007- high conf re:exp
1.880
0.889
3.976
0.099
dosemeci 1999
1.300
0.895
1.889
0.169
pesch 2000 JTEM
1.240
1.030
1.492
0.023
siemiatycki 1991
0.800
0.287
2.233
0.670
1.255
1.114
1.415
0.000
0.1 0.2
random effects model; same for fixed
Figure 4.3.1. Meta-analysis of kidney cancer and overall TCE exposure (The pooled estimate is in the bottom row. Symbol sizes reflect
relative weights of the studies. The horizontal midpoint of the bottom diamond represents the pooled RR estimate and the horizontal
extremes depict the 95% CI limits.)
6/22/2009 418
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE and kidney cancer - highest exposure groups
Study name
Statistics for each study
Risk ratio and 95% CI
Risk
Lower
Upper
ratio
limit
limit
p-Value
Blair 1998 mort
1.500
0.420
5.356
0.532
Bo ice 1999
0.690
0.222
2.142
0.521
Morgan 1998
1.590
0.681
3.714
0.284
Raaschou-Nielsen 2003
1.700
1.189
2.431
0.004
Zhao 2005 inc 20y lag
7.400
0.471
116.249
0.154
bruning 2003
2.690
0.838
8.634
0.096
charbotel 2007 good conf re:exp
3.340
1.273
8.761
0.014
pesch 2000 - JTEM
1.400
0.911
2.151
0.124
siemiatycki 1991
0.800
0.189
3.385
0.762
antilla
1.000
0.250
3.998
1.000
axelson
1.000
0.141
7.099
1.000
hansen
1.000
0.323
3.098
1.000
1.549
1.239
1.937
0.000
0.1
0.2
0.5
10
random effects model; fixed effect same
Figure 4.3.2. Meta-analysis of kidney cancer and TCE exposure - highest exposure groups.
With assumed null RR estimates for Antilla, Axelson, and Hansen (see Appendix C text).
6/22/2009
419
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.3.3 Human studies of Somatic Mutation of von Hippel-Lindau ( VHL) Gene
Studies have been conducted to identify mutations in the VHL gene in renal cell
carcinoma patients, with and without TCE exposures (Charbotel et al., 2007; Schraml et al.,
1999; Brauch et al., 1999, 2004; Toma et al., 2008; Furge et al., 2007; Kenck et al., 1996).
Inactivation of the VHL gene through mutations, loss of heterozygosity and imprinting has been
observed in about 70% of sporadic renal clear cell carcinomas, the most common renal cell
carcinoma subtype (Kenck et al., 1996). Other genes or pathways, including c-myc activation
and vascular endothelial growth factor (VEGF), have also been examined as to their role in
various renal cell carcinoma subtypes (Furge et al., 2007; Toma et al., 2008). Furge et al. (2007)
reported that there are molecularly distinct forms of renal cell carcinoma (RCC) and possibly
molecular differences between clear-cell renal cell carcinoma subtypes. This study was
performed using tissues obtained from paraffin blocks. These results are supported by a more
recent study which examined the genetic abnormalities of clear cell renal cell carcinoma using
frozen tissues from 22 cc-RCC patients and paired normal tissues (Toma et al., 2008). This
study found that 20 (91%) of the 22 cases had loss of heterozygosity (LOH) on chromosome 3p
(harboring the VHL gene). Alterations in copy number were also found on chromosome 9 (32%
of cases), chromosome arm 14q (36% of cases), chromosome arm 5q (45% of cases) and
chromosome 7 (32% of cases), suggesting roles for multiple genetic changes in RCC, and is also
supported by genomes-wide SNP (single-nucleotide polymorphism) analysis (Toma et al., 2008).
Several papers link mutation of the VHL gene in renal cell carcinoma patients to TCE
exposure. These reports are based on comparisons of VHL mutation frequencies in TCE exposed
cases from renal cell carcinoma case-control studies or from comparison to background mutation
rates among renal cell carcinoma case series (Table 4.3.2). Briining et al. (1997a) first reported a
high somatic mutation frequency (100%) in a series of 23 renal cell carcinomas cases with
medium to high intensity TCE exposure as determined by an abnormal SSCP pattern, with most
variations found in exon two. Only four samples were sequenced at the time of publication and
showed mutations in exon one, two and three (Table 4.3.2). Some of the cases in this study were
from the case-control study of Vamvakas et al. (1998) (see Section 4.3.3. and Appendix [meta-
analysis]).
Brauch et al. (1999, 2004) analyzed renal cancer cell tissues for mutations of the VHL
gene and reported increased occurrence of mutations in patients exposed to high concentrations
of TCE. In the first study (Brauch et al., 1999), an employer's liability or worker's
compensation registry was used to identify 44 renal cell carcinoma cases, 18 of whom were also
included in Briining et al. (1997a). Brauch et al. (1999) found multiple mutations in 42% of the
6/22/2009
420
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
exposed patients who experienced any mutation and 57% showed loss of heterozygosity. A hot
spot mutation of cytosine to thymine at nucleotide 454 (C454T) was found in 39% of samples
that had a VHL mutation and was not found in renal cell cancers from nonexposed patients or in
lymphocyte DNA from either exposed or nonexposed cases or controls. As discussed above,
little information was given on how subjects were selected and whether there was blinding of
exposure status during the DNA analysis. In the second study, Brauch et al. (2004) investigated
21 of the 39 renal cell carcinoma patients identified as non-TCE exposed from Vamvakas et al.
(1998) for which tissue specimens were available. The earlier studies of Briining et al. (1997a)
or Brauch et al. (1999) included VHL sequencing of tissue specimens from TCE-exposed cases
from the renal cell carcinoma case-control study of Vamvakas et al. (1998). Brauch et al. (2004)
compared age at diagnosis and histopathologic parameters of tumors as well as somatic mutation
characteristics in the VHL tumor suppressor gene between the TCE-exposed and non-TCE
exposed renal cell carcinoma patient groups (TCE-exposed from their previous 1999 publication
to the non TCE-exposed cases newly sequenced in this study). Renal cell carcinoma did not
differ with respect to histopathologic characteristics in either patient group. Comparing results
from TCE-exposed and nonexposed patients revealed clear differences with respect to (1)
frequency of somatic VHL mutations, (2) incidence of C454T transition, and (3) incidence of
multiple mutations. The C454T hot spot mutation at codon 81 was exclusively detected in
tumors from TCE-exposed patients, as were multiple mutations. Also, the incidence of VHL
mutations in the TCE-exposed group was at least 2-fold higher than in the nonexposed group.
Overall, these finding support the view that the effect of TCE is not limited to clonal expansion
of cells mutated spontaneously or by some other agent.
Brauch et al. (2004) were not able to analyze all RCCs from the Vamvakas study
(Vamvakas et al., 1998), in part because samples were no longer available. Using the data
described by Brauch et al. (2004) (VHL mutation found in 15 exposed and 2 nonexposed
individuals, and VHL mutation not found in 2 exposed and 19 unexposed individuals), the
calculated OR is 71.3. The lower bound of the OR including the excluded RCCs is derived from
the assumption that all 20 cases that were excluded were exposed but did not have mutations in
VHL (VHL mutations were found in 15 exposed and 2 unexposed individuals and VHL was not
found in 22 exposed and 18 unexposed individuals), leading to an OR of 6.5 that remains
statistically significant.
Charbotel et al. (2007) examines somatic mutations in the three VHL coding exons in
RCC cases from their case-control study (Charbotel et al., 2006). Of the 87 RCCs in the case-
control study, tissue specimens were available for 69 cases (19%) of which 48 were cc-RCC.
VHL sequencing was carried out for only the cc-RCC cases, 66% of the 73 cc-RCC cases in
Charbotel et al. (2006). Of the 48 cc-RCC cases available for VHL sequencing, 15 subjects were
6/22/2009
421
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
identified with TCE exposure (31%), an exposure prevalence lower than 43% observed in the
case-control study. Partial to full sequencing of the VHL gene was carried out using PCR
amplification and VHL mutation pattern recognition software of Beroud et al. (1998). Full
sequencing of the VHL gene was possible for only 26 RCC cases (36% of all RCC cases). Single
point mutations were identified in 4 cases (8% prevalence): 2 unexposed cases, a G>C mutation
in exon 2 splice site and a G>A in exon 1; one case identified with low/medium exposure, T>C
mutation in exon 2, and, one case identified with high TCE exposure, T>C in exon 3. It should
be noted that the two cases with T>C mutations were smokers unlike the cases with G>A or G>C
mutations. The prevalence of somatic VHL mutation in this study is quite low compared to that
observed in other RCC case series from this region; around 50% (Bailly et al., 1995; Gallou et
al., 2001). To address possible bias from misclassification of TCE exposure, Charbotel et al.
(2006) examined renal cancer risk for jobs associated with a high level of confidence for TCE
exposure. As would be expected if bias was a result of misclassification, they observed a
stronger association between higher confidence TCE exposure and RCC, suggesting that some
degree of misclassification bias is associated with their broader exposure assessment approach.
Charbotel et al. (2007) do not present findings on VHL mutations for those subjects with higher
level of confidence TCE exposure assignment.
Schraml et al. (1999) did not observe statistically significant differences in DNA
sequence or mutation type in a series of 12 renal cell carcinomas from subjects exposed to
solvents including varying TCE intensity and a parallel series of 113 clear cell carcinomas from
non-TCE exposed patients. Only 9 of the RCC were cc-RCC and were sequenced for mutations.
VHL mutations were observed in clear cell tumors only; four mutations in three TCE-exposed
subjects compared to 50 mutations in tumors of 38 non-exposed cases. Details as to exposure
conditions are limited to a statement that subjects had been exposed to high doses of solvents,
potential for mixed solvent exposures, and that exposure included a range of TCE
concentrations. Limitations of this study include having a wider range of TCE exposure
intensities as compared to the studies described above (Briining et al., 1997a; Brauch et al., 1999,
2004), which focused on patients exposed to higher levels of TCE, and the limited number of
TCE-exposed subjects analyzed, being the smallest of all available studies on RCC, TCE and
VHL mutation. For these reasons, Schraml et al. (1999) is quite limited for examining the
question of VHL mutations and TCE exposure.
A number of additional methodological issues need to be considered in interpreting these
studies. Isolation of DNA for mutation detection has been performed using various tissue
preparations, including frozen tissues, formalin fixed tissues and tissue sections fixed in Bouin's
solution. Ideally, studies would be performed using fresh or freshly frozen tissue samples to
limit technical issues with the DNA extraction. When derived from other sources, the quality
6/22/2009
422
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
and quantity of the DNA isolated can vary, as the formic acid contained in the formalin solution,
fixation time and period of storage of the tissue blocks often affect the quality of DNA. Picric
acid contained in Bouin's solution is also known to degrade nucleic acids resulting in either low
yield or poor quality of DNA. In addition, during collection of tumor tissues, contamination of
neighboring normal tissue can easily occur if proper care is not exercised. This could lead to the
'dilution effect' of the results—i.e., because of the presence of some normal tissue, frequency of
mutations detected in the tumor tissue can be lower than expected. These technical difficulties
are discussed in these papers, and should be considered when interpreting the results.
Additionally, selection bias is possible given tissue specimens were not available for all RCC
cases in Vamvakas et al. (1998) or in Charbotel et al. (2006). Some uncertainty associated with
misclassification bias is possible given the lack of TCE exposure information to individual
subjects in Schraml et al. (1999) and in Charbotel et al. (2007) from their use of broader
exposure assessment approach compared to that associated with the higher confident exposure
assignment approach. A recent study by Nickerson et al. (2008) addresses many of these
concerns by utilizing more sensitive methods to look at both the genetic and epigenetic issues
related to VHL inactivation. This study was performed on DNA from frozen tissue samples and
used a more sensitive technique for analysis for mutations (endonuclease scanning) as well as
analyzing for methylation changes that may lead to inactivation of the VHL gene. This method
of analysis was validated on tissue samples with known mutations. Of the 205 cc-RCC samples
analyzed, 169 showed mutations in the VHL gene (82.4%). Of those 36 without mutation, 11
were hypermethylated in the promoter region, which will also lead to inactivation of the VHL
gene. Therefore, this study showed inactivating alterations in the VHL gene (either by mutation
or hypermethylation) in 91% tumor samples analyzed.
The limited animal studies examining the role of VHL mutation following exposure to
chemicals including TCE are described below in Section 4.3.6.1.1.
4.3.4 Kidney non-cancer toxicity in laboratory animals
Acute, subchronic, and chronic exposures to TCE cause toxicity to the renal tubules in
rats and mice of both sexes. Nephrotoxicity from acute exposures to TCE has only been reported
at relatively high doses, although histopathological changes have not been investigated in these
experiments. Chakrabarty and Tuchweber (1988) found that TCE administered to male F344
rats by intraperitoneal injection (723-2890 mg/kg) or by inhalation (1,000-2,000 ppm for 6 hr)
produced elevated urinary NAG, GGT, glucose excretion, BUN, and high molecular weight
protein excretion, characteristic signs of proximal tubular, and possibly glomerular injury, as
soon as 24h post-exposure. In the intraperitoneal injection experiments, inflammation was
6/22/2009
423
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
observed, although some inflammation is expected due to the route of exposure, and
nephrotoxicity effects were only statistically significantly elevated at the highest dose (2,890
mg/kg). In the inhalation experiments, the majority of the effects were statistically significant at
both 1,000 and 2,000 ppm. Similarly, at these exposures, renal cortical slice uptake of p-
aminohippurate was inhibited, indicating reduced proximal tubular function. Cojocel et al.
(1989) found similar effects in mice administered TCE by intraperitoneal injection (120-1,000
mg/kg) at 6h post-exposure, such as the dose-dependent increase in plasma BUN concentrations
and decrease in p-aminohippurate accumulation in renal cortical slices. In addition,
malondialdehyde (MDA) and ethane production were increased, indicating lipid peroxidation.
Kidney weight increases have been observed following inhalation exposure to TCE in
both mice (Kjellstrand et al., 1983b) and rats (Woolhiser et al., 2006). Kjellstrand et al. (1983b)
demonstrated an increase in kidney weights in both male (20% compared to control) and female
(10% compared to control) mice following intermittent and continuous TCE whole-body
inhalation exposure (up to 120 days). This increase was significant in males as low as 75 ppm
exposure and in females starting at 150ppm exposure. The latter study, an unpublished report by
Woolhiser et al. (2006), was designed to examine immunotoxicity of TCE but also contains
information regarding kidney weight increases in female Sprague Dawley (CD) rats exposed to
0-ppm, 100-ppm, 300-ppm, and 1,000-ppm TCE for 6 hours/day, 5 days/week, for 4 weeks.
Relative kidney weights were significantly elevated (17.4% relative to controls) at l,000ppm
TCE exposure. However, the small number of animals and the variation in initial animal weight
limit the ability of this study to determine statistically significant increases.
Similarly, overt signs of subchronic nephrotoxicity, such as changes in blood or urinary
biomarkers, are also primarily a high dose phenomenon, although histopathological changes are
evident at lower exposures. Green et al. (1997b) reported administration of 2,000 mg/kg-d TCE
by corn oil gavage for 42 days in F344 rats caused increases of around 2-fold of control results in
urinary markers of nephrotoxicity such as urine volume and protein (both 1.8x), NAG (1.6x),
glucose (2.2x) and ALP (2.Ox), similar to the results of the acute study of Chakrabarty and
Tuchweber (1988), above. At lower dose levels, Green et al.(1998b) reported that plasma and
urinary markers of nephrotoxicity were unchanged. In particular, after 1-28 day exposures to
250 or 500 ppm TCE for 6 hours/day, there were no statistically significant differences in plasma
levels of BUN or in urinary levels of creatinine, protein, ALP, NAG, or GGT. However,
increased urinary excretion of formic acid, accompanied by changes in urinary pH and increased
ammonia, was found at these exposures. Interestingly, at the same exposure level of 500 ppm
(6h/day, 5 days/week, for 6 months), Mensing et al. (2002) reported elevated excretion of low
molecular weight proteins and NAG, biomarkers of nephrotoxicity, but after the longer exposure
duration of 6 months.
6/22/2009
424
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Numerous studies have reported histological changes from TCE exposure for subchronic
and chronic durations (Maltoni et al., 1988, 1986; Mensing et al., 2002; NTP, 1990, 1988). As
summarized in Table 4.3.5, in 13-week studies in F344 rats and B6C3F1 mice, NTP (1990)
reported relatively mild cytomegaly and karyomegaly of the renal tubular epithelial cells at the
doses 1,000-6,000 mg/kg/day (at the other doses, tissues were not examined). The NTP report
noted that "these renal effects were so minimal that they were diagnosed only during a
reevaluation of the tissues ... prompted by the production of definite renal toxicity in the 2-year
study." In the 6 month, 500-ppm inhalation exposure experiments of Mensing et al. (2002),
some histological changes were noted in the glomeruli and tubuli of exposed rats, but they
provided no detailed descriptions beyond the statement that "perivascular, interstitial infections
and glomerulonephritis could well be detected in kidneys of exposed rats."
After 1-2 years of chronic TCE exposure by gavage (NCI, 1976; NTP, 1990, 1988) or
inhalation (Maltoni et al., 1988) (Tables 4.3.5-4.3.9), both the incidence and severity of these
effects increases, with mice and rats exhibiting lesions in the tubular epithelial cells of the inner
renal cortex that are characterized by cytomegaly, karyomegaly, and toxic nephrosis. As with
the studies at shorter duration, these chronic studies reported cytomegaly and karyomegaly of
tubular cells. NTP (1990) specified the area of damage as the pars recta, located in the
corticomedullary region. It is important to note that these effects are distinct from the chronic
nephropathy and inflammation observed in control mice and rats (Lash et al., 2000b; Maltoni et
al., 1988; NCI, 1976).
These effects of TCE on the kidney appear to be progressive. Maltoni et al. (1988) noted
that the incidence and degree of renal toxicity increased with increased exposure time and
increased time from the start of treatment. As mentioned above, signs of toxicity were present in
the 13 week study (NTP, 1988), and NTP (1990) noted cytomegaly at 26 weeks. NTP (1990)
noted that as "exposure time increased, affected tubular cells continued to enlarge and additional
tubules and tubular cells were affected," with toxicity extending to the cortical area as kidneys
became more extensively damaged. NTP (1988, 1990) noted additional lesions that increased in
frequency and severity with longer exposure, such as dilation of tubules and loss of tubular cells
lining the basement membrane ("stripped appearance" [NTP, 1988] or flattening of these cells
[NTP, 1990]). NTP (1990) also commented on the intratubular material and noted that the
tubules were empty or "contained wisps of eosinophilic material."
With gavage exposure, these lesions were present in both mice and rats of both sexes, but
were on average more severe in rats than in mice, and in male rats than in female rats (NTP,
1990). Thus it appears that male rats are most sensitive to these effects, followed by female rats
and then mice. This is consistent with the experiments of Maltoni et al. (1988), which only
reported these effects in male rats. The limited response in female rats or mice of either sex in
6/22/2009
425
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
these experiments may be related to dose or strain. The lowest chronic gavage doses in the NCI
(1976) and NTP (1988, 1990) F344 rat experiments was 500 mg/kg-day, and in all these cases at
least 80% (and frequently 100%) of the animals showed cytomegaly or related toxicity. By
comparison, the highest gavage dose in the Maltoni et al. (1988) experiments (250 mg/kg-day)
showed lower incidences of renal cytomegaly and karyomegaly in male Sprague-Dawley rats
(47%) and 67%, overall and corrected incidences) and none in female rats. The B6C3F1 mouse
strain was used in the NCI (1976), NTP (1990), and Maltoni et al. (1988) studies (Tables
4.3.6-4.3.9). While the two gavage studies (NCI, 1976; NTP, 1990) were consistent, reporting
at least 90% incidence of cytomegaly and karyomegaly at all studied doses, whether dose
accounts for the lack of kidney effects in Maltoni et al. (1988) requires comparing inhalation and
gavage dosing. Such comparisons depend substantially on the internal dose metric, so
conclusions as to whether dose can explain differences across studies cannot be addressed
without dose-response analysis using PBPK modeling. Some minor differences were found in
the multi-strain NTP study (NTP, 1988), but the high rate of response makes distinguishing
among them difficult. Soffritti (personal communication with JC Caldwell, February 14, 2006)
did note that the colony from which the rats in Maltoni et al. (1986, 1988) experiments were
derived had historically low incidences of chronic progressive nephropathy and renal cancer.
6/22/2009
426
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 TABLE 4.3.5 Summary of Renal Toxicity and Tumor Findings in Gavage Studies of
2 Trichloroethylene by NTP (1990)
Cytomegaly and
Adenoma
Adenocarcinoma
Karyomegaly
Sex Dose (mg/kg)"
Incidence (Severity6)
(overall; terminal) (overall; terminal)
1/d, 5d/week, 13-wk study, F344/N rats
Male 0, 125, 250, 500, 100
Tissues not evaluated
None reported
2,000 8/9 (Minimal/mild)
Female 0, 62.5, 125, 250, 500 Tissues not evaluated
1,000
5/10 (Equivocal/minimal)
1/d, 5d/week, 13-wk study, B6C3Fi mice
Male 0, 375, 750, 1,500
3,000
6,000
Female 0, 375, 750, 1,500
Tissues not evaluated
7/10C (Mild/moderate)
d
Tissues not evaluated
None reported
3,000
6,000
9/10 (Mild/moderate)
1/10 (Mild/moderate)
1/d, 5d/week, 103-wk study, F344/N rats
Male 0
0% (0)
0/48; 0/33
0/48; 0/33
500
98% (2.8)
2/49; 0/20
0/49; 0/20
1,000
98% (3.1)
0/49; 0/16
3/49; 3/16e
Female 0
0% (0)
0/50; 0/37
0/50; 0/37
500
100% (1.9)
0/49; 0/33
0/49; 0/33
1,000
100% (2.7)
0/48; 0/26
1/48; 1/26
1/d, 5d/week, 103-wk study, B6C3Fi mice
Male 0
0% (0)
1/49; 1/33
0/49; 0/33
1,000
90% (1.5)
0/50; 0/16
1/50; 0/16
Female 0
0% (0)
0/48; 0/32
0/48; 0/32
1,000
98% (1.8)
0/49; 0/23
0/49; 0/23
3 "Corn oil vehicle.
4 ^Numerical scores reflect the average grade of the lesion in each group (1, slight; 2, moderate; 3, well marked; and
5 4, severe).
6 cObserved in four mice that died after 7-13 wk and in three that survived the study.
7 "All mice died during the first week.
6/22/2009
427
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 eP = 0.028
2
6/22/2009
428
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
2 TABLE 4.3.6 Summary of Renal Toxicity and Tumor Findings in Gavage Studies of
3 Trichloroethylene by NCI (1976)
Sex
Dose (mg/kg)"
Toxic Nephrosis
Adenoma or Adenocarcinoma
(overall; terminal)
(overall; terminal)6
1/d, 5d/week, 2-yr study, Osborn-Mendel rats
Males
0
0/20; 0/2
0/20; 0/2
549
46/50; 7/7
l/50c; 0/7
1,097
46/50; 3/3
0/50; 0/3
Females
0
0/20; 0/8
0/20/ 0/8
549
39/48; 12/12
0/48; 0/12
1,097
48/50; 13/13
0/50; 0/13
1/d, 5d/week, 2-yr study, B6C3F1
mice
Males
0
0/20; 0/8
0/20; 0/8
1,169
48/50; 35/35
0/50; 0/35
2,339
45/50; 20/20
l/50rf; 1/20
Females
0
0/20; 0/17
0/20; 0/17
869
46/50; 40/40
0/50; 0/40
1,739
46/47e; 39/39
0/47; 0/39
4 " Treatment period was 48 weeks for rats, 66 weeks for mice. Doses were changed several times during the study
5 based on monitoring of body weight changes and survival. Dose listed here is the time-weighted average dose over
6 the days on which animals received a dose.
7 b A few malignant mixed tumors and hamartomas of the kidney were observed in control and low dose male rats, but
8 are not counted here.
9 " Tubular adenocarcinoma
10 d Tubular adenoma
11 0 One mouse was reported with "nephrosis," but not "nephrosis toxic," and so was not counted here.
6/22/2009
429
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 TABLE 4.3.7 Summary of Renal Toxicity Findings in Gavage Studies of Trichloroethylene by
2 Maltoni et al. (1988)
Sex Dose (mg/kg)" Megalonucleocytosi sb
(overall; correctedc)
1/d, 4-5d/week, 52-wk exposure, observed for lifespan,
Sprague-Dawley rats
Males
0
0/20; 0/22
50
0/30; 0/24
250
14/30; 14/21
Females
0
0/30; 0/30
50
0/30; 0/29
250
0/30; 0/26
3 " Olive oil vehicle.
4 b Renal tubuli megalonucleocytosis is the same as cytomegaly and karyomegaly of renal tubuli cells (Maltoni et al.,
5 1988).
6 " Denominator for "corrected" incidences is the number of animals alive at the time of the first kidney lesion in this
7 experiment (39 weeks).
6/22/2009
430
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 TABLE 4.3.8 Summary of Renal Toxicity and Tumor Incidence in Gavage Studies of
2 Trichloroethylene by NTP (1988)
Toxic Adenoma Adenocarcinoma
Sex Dose (mg/kg)" Cytomegaly Nephropathy (overall; terminal) (overall; terminal)
1/day, 5d/week, 2-yr study, ACI rats
Male 0
0/50
0/50
0/50;
0/38
0/50
0/38
500
40/49
18/49
0/49;
0/19
1/49
0/19
1,000
48/49
18/49
0/49;
0/11
0/49
0/11
Female 0
0/48
0/48
0/48;
0/34
0/48
0/34
500
43/47
21/47
2/47;
1/20
1/47
1/20
1,000
42/43
19/43
0/43;
0/19
1/43
0/19
1/day, 5d/week, 2-yr study, August rats
Male 0
0/50
0/50
0/50;
0/21
0/50
0/21
500
46/50
10/50
1/50;
0/13
1/50
1/13
1,000
46/49
31/49
1/49;
1/16
0/49
0/16
Female 0
0/49
0/49
1/49;
1/23
0/49
0/23
500
46/48
8/48
2/48;
1/26
2/48
2/26
1,000
50/50
29/50
0/50;
0/25
0/50
0/25
1/day, 5d/week, 2-yr study, Marshall rats
Male 0
0/49
0/49
0/49;
0/26
0/49
0/26
500
48/50
18/50
1/50;
0/12
0/50
0/12
1,000
47/47
23/47
0/47;
0/6
1/47
0/6
Female 0
0/50
0/50
1/50;
0/30
0/50
0/30
500
46/48
30/48
1/48;
1/12
1/48
0/12
1,000
43/44
30/44
0/44;
0/10
1/44
1/10
1/day, 5d/week, 2-yr study, Osborne-Mendel rats
Male 0
0/50
0/50
0/50;
0/22
0/50
0/22
500
48/50
39/50
6/50;
5/17
0/50
0/17
1,000
49/50
35/50
1/50;
1/15
1/50
0/15
Female 0
0/50
0/50
0/50;
0/20
0/50
0/20
500
48/50
30/50
0/50;
0/11
0/50
0/11
1,000
49/49
39/49
1/49;
0/7
0/49
0/7
3 "Corn oil vehicle.
4
6/22/2009
431
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TABLE 4.3.9 Summary of Renal Toxicity and Tumor Findings in Inhalation Studies of
Trichloroethylene by Maltoni et al. (1988)"
Concentration Meganucl eocy tosi sb Adenoma Adenocarcinoma
Sex (ppm) (overall; corrected) (overall; corrected) (overall; corrected)
7h/day, 5d/week, 2-yr exposure, observed for lifespan, Sprague-Dawley ratsc
Male 0
0/135; 0/122
0/135; 0/122
0/135;
0/122
100
0/130; 0/121
1/130; 1/121
0/130;
0/121
300
22/130; 22/ 116
0/130; 0/116
0/130;
0/ 116
600
101/130; 101/124
1/130; 1/124
4/130;
4/124
Female 0
0/145; 0/141
0/145; 0/141
0/145;
0/141
100
0/130; 0/128
1/130; 1/128
0/130;
0/128
300
0/130; 0/127
0/130; 0/127
0/130;
0/127
600
0/130; 0/127
0/130; 0/127
1/130;
1/127
7h/day, 5d/week, 78-wk exposure, observed for lifespan, B6C3F1
mice6'
Male 0
0/90
0/90
0/90
100
0/90
0/90
1/90
300
0/90
0/90
0/90
600
0/90
0/90
0/90
Female 0
0/90
0/90
1/90
100
0/90
0/90
0/90
300
0/90
0/90
0/90
600
0/90
0/90
0/90
" Three inhalation experiments in this study found no renal megalonucleocytosis, adenomas, or adenocarcinomas:
BT302 (8-wk exposure to 0, 100, 600 ppm in Sprague-Dawley rats); BT303 (8-wk exposure to 0, 100, 600 ppm in
Swiss mice); and BT305 (78-wk exposure to 0, 100, 300, 600 ppm in Swiss mice).
b Renal tubuli meganucleocytosis is the same as cytomegaly and karyomegaly of renal tubuli cells (Maltoni et al.,
1988).
" Combined incidences from experiments BT304 and BT304bis. Corrected incidences reflect number of rats alive at
47 weeks, when the first renal tubular megalonucleocytosis in these experiments appeared.
d Female incidences are from experiment BT306, while male incidences are from experiment BT306bis, which was
added to the study because of high, early mortality due to aggressiveness and fighting in males in experiment
BT306. Corrected incidences not show, because only the renal adenocarcinomas appeared at 107 weeks in the male
and 136 in the female, when the most of the mice were already deceased.
6/22/2009
432
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.3.5 Kidney cancer in laboratory animals
Kidney cancer is an extremely rare occurrence historically in rats, occurring in only 0.4%
of corn oil gavage controls in NTP studies (Rhomberg, 2000). Carcinogenicity bioassays with
TCE and its metabolites have shown evidence of neoplastic lesions in the kidney, mainly in male
rats. Although these studies have shown limited increases in kidney tumors, given the rarity of
these tumors and the repeatability of this result, these are considered biologically significant.
4.3.5.1 Inhalation Studies of TCE
A limited number of inhalation studies examined the carcinogenicity of TCE, with no
statistically-significantly increases in kidney tumor incidence reported in mice or hamsters
(Fukuda et al., 1983; Henschler et al., 1980; Maltoni et al., 1988, 1986). Maltoni et al. (1988)
observed five renal adenocarcinomas (four/130 males, one/130 female) in Sprague-Dawley rats
after 8 weeks of exposure to 600ppm TCE. In males, these tumors seemed to have originated in
the tubular cells and have not been seen in historical controls. The cortical adenocarcinoma in
the female rat was cortical and similar to that seen infrequently in historical controls. This study
also demonstrated the appearance of increased cytokaryomegaly as a potential precursor to
kidney cancer. This lesion had a significantly and dose-dependently increased in male rats only
(Table 4.3.9). The inhalation studies by Fukuda et al. (1983) in Sprague-Dawley rats and female
ICR mice, reported one clear cell carcinoma in rats exposed to the highest concentration (450
ppm) but saw no increase in kidney tumors in mice. This result was not statistically significant
(Table 4.3.10). The cancer bioassay by (Maltoni et al., 1986, 1988) reported no statistically
significant increase in kidney tumors in mice or hamsters, but renal adenocarcinomas were found
in male rats at the high dose (600 ppm) at 2 years (4/130). This exposed group also experienced
cytokaryomegaly or megalonucleocytosis (101/130), as did a small percentage of the mid-dose
group (300ppm, 22/130). Pathology was not described, so it is not possible to know if increased
levels of nephrotoxicity were observed in the higher dose group, or in animals that then had
tumors (Table 4.3.9). One negative study (Henschler et al., 1980) tested NMRI mice, Wistar rats
and Syrian hamsters of both sexes (60 animals per strain), and observed no significant increase in
renal tubule tumors any of the species tested. An increase in benign adenomas was observed in
male mice and rats, with no renal adenocarcinomas reported in females of either species (Table
4.3.10).
6/22/2009
433
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.3.5.2 Gavage and Drinking Water Studies of TCE
Chronic gavage studies exposing multiple strains of rats and mice to 0-3,000 mg/kg TCE
for at least 52 weeks (Table 4.3.6-4.3.8) reported a statistically-significant excess in kidney
tumors only in males at the highest doses (Henschler et al., 1984; Maltoni et al., 1986; NCI,
1976; NTP, 1988, 1990; Van Duuren et al., 1979).
Van Duuren et al. (1979) examined TCE and 14 other halogenated compounds for
carcinogenicity in both sexes of Swiss mice. While no excess tumors were observed, the dose
rate (0.5 mg once per week, or an average dose rate of approximately 2.4 mg/kg-day for a 30 g
mouse) is about 400-fold lower than that in the other gavage studies. Inadequate design and
reporting of this study limit the ability to use the results as an indicator of TCE carcinogenicity.
In the NCI (1976) study, the results for Osborne-Mendel rats were considered inconclusive due
to significant early mortality. In rats of both sexes, no increase was seen in primary tumor
induction over that observed in controls. While both sexes of B6C3F1 mice showed a
compound-related increase in nephropathy, no increase in tumors over historical controls was
observed. The NCI study (1976) used technical grade TCE which contained two known
carcinogenic compounds as stabilizers (epichlorohydrin and 1,2-epoxybutane). However, later
study by Henschler et al. (1984) in mice reported no significant differences in systemic
tumorigenesis between pure, industrial, and stabilized TCE, suggesting that concentrations of
these stabilizers are too low to be the cause of tumors. A later gavage study by NTP (1988),
using TCE stabilized with diisopropylamine, observed an increased incidence of renal tumors in
all four strains of rats (ACI, August, Marshall and Osborne-Mendel). All animals exposed for up
to 2 years (rats and mice) had non-neoplastic kidney lesions, even if they did not later develop
kidney cancer (Table 4.3.8). The final NTP study (1990) in male and female F344 rats and
B6C3F1 mice used epichlorohydrin-free TCE, and reported early mortality in male rats. Only in
the highest dose group (l,000mg/kg) of male F344 rats was renal carcinoma statistically
significant increased. Cytomegaly and karyomegaly were also increased, particularly in male
rats. The toxic nephropathy observed in both rats and mice led to a poor survival rate, rendering
this study inadequate for determining carcinogenicity (Table 4.3.5). As discussed previously,
this toxic nephropathy was clearly distinguishable from the spontaneous chronic progression
nephropathy commonly observed in aged rats.
6/22/2009
434
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 TABLE 4.3.10 Summary of Renal Tumor Findings in Inhalation Studies of Trichloroethylene
2 by Henschler et al. (1980)a and Fukuda et al. (1983)6
Sex Concentration (ppm)
Adenomas
Adenocarcinomas
6h/day, 5d/week, 18 month exposure, 30 months observation, Han:NMRI mice
(Henschler et al., 1980)
Males 0
0/30
1/30
100
0/29
0/30
500
0/29
0/30
Females 0
0/29
0/29
100
0/30
0/30
500
0/28
0/28
6h/day, 5d/week, 18-month exposure, 36-months observation, Han:WIST rats
(Henschler et al., 1980)
Males 0
0/29
0/29
100
1/30
0/30
500
1/30
1/30
Females 0
0/28
0/28
100
0/30
0/30
500
1/30
0/30
7h/day, 5d/week, 2-yr study, Cij:CD (SD) rats (Fukuda et al.,
,1983)
Females 0
0/50
0/50
50
0/50
0/50
150
0/47
0/47
450
0/51
1/50
3 " Henschler et al. (1980) observed no renal tumors control or exposed Syrian hamsters.
4 h Fukuda et al. (1983) observed no renal tumors in control or exposed Cij :CD-1 (ICR) mice.
5
6/22/2009
435
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.3.6 Role of metabolism in TCE kidney toxicity
It is generally thought that one or more TCE metabolites rather than the parent compound
are the active moieties for TCE nephrotoxicity. As reviewed in Section 3.3, oxidation by P450s,
of which CYP2EI is thought to be the most active isoform, results in the production of chloral
hydrate, trichloroacetic acid, dichloroacetic acid and trichloroethanol. The glutathione
conjugation pathway produces metabolites such as S-(l,2-dichlorovinyl)glutathione (DCVG), S-
(l,2-dichlorovinyl)-L-cysteine (DCVC), dichlorovinylthiol (DCVSH) and N-acetyl-S-(l,2-
dichlorovinyl)-L-cysteine (NAcDCVC). Because several of the steps for generating these
reactive metabolites occur in the kidney, the GSH conjugation pathway has been thought to be
responsible for producing the active moiety or moieties of TCE nephrotoxicity. A comparison of
TCE's nephrotoxic effects with the effects of TCE metabolites, both in vivo and in vitro, thus
provides a basis for assessing the relative roles of different metabolites. While most of the
available data have been on metabolites from GSH conjugation, such as DCVC, limited
information is also available on the major oxidative metabolites TCOH and TCA.
4.3.6.1 In vivo studies of the kidney toxicity of TCE metabolites
4.3.6.1.1 Role of GSH conjugation metabolites of TCE
In numerous studies, DCVC has been shown to be acutely nephrotoxic in rats and mice.
Mice receiving a single dose of 1 mg/kg DCVC (the lowest dose tested in this species) exhibited
karyolytic proximal tubular cells in the outer stripe of the outer medulla, with some sloughing of
cells into the lumen and moderate desquamation of the tubular epithelium (Eyre et al., 1995b).
Higher doses in mice were associated with more severe histological changes similar to those
induced by TCE, such as desquamation and necrosis of the tubular epithelium (Darnerud et al.,
1989; Terracini and Parker, 1965a; Vaidya et al., 2003a,b). In rats, no histological changes in the
kidney were reported after single doses of 1, 5, and 10 mg/kg DCVC (Eyre et al., 1995a; Green
et al., 1997a), but cellular debris in the tubular lumen was reported at 25 mg/kg (Eyre et al.,
1995b) and slight degeneration and necrosis were seen at 50 mg/kg (Green et al., 1997). Green
et al. (1997) reported no histological changes were noted in rats after 10 doses of 0.1-5.0 mg/kg
DCVC (although increases in urinary protein and GGT were found), but some karyomegaly was
noted in mice after 10 daily doses of 1 mg/kg. Therefore, mice appear more sensitive than rats to
the nephrotoxic effects of acute exposure to DCVC, although the number of animals used at each
dose in these studies was limited (10 or less). Although the data are not sufficient to assess the
relatively sensitivity of other species, it is clear that multiple species, including rabbits, guinea
6/22/2009
436
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
pigs, cats, and dogs, are responsive to DCVC's acute nephrotoxic effects (Jaffe et al., 1984;
Krejci et al., 1991; Terracini and Parker, 1965b; Wolfgang et al., 1989b).
Very few studies are available at longer durations. Terracini and Parker (1965) gave
DCVC in drinking water to rats at a concentration of 0.01% for 12 weeks (approximately 10
mg/kg-day), and reported consistent pathological and histological changes in the kidney. The
progression of these effects was as follows: (i) during the first few days, completely necrotic
tubules, with isolated pyknotic cells being shed into the lumen; (ii) after 1 week, dilated tubules
in the inner part of the cortex, lined with flat epithelial cells that showed thick basal membranes,
some with big hyperchromatic nuclei; (iii) in the following weeks, increased prominence of
tubular cells exhibiting karyomegaly, seen in almost all animals, less pronounced tubular
dilation, and cytomegaly in the same cells showing karyomegaly. In addition, increased mitotic
activity was reported the first few days, but was not evident for the rest of the experiment.
Terracini and Parker (1965) also reported the results of a small experiment (13 male and 5
female rats) given the same concentration of DCVC in drinking water for 46 weeks, and
observed for 87 weeks. They noted renal tubular cells exhibiting karyomegaly and cytomegaly
consistently throughout the experiment. Moreover, a further group of 8 female rats given DCVC
in drinking water at a concentration of 0.001% (approximately 1 mg/kg-day) also exhibited
similar, though less severe, changes in the renal tubules. In mice, Jaffe et al. (1984) gave DCVC
in drinking water at concentrations of 0.001%, 0.005%), and 0.01%> (estimated daily dose of 1-2,
7-13, and 17-22 mg/kg-day), and reported similar effects in all dose groups, including
cytomegaly, nuclear hyperchromatism, and multiple nucleoli, particularly in the pars recta
section of the kidney. Thus, effects were noted in both mice and rats under chronic exposures at
doses as low as 1-2 mg/kg-day (the lowest dose tested). Therefore, while limited, the available
data do not suggest differences between mice and rats to the nephrotoxic effects of DCVC under
chronic exposure conditions, in contrast to the greater sensitivity of mice to acute and sub-
chronic DCVC-induced nephrotoxicity.
Importantly, as summarized in Table 4.3.11, the histological changes and their location in
these subchronic and chronic experiments with DCVC are quite similar to those reported in
chronic studies of TCE, described above, particularly the prominence of karyomegaly and
cytomegaly in the pars recta section of the kidney. Moreover, the morphological changes in the
tubular cells, such as flattening and dilation, are quite similar. Similar pathology is not observed
with the oxidative metabolites alone (Section 4.3.6.1.2).
Additionally, it is important to consider whether sufficient DCVC may be formed from
TCE exposure to account for TCE nephrotoxicity. While direct pharmacokinetic measurements,
such as the excretion of NAcDCVC, have been used to argue that insufficient DCVC would be
formed to be the active moiety for nephrotoxicity (Green et al., 1997), as discussed in Chapter 3,
6/22/2009
437
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
urinary NAcDCVC is a poor marker of the flux through the GSH conjugation pathway because
of the many other possible fates of metabolites in that pathway. In another approach, Eyre et al.
(1995b) using acid-labile adducts as a common internal dosimeter between TCE and DCVC, and
reported that a single TCE dose of 400 mg/kg in rats (similar to the lowest daily doses in the NCI
and NTP rat bioassays) and 1,000 mg/kg (similar to the lowest daily doses in the NCI and NTP
mouse bioassays) corresponded to a single equivalent DCVC dose of 6 mg/kg-day and 1 mg/kg-
day in rats and mice, respectively. These equivalent doses of DCVC are greater or equal to those
in which nephrotoxicity has been reported in these species under chronic conditions. Therefore,
assuming that this dose correspondence is accurate under chronic conditions, sufficient DCVC
would be formed from TCE exposure to explain the observed histological changes in the renal
tubules.
The Eker rat model (Tsc-2+/) is at increased risk for the development of spontaneous
renal cell carcinoma and as such has been used to understand the mechanisms of renal
carcinogenesis (Stemmer et al., 2007; Wolf et al., 2000). One study has demonstrated similar
pathway activation in Eker rats as that seen in humans with VHL mutations leading to renal cell
carcinoma, suggesting Tsc-2 inactivation is analogous to inactivation of VHL in human renal cell
carcinoma (Liu et al., 2003). Although the Eker rat model is a useful tool for analyzing
progression of renal carcinogenesis, it has some limitations in analysis of specific genetic
changes, particularly given the potential for different genetic changes depending on type of
exposure and tumor. The results of short-term assays to genotoxic carcinogens in the Eker rat
model (Morton et al., 2002; Stemmer et al., 2007) reported limited pre-neoplastic and neoplastic
lesions which may be related to the increased background rate of renal carcinomas in this animal
model.
Recently, Mally et al. (2006) exposed male rats carrying the Eker mutation to TCE
(0-l,000mg/kg bw) by corn oil gavage and demonstrated no increase in renal preneoplastic
lesions or tumors. Primary Eker rat kidney cells exposed to DCVC in this study did induce an
increase in transformants in vitro but no DCVC-induced vhl or Tsc-2 mutations were observed.
In vivo exposure to TCE (5 days/week for 13 weeks), decreased body weight gain and increased
urinary excretion at the two highest TCE concentrations analyzed (500 and l,000mg/kg bw) but
did not change standard nephrotoxicity markers (GGT, creatinine and urinary protein). Renal
tubular epithelial cellular proliferation as measured by BrdU incorporation was demonstrated at
the three highest concentrations of TCE (250, 500 and l,000mg/kg/day). A minority of these
cells also showed karyomegaly at the two higher TCE concentrations. Although renal cortical
tumors were demonstrated in all TCE exposed groups, these were not significantly different from
controls (13 weeks). These studies were complemented with in vitro studies of DCVC
(10-50uM) in rat kidney epithelial (RKE) cells examining proliferation at 8, 24, and 72 h and
6/22/2009
438
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
cellular transformation at 6-7 weeks. Treatment of RKE cells from susceptible rats with DCVC
gave rise to morphologically transformed colonies consistently higher than background (Mally et
al., 2006). Analyzing ten of the renal tumors from the TCE exposed rats and nine of the DCVC
transformants from these studies for alterations to the VHL gene that might lead to inactivation
found no alterations to VHL gene expression or mutations.
One paper has linked the VHL gene to chemical-induced carcinogenesis. Shiao et al.
(1998) demonstrated VHL gene somatic mutations in N-nitrosodimethylamine-induced rat
kidney cancers that were of the clear cell type. The clear cell phenotype is rare in rat kidney
cancers, but it was only the clear cell cancers that showed VHL somatic mutation (three of eight
tumors analyzed). This provided an additional link between VHL inactivation and clear cell
kidney cancer. However, this study examined archived formalin fixed paraffin embedded tissues
from previous experiments. As described previously (Sec 4.3.2), DNA extraction from this type
of preparation creates some technical issues. Similarly, archived formalin-fixed paraffin
embedded tissues from rats exposed to potassium bromide were analyzed in a later study by
Shiao et al. (2002). This later study examined the VHL gene mutations following exposure to
potassium bromide, a rat renal carcinogen known to induce clear cell renal tumors. Clear cell
renal tumors are the most common form of human renal epithelial neoplasms, but are extremely
rare in animals. Although F344 rats exposed to potassium bromide in this study did develop
renal clear cell carcinomas, only two of nine carried the same C to T mutation at the core region
of the Spl transcription-factor binding motif in the VHL promoter region, and one of four
untreated animals had a C to T mutation outside the conserved core region. Mutation in the VHL
coding region was only detected in one tumor, so although the tumors developed following
exposure to potassium bromide were morphologically similar to those found in humans; no
similarities were found in the genetic changes.
Elfarra et al. (1984) found that both DCVG and DCVC administered to male F344 rats by
intraperitoneal injections in isotonic saline resulted in elevations in BUN and urinary glucose
excretion. Furthermore, inhibition of renal GGT activity with acivicin protected rats from
DCVG-induced nephrotoxicity. In addition, both the B-lyase inhibitor AOAA and the renal
organic anion transport inhibitor probenecid provided protection from DCVC, demonstrating a
requirement for metabolism of DCVG to the cysteine conjugate by the action of renal GGT and
dipeptidase, uptake into the renal cell by the organic anion transporter, and subsequent activation
by the B-lyase. This conclusion was supported further by showing that the -methyl analog of
DCVC, which cannot undergo a B-elimination reaction due to the presence of the methyl group,
was not nephrotoxic.
Korrapati et al. (2005) builds upon a series of investigations of hetero- (by HgCh) and
homo-(by DCVC, 15 mg/kg) protection against a lethal dose of DCVC (75 mg/kg). Priming, or
6/22/2009
439
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
preconditioning, with pre-exposure to either HgCh or DCVC of male Swiss-Webster mice was
said to augment and sustain cell division and tissue repair, hence protecting against the
subsequent lethal DCVC dose(Vaidya et al., 2003a, b, c). Korrapati et al. (2005) showed that a
lethal dose of DCVC downregulates phosphorylation of endogenous retinoblastoma protein
(pRb), which is considered critical in renal proximal tubular and mesangial cells for the passage
of cells from G1 to S-phase, thereby leading to a block of renal tubule repair. Priming, in
contrast, upregulated P-pRB which was sustained even after the administration of a lethal dose of
DCVC, thereby stimulating S-phase DNA synthesis, which was concluded to result in tissue
repair and recovery from acute renal failure and death. These studies are more informative about
the mechanism of autoprotection than on the mechanism of initial injury caused by DCVC. In
addition, the priming injury (not innocuous, as it caused 25-50% necrosis and elevated blood
urea nitrogen) may have influenced the toxicokinetics of the second DCVC injection.
4.3.6.1.2 Role of oxidative metabolites of TCE
Some investigators (Green et al., 1998, 2003; Dow and Green, 2000) have proposed that
TCE nephrotoxicity is related to formic acid formation. They demonstrated that exposure to
either trichloroethanol or trichloroacetic acid causes increased formation and urinary excretion of
formic acid (Green et al., 1998). The formic acid does not come from trichloroethylene. Rather,
trichloroethylene (or a metabolite) has been proposed to cause a functional depletion of vitamin
B12, which is required for the methionine salvage pathway of folate metabolism. Vitamin Bi2
depletion results in folate depletion. Folate is a cofactor in one-carbon metabolism and depletion
of folate allows formic acid to accumulate, and then to be excreted in the urine (Dow and Green,
2000).
TCE (1 and 5 g/L), TCA (0.25, 0.5 and 1 g/L) and TCOH (0.5 and 1.0 g/L) exposure in
male Fisher rats substantially increased excretion of formic acid in urine, an effect suggested as a
possible explanation for TCE-induced renal toxicity in rats (Green et al., 1998a). Green et al.
(2003a) reported tubular toxicity as a result of chronic (1 year) exposure to TCOH (0, 0.5 and 1.0
g/L). Although TCOH causes tubular degeneration in a similar region of the kidney as TCE,
there are several dissimilarities between the characteristics of nephrotoxicity between the two
compounds, as summarized in Table 4.3.11. In particular, Green et al. (1998) did not observe
TCOH causing karyomegaly and cytomegaly. These effects were seen as early as 13 weeks after
the commencement of TCE exposure (NTP, 1990), with 300 ppm inhalation exposures to TCE
(Maltoni et al., 1988), as well as at very low chronic exposures to DCVC (Terracini and Parker,
1965; Jaffe et al., 1984). In addition, Green et al. (2003) reported neither flattening nor loss of
the tubular epithelium nor hyperplasia, but suggested that the increased early basophilia was due
to newly divided cells, and therefore represented tubular regeneration in response to damage.
Furthermore, they noted that such changes were seen with the spontaneous damage that occurs in
6/22/2009
440
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
aging rats. However, several of the chronic studies of TCE noted that the TCE-induced damage
observed was distinct from the spontaneous nephropathy observed in rats. A recent in vitro
study of rat hepatocytes and primary human renal proximal tubule cells from two donors
measured formic acid production following exposure to CH (0.3-3mM, 3-10 days) (Lock et al.,
2007). This study observed increased formic acid production at day 10 in both human renal
proximal tubule cell strains, but a similar level of formic acid was measured when CH was added
to media alone. The results of this study are limited by the use of only two primary human cell
strains, but suggest exposure to CH does not lead to significant increases in formic acid
production in vivo.
Interestingly, it appears that the amount of formic acid excreted reaches a plateau at a
relatively low dose. Green et al. (2003) added folic acid to the drinking water of the group of
rats receiving the lower dose of TCOH (18.3 mg/kg-day) in order to modulate the excretion of
formic acid in that dose group, and retain the dose-response in formic acid excretion relative to
the higher dose group (54.3 mg/kg-day). These doses of TCOH are much lower than what would
be expected to be formed in vivo at chronic gavage doses. For instance, after a single 500 mg/kg
dose of TCE (the lower daily dose in the NTP rat chronic bioassays), Green and Prout (1985)
reported excretion of about 41% of the TCE gavage dose in urine as TCOH or TCOG in 24 hr.
Thus, using the measure of additional excretion after 24 hr and the TCOH converted to TCA as a
lower bound as to the amount of TCOH formed by a single 500 mg/kg dose of TCE, the amount
of TCOH would be about 205 mg/kg, almost 4-fold greater than the high dose in the Green et al.
(2003) study. By contrast, these TCOH doses are somewhat smaller than those expected from
the inhalation exposures of TCE. For instance, after 6 hour exposure to 100 and 500 ppm TCE
(similar to the daily inhalation exposures in Maltoni et al., 1988), male rats excreted 1.5 and 4.4
mg of TCOH over 48 hr, corresponding to 5 and 15 mg/kg for a rat weighing 0.3 kg (Kaneko et
al., 1994). The higher equivalent TCOH dose is similar to the lower TCOH dose used in Green
et al. (2003), so it is notable that while Maltoni et al. (1988) reported a substantial incidence of
cytomegaly and karyomegaly after TCE exposure (300 and 600 ppm), none was reported in
Green et al. (2003).
TCOH alone does not appear sufficient to explain the range of renal effects observed
after TCE exposure, particularly cytomegaly, karyomegaly, and flattening and dilation of the
tubular epithelium. However, given the studies described above, it is reasonable to conclude that
TCOH may contribute to the nephrotoxicity of TCE, possibly due to excess formic acid
production, because (i) there are some similarities between the effects observed with TCE and
TCOH and (ii) the dose at which effects with TCOH are observed overlap with the approximate
equivalent TCOH dose from TCE exposure in the chronic studies.
6/22/2009
441
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Dow and Green (2000) noted that TCA also induced formic acid accumulation in rats,
and suggested that TCA may therefore contribute to TCE-induced nephrotoxicity. However,
TCA has not been reported to cause any similar histologic changes in the kidney. Mather et al.
(1990) reported an increase of kidney-weight to body-weight ratio in rats after 90 days of
exposure to trichloroacetic acid in drinking water at 5,000 ppm (5 g/L) but reported no
histopathologic changes in the kidney. DeAngelo et al. (1997) reported no effects of
trichloroacetic acid on kidney weight or histopathology in rats in a 2-year cancer bioassay. Dow
and Green (2000) administered TCA at quite high doses (1 and 5 g/L in drinking water), greater
than the subsequent experiments of Green et al. (2003) with TCOH (0.5 and 1 g/L in drinking
water), and reported similar amounts of formic acid produced (about 20 mg/day for each
compound). However, cytotoxicity or karyomegaly did not appear to be analyzed. Furthermore,
much more TCOH is formed from TCE exposure than TCA. Therefore, if TCA contributes
substantially to the nephrotoxicity of TCE, its contribution would be substantially less than that
of TCOH. Lock et al. (2007) also measured formic acid production in human renal proximal
tubule cells exposed to 0.3-3mM CH for 10 days CH. This study measured metabolism of CH
to TCOH and TCA as well as formic acid production and subsequent cytotoxicity. Increased
formic acid was not observed in this study, and limited cytotoxicity was observed. However, this
study was performed in human renal proximal tubular cells from only two donors, and there is
potential for large inter-individual variability in response, particularly with CYP450 enzymes.
In order to determine the ability of various chlorinated hydrocarbons to induce
peroxisomal enzymes, Goldsworthy and Popp (1987) exposed male Fisher-344 rats and male
B6C3F1 mice to TCE (1,000 mg/kg bw) and TCA (500 mg/kg bw) by corn oil gavage for 10
consecutive days. Peroxisomal activation was measured by palmitoyl CoA oxidase activity
levels. TCE led to increased peroxisomal activation in the kidneys of both rats (300% of control)
and mice (625% of control), while TCA led to an increase only in mice (280% of control). A
study by Zanelli et al. (1996) exposed Sprague-Dawley rats to TCA for four days and measured
both renal and hepatic peroxisomal and cytochrome P450 enzyme activities. TCA-treated rats
had increased activity in CYP450 4A subfamily enzymes and peroxisomal palmitoyl-CoA
oxidase. Both of these acute studies focused on enzyme activities and did not further analyze
resulting histopathology.
6/22/2009
442
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 TABLE 4.3.11 Summary of Histological Changes in Renal Proximal Tubular Cells Induced by Chronic Exposure to TCE, DCVC,
2 and TCOHa
Effects
TCE
DCVC
TCOH
Karyomegaly
Enlarged, hyperchromatic nuclei, irregular to
oblong in shape. Vesicular nuclei containing
prominent nucleoli.
Enlarged, hyperchromatic nuclei with
and multiple nucleoli. Nuclear
pyknosis and karyorrhexis.
None reported.
Cytomegaly
Epithelial cells were large, elongated and
flattened.
Epithelial cells were large, elongated
and flattened cells.
No report of enlarged cells.
Cell necrosis/
hyperplasia
Stratified epithelium that partially or completely
filled the tubular lumens. Cells in mitosis were
variable in number or absent. Cells had
abundant eosinophilic or basophilic cytoplasm.
Thinning of tubular epithelium, frank
tubular necrosis, re-epitheliation.
Tubular atrophy, interstitial fibrosis
and destruction of renal parenchyma.
More basophilic and finely vacuolated.
No flattening or loss of
epithelium reported. Increased
tubular cell basophilia,
followed by increased cellular
eosinophilia, tubular cell
vacuolation.
Morphology/
content of
tubules
Some tubules enlarged/dilated to the extent that
they were difficult to identify. Portions of
basement membrane had a stripped appearance.
Tubules were empty or contained "wisps of
eosinophilic material."
Tubular dilation, denuded tubules.
Thick basal membrane. Focal areas of
dysplasia, intraluminal casts.
No tubular dilation reported.
Intra-tubular cast formation.
3 a Sources: NCI (1976); NTP (1988, 1990); Maltoni et al. (1988); Terracini and Parker (1965); Jaffe et al. (1985); Green et al. (2003).
4
6/22/2009
443
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.3.6.2 In vitro studies of kidney toxicity of TCE and metabolites
Generally, it is believed that TCE metabolites are responsible for the bulk of kidney
toxicity observed following exposure. In particular, studies have demonstrated a role for DCVG
and DCVC in kidney toxicity. The work by Lash and colleagues (Cummings et al., 2000a,b;
Cummings and Lash, 2000; Lash et al., 2000a) examined the effect of trichloroethylene and its
metabolites in vitro. Trichloroethylene and DCVC are toxic to primary cultures of rat proximal
and distal tubular cells (Cummings et al., 2000b) while the TCE metabolites DCVG and DCVC
have been demonstrated to be cytotoxic to rat and rabbit kidney cells in vitro (Groves et al.,
1993; Hassall et al., 1983; Lash et al., 2000a, 2001; Wolfgang et al., 1989a). Glutathione-related
enzyme activities were well maintained in the cells, whereas CYP activities were not. The
enzyme activity response to DCVC was greater than the response to trichloroethylene; however,
the proximal and distal tubule cells had similar responses even though the proximal tubule is the
target in vivo. The authors attributed this to the fact that the proximal tubule is exposed before
the distal tubule in vivo and to possible differences in uptake transporters. They did not address
the extent to which transporters were maintained in the cultured cells.
In further studies, Lash et al. (2001) assessed the toxicity of trichloroethylene and its
metabolites DCVC and DCVG using in vitro techniques (Lash et al., 2001) as compared to in
vivo studies. Experiments using isolated cells were performed only with tissues from Fischer
344 rats, and lactate dehydrogenase release was used as the measure of cellular toxicity. The
effects were greater in males. DCVC and trichloroethylene had similar effects, but DCVG
exhibited increased efficacy compared with trichloroethylene and DCVC.
In vitro mitochondrial toxicity was assessed in renal cells from both Fischer 344 rats and
B6C3F1 mice following exposure to both DCVC and DCVG (Lash et al., 2001). Renal
mitochondria from male rats and mice responded similarly; a greater effect was seen in cells
from the female mice. These studies show DCVC to be slightly more toxic than
trichloroethylene and DCVG, but species differences are not consistent with the effects observed
in long-term bioassays. This suggests that in vitro data be used with caution in risk assessment,
being mindful that in vitro experiments do not account for in vivo pharmacokinetic and metabolic
processes.
In LLC-PK1 cells, DCVC causes loss of mitochondrial membrane potential,
mitochondrial swelling, release of cytochrome c, caspase activation, and apoptosis (Chen et al.,
2001). Thus, DCVC is toxic to mitochondria, resulting in either apoptosis or necrosis. DCVC-
induced apoptosis also has been reported in primary cultures of human proximal tubule cells
(Lash et al., 2001).
6/22/2009
444
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
DCVC was further studied in human renal proximal tubule cells for alterations in gene
expression patterns related to proposed modes of action in nephrotoxicity (Lock et al., 2006). In
cells exposed to subtoxic levels of DCVC to better mimic workplace exposures, the expression
of genes involved with apoptosis (caspase 8, FADD-like regulator) was increased at the higher
dose (luM) but not at the lower dose (0. luM) of DCVC exposure. Genes related to oxidative
stress response (SOD, NFkB, p53, c-Jun) were altered at both subtoxic doses, with genes
generally upregulated at 0. luM DCVC being downregulated at luM DCVC. The results of this
study support the need for further study, and highlight the involvement of multiple pathways and
variability of response based on different concentrations.
Lash et al. (2007) examined the effect of modulation of renal metabolism on toxicity of
TCE in isolated rat cells and microsomes from kidney and liver. Following exposure to
modulating chemicals, LDH was measured as a marker of cytotoxicity, and the presence of
specific metabolites was documented (DCVG, TCA, TCOH, CH). Inhibition of the CYP450
stimulated an increase of GSH conjugation of TCE and increased cytotoxicity in kidney cells.
This modulation of CYP450 had a greater effect on TCE-induced cytotoxicity in liver cells than
in kidney cells. Increases in GSH concentrations in the kidney cells led to increased cytotoxicity
following exposure to TCE. Depletion of GSH in hepatocytes exposed to TCE, however, led to
an increase in hepatic cytotoxicity. The results of this study highlight the role of different
bioactivation pathways needed in both the kidney and the liver, with the kidney effects being
more affected by the GSH conjugation pathways metabolic products.
In addition to the higher susceptibility of male rats to TCE-induced
nephrocarcinogenicity and nephrotoxicity, isolated renal cortical cells from male F344 rats are
more susceptible to acute cytotoxicity from TCE than cells from female rats. TCE caused a
modest increase in lactate dehydrogenase (LDH) release from male rat kidney cells but had no
significant effect on LDH release from female rat kidney cells. These results on male
susceptibility to TCE agree with the in vivo data.
4.3.6.3 Conclusions as to the active agents of TCE-induced nephrotoxicity
In summary, the TCE metabolites DCVC, TCOH, and TCA have all been proposed as
possible contributors to the nephrotoxicity of TCE. Both in vivo and in vitro data strongly
support the conclusion that DCVC and related GSH conjugation metabolites are the active agents
of TCE-induced nephrotoxicity. Of these, DCVC induces effects in renal tissues, both in vivo
and in vitro, that are most similar to those of TCE, and formed in sufficient amounts after TCE
exposure to account for those effects. A role for formic acid due to TCOH or TCA formation
from TCE cannot be ruled out, as it is known that substantial TCOH and TCA are formed from
6/22/2009
445
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE exposure, that formic acid is produced from all three compounds, and that TCOH exposure
leads to toxicity in the renal tubules. However, the characteristics of TCOH-induced
nephrotoxicity do not account for the range of effects observed after TCE exposure while those
of DCVC-induced nephrotoxicity do. Also, TCOH does not induce the same pathology as TCE
or DCVC. TCA has also been demonstrated to induce peroxisomal proliferation (Goldsworthy
and Popp, 1987), but this has not been associated with kidney cancer. Therefore, although
TCOH and possibly TCA may contribute to TCE-induced nephrotoxicity, their contribution is
likely to be small compared to that of DCVC.
4.3.7 Mode(s) of Action for Kidney Carcinogenicity
This section will discuss the evidentiary support for several hypothesized modes of action
for kidney carcinogenicity, including mutagenicity, cytotoxicity and regenerative proliferation,
peroxisome proliferation, a2|i-related nephropathy and formic acid-related nephropathy,
following the framework outlined in the Cancer Guidelines (U.S. EPA, 2005a; 2005b).
4.3.7.1 Hypothesized Mode of Action: Mutagenicity
One hypothesis is that TCE acts by a mutagenic mode of action in TCE-induced renal
carcinogenesis. According to this hypothesis, the key event leading to TCE-induced kidney
tumor formation constitute the following: TCE GSH conjugation metabolites (e.g., DCVG,
DCVC, NAcDCVC, and/or other reactive metabolites derived from subsequent beta-lyase, FMO,
or P450 metabolism) derived from the GSH-conjugation pathway, after being either produced in
situ in or delivered systemically to the kidney, cause direct alterations to DNA (e.g., mutation,
DNA damage, and/or micronuclei induction). Mutagenicity is a well-established cause of
carcinogenicity.
Experimental Support for the Hypothesized Mode of Action
Evidence for the hypothesized mode of action for TCE includes 1) the formation of GSH-
conjugation pathway metabolites in the kidney demonstrated in TCE toxicokinetics studies; and
2) the genotoxicity of these GSH-conjugation pathway metabolites demonstrated in most
existing in vitro and in vivo assays of gene mutations (i.e., Ames test) and in assays of
unscheduled DNA synthesis, DNA strand breaks, and micronuclei using both "standard" systems
and renal cells/tissues. Additional relevant data come from analyses of VHL mutations in human
kidney tumors and studies using the Eker rat model. These lines of evidence are elaborated
below.
Toxicokinetic data are consistent with these genotoxic metabolites either being delivered
to or produced in the kidney. As discussed in Chapter 3, following in vivo exposure to TCE, the
6/22/2009
446
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
metabolites DCVG, DCVC, and NAcDCVC have all been detected in the blood, kidney, or urine
of rats, and DCVG in blood and NAcDCVC in urine have been detected in humans (Birner et al.,
1993; Bernauer et al., 1996; Lash et al., 1999a, 2006). In addition, in vitro data have shown
DCVG formation from TCE in cellular and subcellular fractions from the liver, from which it
would be delivered to the kidney via systemic circulation, and from the kidney (see Table
3.3.11-3.3.12, and references therein). Furthermore, in vitro data in both humans and rodents
support the conclusion that DCVC is primarily formed from DCVG in the kidney itself, with
subsequent in situ transformation to NAcDCVC by N-Acetyl transferase or to reactive
metabolites by beta-lyase, FMO, or P450s (see Sections 3.3.3.2.2-3.3.3.2.5). Therefore it is
highly likely that both human and rodent kidneys are exposed to these TCE metabolites.
As discussed in section 4.1.1.4.2, DCVG, DCVC, and NAcDCVC have been
demonstrated to be genotoxic in most available in vitro assays. In particular, DCVC was
mutagenic in the Ames test in three of the tested strains of S. typhimurium (TA100, TA2638,
TA98) (Dekant et al., 1986; Vamvakas et al., 1988a), and caused dose-dependent increases in
unscheduled DNA synthesis in the two available assays: porcine kidney tubular epithelial cell
line (Vamvakas et al., 1996) and Syrian hamster embryo fibroblasts (Vamvakas et al., 1988b).
DCVC has also been shown to induce DNA strand breaks in both available studies (Jaffe et al.,
1985, Robbiano et al., 2004), and induce micronucleus formation in primary kidney cells from
rats and humans (Robbiano et al., 2004) but not in Syrian hamster embryo fibroblasts (Vamvakas
et al., 1988b). Only one study each is available for DCVG and N-AcDCVC, but notably both
were positive in the Ames test (Vamvakas et al., 1988a; Vamvakas et al., 1987). Although the
number of test systems was limited, these results are consistent.
These in vitro results are further supported by studies reporting kidney-specific
genotoxicity after in vivo administration of TCE or DCVC. In particular, Robbiano et al. (1998)
reported increased numbers of micronucleated cells in the rat kidney following oral TCE
exposure. Oral exposure to DCVC in both rabbits (Jaffe et al., 1985) and rats (Clay, 2008)
increased DNA strand breaks in the kidney. However, in one inhalation exposure study in rats,
TCE did not increase DNA breakage in the rat kidney, possibly due to study limitations [limited
exposure time (6h/day for only 5d) and small number of animals exposed (n = 5) (Clay, 2008)).
One study of TCE exposure in the Eker rat, a rat model heterozygous for the tumor suppressor
gene Tsc-2, reported no significant increase in kidney tumors as compared to controls (Mally et
al., 2006). Inactivation of Tsc-2 in this rat model is associated with spontaneous renal cell
carcinoma with activation of pathways similar to that of VHL inactivation in humans (Liu et al.,
2003). TCE exposure for 13-weeks (corn oil gavage) led to increased nephrotoxicity but no
significant increases in preneoplastic or neoplastic lesions as compared to controls (Mally et al.,
2006). This lack of increased incidence of neoplastic or preneoplastic lesions reported by Mally
6/22/2009
447
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
et al. (2006) in the tumor-prone Eker rat is similar to lack of significant short-term response
exhibited by other genotoxic carcinogens in the Eker rat (Morton et al., 2002; Stemmer et al.,
2007) and may be related to the increased background rate of renal carcinomas in this animal
model. Mally et al. (2006) also exposed primary kidney epithelial cells from the Eker rat to
DCVC in vitro and demonstrated increased transformation similar to that of other renal
carcinogens (Horesovsky et al., 1994). Similar to other genotoxic renal carcinogens analyzed by
NTP, there is limited evidence of mouse kidney tumors following TCE exposure. However,
given the already low incidences of kidney tumors observed in rats, a relatively small difference
in potency in mice would be undetectable in available chronic bioassays. In addition, limited,
mostly in vitro, toxicokinetic data do not suggest mice have less GSH conjugation or subsequent
renal metabolism/bioactivation (see Section 3.3.3.2.7). Notably, of seven chemicals categorized
as direct-acting genotoxic carcinogens that induced rat renal tumors in NTP studies, only two
also led to renal tumors in the mouse (tris[2,3-dibromopropyl]phosphate and ochratoxin A)
(Reznik et al., 1979; Kanisawa and Suzuki, 1978), so the lack of response in mouse bioassays
(albeit with low power) does not preclude a genotoxic MOA.
VHL inactivation (via mechanisms such as deletion, silencing or mutation) observed in
human renal clear cell carcinomas, is the basis of a hereditary syndrome of kidney cancer
predisposition, and is hypothesized to be an early and causative event in this disease (e.g.,
Nickerson et al., 2008). Therefore, specific actions of TCE metabolites that produce or select for
mutations of the VHL suppressor gene could lead to kidney tumorigenesis. Several studies have
compared VHL mutation frequencies in cases with TCE exposures with those from control or
background populations. Briining et al. (1997a) and Brauch et al. (1999, 2004) reported
differences between TCE-exposed and non-exposed renal cell carcinoma patients in the
frequency of somatic VHL mutations, the incidence of a hot spot mutation of cytosine to thymine
at nucleotide 454, and the incidence of multiple mutations. These data suggest that kidney tumor
genotype data in the form of a specific mutation pattern may potentially serve to discriminate
TCE-induced tumors from other types of kidney tumors in humans. If validated, this would also
suggest that TCE-induced kidney tumors are dissimilar from those occurring in unexposed
individuals. Thus, while not confirming a mutation MOA, these data suggest that TCE-induced
tumors may be distinct from those induced spontaneously in humans. However, it has not been
examined whether a possible linkage exists between VHL loss or silencing and mutagenic TCE
metabolites.
By contrast, Schraml et al. (1999) and Charbotel et al. (2007) reported that TCE-exposed
renal cell carcinoma patients did not have significantly higher incidences of VHL mutations
compared to non-exposed patients. However, details as to the exposure conditions were lacking
in Schraml et al. (1999). In addition, the sample preparation methodology employed by
6/22/2009
448
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Charbotel et al. (2007) and others (Briining et al., 1997a; Brauch et al., 1999) often results in
poor quality and/or low quantity DNA, leading to study limitations (less than 100% of samples
were able to be analyzed). Therefore, further investigations are necessary to either confirm or
contradict the validity of the genetic biomarkers for TCE-related renal tumors reported by
Briining et al. (1997a) and Brauch et al. (1999, 2004).
In addition, while exposure to mutagens is certainly associated with cancer induction (as
discussed with respect to the liver in Appendix E, Sections 3.1 and 3.2), examination of end-
stage tumor phenotype or genotype has limitations concerning determination of early key events.
The mutations that are observed with the progression of neoplasia are associated with increased
genetic instability and an increase in mutation rate. Further, inactivation of the VHL gene also
occurs through other mechanisms in addition to point mutations, such as loss of heterozygosity
or hypermethylation (Kenck et al., 1996; Nickerson et al., 2008) not addressed in these studies.
Recent studies examining the role of other genes or pathways suggest roles for multiple genes in
renal cell carcinoma development (Furge et al., 2007; Toma et al., 2008). Therefore, the
inconsistent results with respect to VHL mutation status do not constitute negative evidence for a
mutational MOA and the positive studies are suggestive of a TCE-induced kidney tumor
genotype.
In sum, the predominance of positive genotoxicity data in the database of available
studies of TCE metabolites derived from GSH conjugation (in particular the evidence of kidney-
specific genotoxicity following in vivo exposure to TCE or DCVC), coupled with the
toxicokinetic data consistent with the in situ formation of these GSH-conjugation metabolites of
TCE in the kidney, is consistent with a mutagenic MOA is operative in TCE-induced kidney
tumors. Available data on the VHL gene in humans adds biological plausibility to these
conclusions.
4.3.7.2 Hypothesized Mode of Action: Cytotoxicity and Regenerative Proliferation
Another hypothesis is that TCE acts by a cytotoxicity mode of action in TCE-induced
renal carcinogenesis. According to this hypothesis, the key events leading to TCE-induced
kidney tumor formation comprise the following: the TCE GSH-conjugation metabolite DCVC,
after being either produced in situ in or delivered systemically to the kidney, causes cytotoxicity,
leading to compensatory cellular proliferation and subsequently increased mutations and clonal
expansion of initiated cells.
Experimental Support for the Hypothesized Mode of Action
Evidence for the hypothesized MOA consist primarily of 1) the demonstration of
nephrotoxicity following TCE exposure at current occupational limits in human studies and
6/22/2009
449
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
chronic TCE exposure in animal studies; 2) the relatively high potential of the TCE metabolite
DCVC to cause nephrotoxicity; and 3) toxicokinetic data demonstrating that DCVC is formed in
the kidney following TCE exposure. Data on nephrotoxicity of TCE and DCVC are discussed in
more detail below, while the toxicokinetic data were summarized previously in the discussion of
mutagenicity. However, there is a lack of experimental support linking TCE nephrotoxicity and
sustained cellular proliferation to TCE-induced nephrocarcinogenicity.
There is substantial evidence that TCE is nephrotoxic in humans and laboratory animals
and that its metabolite DCVC is nephrotoxic in laboratory animals. Epidemiological studies
have consistently demonstrated increased excretion of nephrotoxicity markers (NAG, protein,
albumin) at occupational (Green et al., 2004) and higher (Bolt et al., 2004; Briining et al.,
1999a, b) levels of TCE exposure. However, direct evidence of tubular toxicity, particularly in
renal cell carcinoma cases, is not available. These studies are supported by the results of
multiple laboratory animal studies. Chronic bioassays have reported very high (nearly 100%)
incidences of nephrotoxicity of the proximal tubule in rats (NTP, 1988, 1990) and mice (NCI,
1976; NTP, 1990) at the highest doses tested. In vivo studies examining the effect of TCE
exposure on nephrotoxicity showed increased proximal tubule damage following intraperitoneal
injection and inhalation of TCE in rats (Chakrabarty and Tuchweber, 1988) and intraperitoneal
injection in mice (Cojocel et al., 1989). Studies examining DCVC exposure in rats (Terracini
and Parker, 1965; Elfarra et al., 1986) and mice (Jaffe et al., 1984; Darnerud et al., 1989) have
also shown increases in kidney toxicity. The greater potency for kidney cytotoxicity for DCVC
compared to TCE was shown by in vitro studies (Lash et al., 1995, 1986, Stevens et al., 1986).
These studies also further confirmed the higher susceptibility of male rats or mice to DCVC-
induced cytotoxicity. Cytokaryomegaly (an effect specific to TCE and not part of the chronic
progressive nephropathy or the pathology that occurs in aging rat kidneys) was observed in the
majority of rodent studies and may or may not progress to carcinogenesis. Finally, as discussed
extensively in Section 4.3.6.1, a detailed comparison of the histological changes in the kidney
caused by TCE and its metabolites supports the conclusion that DCVC is the predominant
moiety responsible for TCE-induced nephrotoxicity.
Because it is known that not all cytotoxins are carcinogens (i.e., cytotoxicity is not a
specific predictor of carcinogenicity), additional experimental support is required to link
nephrotoxicity to carcinogenicity. Clearly, cytotoxicity occurs at doses below those causing
carcinogenicity, as the incidence of nephrotoxicity in chronic bioassays is an order of magnitude
higher than that of renal tumors. However, there are multiple mechanisms by which TCE has
been hypothesized to induce cytotoxicity, including oxidative stress, disturbances in calcium ion
homeostasis, mitochondrial dysfunction, and protein alkylation (Lash et al., 2000a). Some of
these effects may therefore have ancillary consequences related to tumor induction which are
6/22/2009
450
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
independent of cytotoxicity per se. Under the hypothesized MO A, cytotoxicity leads to the
induction of repair processes and compensatory proliferation that could lead to an increased
production or clonal expansion of cells previously initiated by mutations occurred spontaneously,
from co-exposures, or from TCE or its metabolites. Data on compensatory cellular proliferation
and the subsequent hypothesized key events in the kidney are few, with no data from rat strains
used in chronic bioassays. In rats carrying the Eker mutation, Mally et al. (2006) reported
increased DNA synthesis as measured by BrdU incorporation in animals exposed to the high
dose of TCE (1,000 mg/kg-d) for 13 weeks, but there was no evidence of clonal expansion or
tumorigenesis in the form of increased preneoplastic or neoplastic lesions as compared to
controls. While chronic nephrotoxicity was reported in the same bioassays showing increased
kidney tumor incidences, the use of such data to inform MOA is indirect and associative.
Moreover, chronic animal studies with reduced (in female rats) or absent (in mice of both sexes)
carcinogenic response have also demonstrated cytotoxicity (NTP, 1990, NCI, 1976). Therefore,
in both rodent and human studies of TCE, data demonstrating a causal link between tubular
toxicity and the induction of kidney tumors are lacking.
4.3.7.3 Additional Hypothesized Modes of Action with Limited Evidence or Inadequate
Experimental Support
Along with metabolites derived from GSH conjugation of TCE, oxidative metabolites are
also present and could induce toxicity in the kidney. After TCE exposure, the oxidative
metabolite and peroxisome proliferator TCA is present in the kidney and excreted in the urine as
a biomarker of exposure. Hypotheses have also been generated regarding the roles of o.2\i-
globulin or formic acid in nephrotoxicity induced by TCE oxidative metabolites TCA or TCOH.
4.3.7.3.1 Peroxisome proliferation
Although not as well studied as the effects of glutathione metabolites in the kidney, there
is evidence that oxidative metabolites affect the kidney after TCE exposure. Both TCA and
DCA are PPARa agonists although most activity has been associated with TCA production after
TCE exposure. Exposure to TCE has been found to induce peroxisome proliferation not only in
the liver but also the kidney. Peroxisome proliferation in the kidney has been evaluated by only
one study of TCE (Goldsworthy and Popp, 1987), using increases in cyanide-insensitive
palmitoyl-CoA oxidation (PCO) activity as a marker. Increases in renal PCO activity were
observed in rats (3.0-fold) and mice (3.6-fold) treated with TCE at 1,000 mg/kg-d for 10 days,
with smaller increases in both species from TCA treatment at 500 mg/kg-d for 10 days.
However, no significant increases in kidney/body weight ratios were observed in either species.
There was no relationship between induction of renal peroxisome proliferation and renal tumors
6/22/2009
451
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(i.e., a similar extent of peroxisome proliferation-associated enzyme activity occurred in species
with and without TCE-induced renal tumors). However, the increased peroxisomal enzyme
activities due to TCE exposure are indicative of oxidative metabolites being present and
affecting the kidney. Such metabolites have been associated with other tumor types, especially
liver, and whether coexposures to oxidative metabolites and glutathione metabolites contribute to
kidney tumorigenicity has not been examined.
4.3.7.3.2 a2fi-globulin-related nephropathy
Induction of a.2|i-globulin nephropathy by TCE has been investigated by Goldsworthy et
al. (1988), who reported that TCE did not induce increases in this urinary protein, nor did it
stimulate cellular proliferation in rats. In addition, whereas kidney tumors associated with o.2\i-
globulin nephropathy are specific to the male rat, as discussed above, nephrotoxicity is observed
in both rats and mice and kidney tumor incidence is elevated (though not always statistically
significant) in both male and female rats. TCOH was recently reported to cause hyaline droplet
accumulation and an increase in a.2|i-globulin, but these levels were insufficient to account for
the observed nephropathy as compared to other exposures (Green et al., 2003b). Therefore, it is
unlikely that a.2|i-globulin nephropathy contributes significantly to TCE-induced renal
carcinogenesis.
4.3.7.3.3 Formic acid-related nephrotoxicity
Another MOA hypothesis proposes that TCE nephrotoxicity is mediated by increased
formation and urinary excretion of formic acid mediated by the oxidative metabolites TCA or
TCOH (Green et al., 1998, 2003; Dow and Green, 2000). The subsequent hypothesized key
events are the same as those for DCVC-induced cytotoxicity, discussed above (Section 4.3.7.2).
As discussed extensively in Section 4.3.6.1.2, these oxidative metabolites do not appear
sufficient to explain the range of renal effects observed after TCE exposure, particularly
cytomegaly, karyomegaly, and flattening and dilation of the tubular epithelium. Although
TCOH and possibly TCA may contribute to the nephrotoxicity of TCE, perhaps due to excess
formic acid production, these metabolites do not show the same range of cytotoxic effects
observed following TCE exposure (Table 4.3.11). Therefore, without specific evidence linking
the specific nephrotoxic effects caused by TCOH or TCA to carcinogenesis, and in light of the
substantial evidence that DCVC itself can adequately account for the nephrotoxic effects of TCE,
the weight of evidence supports a conclusion that cytotoxicity mediated by increased formic acid
production induced by oxidative metabolites TCOH and possibly TCA is not responsible for the
majority of the TCE-induced cytotoxicity in the kidneys, and therefore would not be the major
contributor to the other hypothesized key events in this MOA, such as subsequent regenerative
proliferation.
6/22/2009
452
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.3.7.4 Conclusions about the Hypothesized Modes of Action
1. Is the hypothesized mode of action sufficiently supported in the test animals?
Mutagenicity: The predominance of positive genotoxicity data in the database of available
studies of TCE metabolites derived from GSH conjugation (in particular the evidence of kidney-
specific genotoxicity following in vivo exposure to TCE or DCVC), coupled with the
toxicokinetic data consistent with the in situ formation of these GSH-conjugation metabolites of
TCE in the kidney, supports the conclusion that a mutagenic MOA is operative in TCE-induced
kidney tumors.
Cytotoxicity: As reviewed above, in vivo and in vitro studies have shown a consistent
nephrotoxic response to TCE and its metabolites in proximal tubule cells from male rats.
Therefore, it has been proposed that cytotoxicity seen in this region of the kidney is a precursor
to carcinogenicity. However, it has not been determined whether tubular toxicity is a necessary
precursor of carcinogenesis, and there is a lack of experimental support for causal links, such as
compensatory cellular proliferation or clonal expansion of initiated cells, between nephrotoxicity
and kidney tumors induced by TCE. Nephrotoxicity is evidently not sufficient in and of itself, as
mice exhibit a similar nephrotoxic response without an increase in kidney tumors, and an
explanation for this species difference has not been found.
Additional hypotheses: The kidney is also exposed to oxidative metabolites that have been
shown to be carcinogenic in other target organs. TCA is excreted in kidney after its metabolism
from TCE and also can cause peroxisome proliferation in the kidney, but there are inadequate
data to define a MOA for kidney tumor induction based on peroxisome proliferation. TCE
induced little or no a2|i-globulin and hyaline droplet accumulation to account for the observed
nephropathy, so available data do not support this hypothesized MOA. The production of formic
acid following exposure to TCE and its oxidative metabolites TCOH and TCA may also
contribute to nephrotoxicity; however, the available data indicate that TCOH and TCA are minor
contributors to TCE-induced nephrotoxicity, and therefore do not support this hypothesized
MOA. Because these additional MOA hypotheses are either inadequately defined or are not
supported by the available data, they are not considered further in the conclusions below.
2. Is the hypothesized mode of action relevant to humans?
Mutagenicity: The evidence discussed above demonstrates that TCE GSH-conjugation
metabolites are mutagens in microbial as well as test animal species. Therefore the presumption
that they would be mutagenic in humans. Available data on the VHL gene in humans adds
biological plausibility to this hypothesis. The few available data from human studies concerning
6/22/2009
453
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
the mutagenicity of TCE and its metabolites suggest consistency with this MO A, but are not
sufficiently conclusive to provide direct supporting evidence for a mutagenic MOA. Therefore,
this MOA is considered relevant to humans.
Cytotoxicity: Although data are inadequate to determine that the MOA is operative, none of the
available data suggest that this MOA is biologically precluded in humans. Furthermore, both
animal and human studies suggest that TCE causes nephrotoxicity at exposures that also induce
renal cancer, constituting positive evidence of the human relevance of this hypothesized MOA.
3. Which populations or lifestages can be particularly susceptible to the hypothesized mode of
action?
Mutagenicity: The mutagenic MOA is considered relevant to all populations and lifestages.
According to EPA's Cancer Guidelines (U.S. EPA, 2005a) and Supplemental Guidance (U.S.
EPA, 2005b), there may be increased susceptibility to early-life exposures for carcinogens with a
mutagenic mode of action. Therefore, because the weight of evidence supports a mutagenic
mode of action for TCE carcinogenicity and in the absence of chemical-specific data to evaluate
differences in susceptibility, early-life susceptibility should be assumed and the age-dependent
adjustment factors (ADAFs) should be applied, in accordance with the Supplemental Guidance.
In addition, because the MOA begins with GSH-conjugation metabolites being delivered
systemically or produced in situ in the kidney, toxicokinetic differences - i.e., increased
production or bioactivation of these metabolites - may render some individuals more susceptible
to this MOA. Toxicokinetic-based susceptibility is discussed further in Section 4.9.
In rat chronic bioassays, TCE-treated males have higher incidence of kidney tumors than
similarly treated females. However, the basis for this sex-difference is unknown, and whether it
is indicative of a sex difference in human susceptibility to TCE-induced kidney tumors is
likewise unknown. The epidemiologic studies generally do not show sex differences in kidney
cancer risk. Lacking exposure-response information, it is not known if the sex-difference in one
renal cell carcinoma case-control study (Dosemeci et al., 1999) may reflect exposure differences
or susceptibility differences.
Cytotoxicity: Populations which may be more susceptible based on the toxicokinetics of the
production of GSH conjugation metabolites and the sex differences observed in rat chronic
bioassays are the same as for a mutagenic MOA. No data are available as to whether other
factors may lead to different populations or lifestages being more susceptible to a cytotoxic
MOA for TCE-induced kidney tumors. For instance, it is not known how the hypothesized key
events in this MOA interact with known risk factors for human renal cell carcinoma.
6/22/2009
454
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
The weight of evidence sufficiently supports a mutagenic MOA for TCE in the kidney,
based on supporting data that GSH-metabolites are genotoxic and produced in sufficient
quantities in the kidney to lead to tumorigenesis. Cytotoxicity and regenerative proliferation
were considered as an alternate MOA, however, there is inadequate data to support a causal
association between cytotoxicity and kidney tumors. Further, hypothesized MO As relating to
peroxisomal proliferation, a2[j,-globulin nephropathy and formic acid-related nephrotoxicity
were considered and rejected due to limited evidence and/or inadequate experimental support.
4.3.8 Summary: TCE kidney toxicity, carcinogenicity, and mode-of-action
Human studies have shown increased levels of proximal tubule damage in workers
exposed to high levels of TCE (NRC, 2006). These studies analyzed workers exposed to TCE
alone or in mixtures and reported increases in various urinary biomarkers of kidney toxicity (P2-
microglublin, total protein, NAG, a 1-microglobulin) (Nagaya et al., 1989; Selden et al., 1993;
Briining et al. 1999a, b; Bolt et al., 2004; Green et al., 2004; Radican et al., 2006). Laboratory
animal studies examining TCE exposure provide additional support, as multiple studies by both
gavage and inhalation exposure show that TCE causes renal toxicity in the form of cytomegaly
and karyomegaly of the renal tubules in male and female rats and mice. By gavage, incidences
of these effects under chronic bioassay conditions approach 100%, with male rats appearing to
be more sensitive than either female rats or mice of either sex based on the severity of effects.
Under chronic inhalation exposures, only male rats exhibited these effects. Further studies with
TCE metabolites have demonstrated a potential role for DCVC, TCOH and TCA in TCE-
induced nephrotoxicity. Of these, DCVC induces the renal effects that are most like TCE, and it
is formed in sufficient amounts following TCE exposure to account for these effects.
Kidney cancer risk from TCE exposure has been studied related to TCE exposure in
cohort, case-control and geographical studies. These studies have examined TCE in mixed
exposures as well as alone. Elevated risks are observed in many of the cohort and case-control
studies examining kidney cancer incidence in industries or job titles with historical use of TCE
(Table 4.3-3, 4.3-4). Although there are some controversies related to deficiencies of the
epidemiological studies (Vamvakas et al., 1998; Henschler et al., 1995), many of these are
overcome in later studies (Briining et al., 2003; Charbotel et al., 2006). A meta-analysis of the
overall effect of TCE exposure on kidney cancer suggests a small, statistically significant
increase in risk (pooled RR = 1.26 95% CI: 1.11, 1.42) with a pooled relative risk estimate in the
higher exposure group of 1.61, (95% CI: 1.27, 2.03). In vivo laboratory animal studies to date
suggesting a small increase in renal tubule tumors in male rats and, to a lesser extent, in female
rats, with no increases seen in mice or hamsters. These results are based on limited studies of
6/22/2009
455
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
both oral and inhalation routes, some of which were deemed insufficient to determine
carcinogenicity based on various experimental issues. However, because of the rarity of renal
clear cell carcinomas in rodents, the repeatability of this finding across strains and studies
supports their biological significance.
Some but not all human studies have suggested a role for VHL mutations in TCE-induced
kidney cancer (Briining et al., 1997a; Brauch et al., 1999, 2004; Schraml et al., 1999; Charbotel
et al., 2007). Certain aspects of these studies may explain some of these discrepant results. The
majority of these studies have examined paraffinized tissue that may lead to technical difficulties
in analysis, as paraffin extractions yield small quantities of often low-quality DNA. The
chemicals used in the extraction process itself may also interfere with enzymes required for
further analysis (PCR, sequencing). Although these studies do not clearly show mutations in all
TCE-exposed individuals, or in fact in all kidney tumors examined, this does not take into
account other possible means of VHL inactivation, including silencing or loss, and other potential
targets of TCE mutagenesis were not systematically examined. A recent study by Nickerson et
al. (2008) analyzed both somatic mutation and promoter hypermethylation of the VHL gene in
cc-RCC frozen tissue samples using more sensitive methods. The results of this study support
the hypothesis that VHL alterations are an early event in clear cell RCC carcinogenesis, but these
alterations may not be gene mutations. No experimental animal studies have been performed
examining vhl inactivation following exposure to TCE, although one in vitro study examined vhl
mutation status following exposure to the TCE-metabolite DCVC (Mally et al., 2006). This
study found no mutations following DCVC exposure, although this does not rule out a role for
DCVC in vhl inactivation by some other method or vhl alterations caused by other TCE
metabolites.
Although not encompassing all of the actions of TCE and its metabolites that may be
involved in the formation and progression of neoplasia, available evidence supports the
conclusion that a mutagenic MOA mediated by the TCE GSH-conjugation metabolites
(predominantly DCVC) is operative in TCE-induced kidney cancer. This conclusion is based on
substantial evidence that these metabolites are genotoxic and are delivered to or produced in the
kidney, including evidence of kidney-specific genotoxicity following in vivo exposure to TCE or
DCVC. Cytotoxicity caused by DCVC leading to compensatory cellular proliferation is also a
potential MOA in renal carcinogenesis, but available evidence is inadequate to conclude that this
MOA is operative, either together with or independent of a mutagenic MOA. The additional
MOA hypotheses of peroxisome proliferation, accumulation of a2[j.-globulin, and cytotoxicity
mediated by TCE-induced excess formic acid production are not supported by the available data.
6/22/2009
456
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 References
2 Andersson L, Heraldsson B, Johansson C, Barregard L. 2008. Methodological issues on the use
3 of urinary alpha-1-microglobulin in epidemiological studies. Nephrol Dial Transplant
4 23:1252-1256.
5 Anttila A, Pukkala E, Sallmen M, Hernberg S, Hemminki K. 1995. Cancer incidence among
6 Finnish workers exposed to halogenated hydrocarbons. J Occup Environ Med 37:797-806.
7 Asal NR, Geyer JR, Risser DR, Lee ET, Kadamani S, Cherng N. 1988. Risk factors in renal cell
8 carcinoma. II. Medical history, occupation, multivariate analysis and conclusions. Cancer
9 Detect Prev 13:263-279.
10 Auperin A, Asal NR, Geyer JR, Risser DR, Lee ET, Kadamani S, Cherng N. 1988 Risk factors
11 in renal cell carcinoma. II. Medical history, occupation, multivariate analysis, and conclusions.
12 Cancer Detect Preven 13:263-279.
13 ATSDR (Agency for Toxic Substance and Disease Registry). 2004. Feasibility Investigation of
14 Worker Exposure to Trichloroethylene at the View-Master Factory in Beaverton, Oregon.
15 Prepared by the Oregon Department of Human Services. Report No. PB2005-102173.
16 ATSDR(Agency for Toxic Substances and Disease Registry). 2006. Health Consultation.
17 Cancer and Birth Outcome Analysis, Endicott Area, Town of Union, Broome County, New
18 York. Health Statistics Review. Atlanta, GA: U.S. Department of Health and Human Services,
19 Public Health Service, Agency for Toxic Substances and Disease Registry. May 26, 2006.
20 ATSDR. (2008) Health statistics review follow-up: cancer and birth outcome analysis. Endicott
21 Area Investigation, Endicott Area, Town of Union, Broome County, New York. U.S.
22 Department of Health and Human Services, May 2008, Atlanta, GA.
23 Auperin A, Benhamou S, Ory-Paoletti C, Flamant R. 1994. Occupational risk factors for renal
24 cell carcinoma: a case-control study. Occup Environ Med 51:426-428
25 Axelson O, Selden A, Andersson K, Hogstedt C. 1994. Updated and expanded Swedish cohort
26 study on trichloroethylene and cancer risk. J Occup Environ 36:556-562.
27 Bailly M, Bain C, Favrot M-C, Ozturk M. 1995. Somatic mutations of von Hippel-Lindau
28 (VHL) tumor-suppressor gene in European kidney cancer. Int J Cancer 63:660-664.
6/22/2009
457
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Bakke B, Stewart PA, Waters MA. 2007. uses of and exposure to trichloroethylene in U.S.
2 Industry: a systematic literature review. J Occup Environ Hgy 4:375-390.
3 Bernauer, U., Birner, G., Dekant, W., and Henschler, D. (1996). Biotransformation of
4 trichloroethene: dose-dependent excretion of 2,2,2-trichloro-metabolites and mercapturic acids in
5 rats and humans after inhalation. Arch Toxicol 70, 338-346.
6 Beroud C, Joly D, Gallou C, Staroz F, Orfanelli MT, Junien C. 1998. Software and database for
7 the analysis of mutation in the VHL gene. Nucleic Acids Research 26:256-258.
8 Birner G, Vamvakas S, Dekant W, Henschler D. (1993). Nephrotoxic and genotoxic N-acetyl-S-
9 dichlorovinyl-L-cysteine is a urinary metabolite after occupational 1,1,2-trichloroethene
10 exposure in humans: implications for the risk of trichloroethene exposure. Environ Health
11 Perspect. 99:281-4.
12 Blair A, Hartge P, Stewart PA, McAdams M, Lubin J. 1998. Mortality and cancer incidence of
13 aircraft maintenance workers exposed to trichloroethylene and other organic solvents and
14 chemicals: extended follow up. Occup Environ Med 55:161-171.
15 Blair A, Stewart P, Lubin JH, Forastiere F. 2007. Methodological issues regarding confounding
16 and exposure misclassification in epidemiological studies of occupational exposures. Am J Ind
17 Med 50:199-207.
18 Bloemen L, Tomenson J. 1995. Increased incidence of renal cell tumours in a cohort of
19 cardboard workers exposed to trichloroethylene [Letter], Arch Toxicol 70:129-130.
20 Boice JD, Marano DE, Fryzek JP, Sadler CJ, McLaughlin JK. 1999. Mortality among aircraft
21 manufacturing workers. Occup Environ Med 56:581-597.
22 Boice JD, Marano DE, Cohen SS, Mumma MR, Blott WJ, Brill AB, Fryzek JP, Henderson BE,
23 McLaughlin JK. 2006. Mortality among Rocketdyne worker who tested rocket engines, 1948-
24 1999. J Occup Environ Med 48:1070-1092.
25 Bolt H, Lammert M, Selinski S, Briining T. 2004. Urinary al-microglobulin excretion as
26 biomarker of renal toxicity in trichloroethylene-exposed persons. Int Arch Occup Environ
27 Health 77:186-190.
6/22/2009
458
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Brauch H, Weirich G, Hornauer MA, Storkel S, Wohl T, Briining T. 1999. Trichloroethylene
2 exposure and specific somatic mutations in patients with renal cell carcinoma. J Natl Cancer Inst
3 91:854:861.
4 Brauch H, Weirich G, Klein B, Rabstein S, Bolt HM, Briining T. 2004. VHL mutations in renal
5 cell cancer: does occupational exposure to trichloroethylene make a difference? Toxicol Lett
6 151:301-310.
7 BrownsonRC. 1988. A case-control study of renal cell carcinoma in relation to occupation,
8 smoking, and alcohol consumption. Arch Environ Health 43:238-241.
9 Briining T, Weirich G, Hornauer MA, Hofler H, Brauch H. 1997a. Renal cell carcinomas in
10 trichloroethene TRI) exposed persons are associated with somatic mutations in the von Hippel-
11 Lindau {VHL) tumour suppressor gene. Arch Toxicol 71:332-335.
12 Briining T, Lammert M, Kempkes M, Their R, Golka K, Bolt HM. 1997b. Influence of
13 polymorphisms of GSTM1 and GSTT2 For risk of renal cell cancer in workers with long-term
14 high occupational exposure to trichloroethene. Arch Toxicol 71:596-599.
15 Briining T, Mann H, Melzer H, Sundberg AGM, Bolt HM. 1999a. Pathological excretion
16 patterns of urinary proteins in renal cell cancer patients exposed to trichloroethylene. Occup
17 Med. 29:299-305.
18 Briining T, Sundberg AGM, Birner G, Lammert M, Bolt HM, Appelkvist E-L, Nilsson R,
19 Dallner G. 1999b. Glutathione transferase alpha as a marker for tubular damage after
20 trichloroethylene exposure. Arch Toxicol 73:246-254.
21 Briining T, Pesch B, Wiesenhiitter B, Rabstein S, Lammert M, Baumiiller A, Bolt HM. 2003.
22 Renal cell cancer risk and occupational exposure to trichloroethylene: results of a consecutive
23 case-control study in Arnsberg, Germany. Am J Ind Med. 43:274-85.
24 Carried M, Magosso D, Piccoli P, Zanetti E, Trevisan A, Bartolucci GB. 2007. Acute nonfatal
25 intoxication with trichloroethylene. Arch Toxicol 81:529-32.
26 Chakrabarti, SK; Tuchweber, B. (1988) Studies of acute nephrotoxic potential of
27 trichloroethylene in Fischer 344 rats. J Toxicol Environ Health 23:147-158.
6/22/2009
459
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Chang Y-M, Tai C-F, Yang S-C, Chan C-J, S Shin T-S, Lin RS, Liou S-H. 2003. A cohort
2 mortality study of workers exposed to chlorinated organic solvents in Taiwan. Ann Epidemiol
3 13:652-660.
4 Chang Y-M, Tai C-F, Yang S-C, Lin R, Sung F-C, Shin T-S, Liou S-H. 2005. Cancer Incidence
5 among Workers Potentially Exposed to Chlorinated Solvents in An Electronics Factory. J Occup
6 Health 47:171-180.
7 Charbotel B, Fevotte J, Hours M, Martin JL, Bergeret A. 2005. Case-control study on renal cell
8 cancer and occupational trichloroethylene exposure, in the Arve valley (France). Final report.
9 Submitted to U.S. EPA by Paul Dugard, Ph.D., Halogenated Solvents Industry Alliance.
10 #8EHQ-1105-16030.
11 Charbotel B, Fevotte J, Hours M, Martin J-L, Bergeret A. 2006. Case-control study on renal
12 cell cancer and occupational exposure to trichloroethylene. Part II: epidemiological aspects.
13 Ann Occup Hyg 50:777-787.
14 Charbotel B, Gad S, Caiola D, Beroud C, Fevotte J, et al. 2007. Trichloroethylene exposure and
15 somatic mutations of the VHL gene in patients with renal cell carcinoma. J Occup Med Toxicol
16 2:2-13.
17 Charbotel B, Fevotte J, martin JL, Bergeret A. 2009. Cancer du rein et expositions au
18 trichloroethylene: les valeurs limites d'exposition professionnelle fra9aises en vigueur sont-elles
19 adaptees. Rev Epidemiol Sante Publique 57:41-47.
20 Chen Y, Cai J, Anders MW, Stevens JL, Jones DP. 2001. Role of mitochondrial dysfunction in
21 S-(l,2-dichlorovinyl)-L-cysteine-induced apoptosis. Toxicology and Applied Pharmacology
22 170:172-180.
23 Cherrie JW, Kromhout H, Semple S. 2001. The importance of reliable exposure estimates in
24 deciding whether trichloroethylene can cause kidney cancer. J Cancer Res Clin Oncol
25 127:400-404.
26 Chow WH, Gridley G, McLaughlin JK, Mandel JS, Wacholder S, Blot WJ, Niwa S, Fraumeni
27 JF. 1994. Protein intake and risk of renal cell cancer. J of National Cancer Institute
28 86:1131-1139.
29 Clapp RW, Hoffman K. 2008. Cancer mortality in IBM Endicott plant workers, 1969-2001: an
30 update on a NY production plant. Environ Health 7:13.
6/22/2009
460
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Clay P. 2008. 2008. Assessment of the genotoxicity of trichloroethylene and its metabolite, S-
2 (l,2-dichlorovinyl)-L-cysteine (DCVC), in the comet assay in rat kidney. Mutagenesis 23:27-33.
3 Cojocel, C; Beuter, W; Muller, W; et al. (1989) Lipid peroxidation: a possible mechanism of
4 trichloroethylene-induced nephrotoxicity. Toxicology 55:131-141.
5 Costa G, Merletti F, Segnan N. 1989. A mortality cohort study in a north Italian aircraft factory.
6 Br J Ind Med 46:738-743.
7 Cummings, BS; Lash, LH. (2000) Metabolism and toxicity of trichloroethylene and S-(l,2-
8 dichlorovinyl)-L-cysteine in freshly isolated human proximal tubular cells. Toxicol Sci
9 53:458-466.
10 Cummings, BS; Lasker, JM; Lash, LH. (2000a) Expression of glutathione-dependent enzymes
11 and cytochrome P450s in freshly isolated and primary cultures of proximal tubular cells from
12 human kidney. J Pharmacol Exp Ther 293:677-685.
13 Cummings, BS; Zangar, RC; Novak, RF; et al. (2000b) Cytotoxicity of trichloroethylene and S-
14 (1, 2-dichlorovinyl)-L-cysteine in primary cultures of rat renal proximal tubular and distal
15 tubular cells. Toxicology 150:83-98.
16 Darnerud, P; Brandt, I; Feil, V; et al. (1989) Dichlorovinyl cysteine (DCVC) in the mouse
17 kidney: tissue-binding and toxicity after glutathione depletion and probenecid treatment. Arch
18 Toxicol 63:345-350.
19 DeAngelo, AB; Daniel, FB; Most, BM; et al. (1997) Failure of monochloroacetic acid and
20 trichloroacetic acid administered in the drinking water to produce liver cancer in male F344/N
21 rats. J Toxicol Environ Health 52:425-445.
22 de Jong PE, Brenner BM. 2004. From secondary to primary prevention of progressive renal
23 disease: the case for screening for albuminuria. Kidney Int. 66(6):2109-18.
24 Dekant W, Vamvakas S, Berthold K, Schmidt S, Wild D, Henschler D. 1986. Bacterial beta-
25 lyase mediated cleavage and mutagenicity of cysteine conjugates derived from the
26 nephrocarcinogenic alkenes trichloroethylene, tetrachloroethylene and hexachlorobutadiene.
27 Chem Biol Interact 60:31-45.
28 Delahunt B, Bethwaite PB, Nacey JN. 1995. Occupational risk for renal cell carcinoma: A case-
29 control study based on the New Zealand Cancer Registry. Br J Urol 75:578-582.
6/22/2009
461
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Dosemeci M, Cocco P, Chow WH. 1999. Gender differences in risk of renal cell carcinoma and
2 occupational exposures to chlorinated aliphatic hydrocarbons. Am J Ind Med 36:54-59.
3 Dow, JL; Green, T. (2000) Trichloroethylene induced vitamin B(12) and folate deficiency leads
4 to increased formic acid excretion in the rat. Toxicology 146:123-136.
5 Elfarra, AA; Anders, MW. (1984) Renal processing of glutathione conjugates. Role in
6 nephrotoxicity. Biochem Pharmacol 33:3729-3732.
7 Elfarra AA, Jakobson I, Anders MW. 1986. Mechanism of S-(l,2-dichlorovinyl)glutathione-
8 induced nephrotoxicity. Biochem Pharmacol 35:283-288.
9 Environmental Health Strategies. 1997. Final Report. Correspondence from Paul A. Cammer,
10 Ph.D., Trichloroethylene Issues Group, to Cheryl Siegel Scott, US Environmental Protection
11 Agency. December 22, 1997.
12 Eyre, RJ; Stevens, DK; Parker, JC; et al. (1995a) Renal activation of trichloroethene and S-(l,2-
13 dichlorovinyl)-L-cysteine and cell proliferative responses in the kidneys of F344 rats and
14 B6C3F1 mice. J Toxicol Environ Health 46:465-481.
15 Eyre, RJ; Stevens, DK; Parker, JC; et al. (1995b) Acid-labile adducts to protein can be used as
16 indicators of the cysteine S-conjugate pathway of trichloroethene metabolism. J Toxicol Environ
17 Health 46:443-464.
18 Fukuda, K; Takemoto, K; Tsuruta, H. (1983) Inhalation carcinogenicity of trichloroethylene in
19 mice and rats. Ind Health 21:243-254.
20 Furge KA, Dykema K, Petillo D, Westphal M, Zhang Z, Kort EJ, Teh BT. 2007. Combining
21 differential expression, chromosomal and pathway analyses for the molecular characterization of
22 renal cell carcinoma. Canadian Urological Association Journal (CUAJ) 1 (2 Suppl): S21-S27.
23 Gallou C, Longuemaux S, Delomenie C, Mejean A, Martin N, Martinet S, Palais G, et al. 2001.
24 Association of GSTT1 non-null and NAT1 slow/rapid genotypes with von Hippel-Lindau
25 tumour suppressor gene tranversions in sporadic renal cell carcinoma. Pharmacogenetics
26 11:521-535.
27 Garabrant DH, Held J, Langholz B, Bernstein L. 1988. Mortality of Aircraft Manufacturing
28 Workers in Southern California. Am J Ind Med 13:683-693.
6/22/2009
462
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Goldsworthy, TL; Popp, JA. (1987) Chlorinated hydrocarbon-induced peroxisomal enzyme
2 activity in relation to species and organ carcinogenicity. Toxicol Appl Pharmacol 88:225-233.
3 Goldsworthy TL, Lyght O, Burnett VL, Popp JA. 1988. Potential role of a-2[j,-globulin, protein
4 droplet accumulation, and cell replication in the renal carcinogenicity of rats exposed to
5 trichloroethylene, perchloroethylene, and pentachloroethane. Toxicology and Applied
6 Pharmacology 96:367-379.
7 Green LC, Lash TL. 1999. Re: "Renal cell cancer correlated with occupational exposure to
8 trichloroethylene". J Cancer Res Clin Oncol 125:430-432.
9 Green, T; Prout, MS. (1985) Species differences in response to trichloroethylene. II.
10 Biotransformation in rats and mice. Toxicol Appl Pharmacol 79:401-411.
11 Green, T; Dow, J; Ellis, MK; et al. (1997) The role of glutathione conjugation in the
12 development of kidney tumours in rats exposed to trichloroethylene. Chem Biol Interact
13 105:99-117.
14 Green, T; Dow, J; Foster, JR; et al. (1998) Formic acid excretion in rats exposed to
15 trichloroethylene: a possible explanation for renal toxicity in long-term studies. Toxicology
16 127:39-47.
17 Green, T; Dow, J; Foster, J. (2003) Increased formic acid excretion and the development of
18 kidney toxicity in rats following chronic dosing with trichloroethanol, a major metabolite of
19 trichloroethylene. Toxicology 191:109-119.
20 Green T, Dow J, Ong CN, Ng V, Ong HY, Zhuang ZX, Yang XF, Bloemen L. 2004. Biological
21 monitoring of kidney function among workers occupationally exposed to trichloroethylene.
22 Occup Environ Med 61:312-317.
23 Greenland S, Salvan A, Wegman DH, Hallock MF, Smith TH. 1994. A case-control study of
24 cancer mortality at the transformer-assembly facility. Int Arch Occup Environ Health 66:49-54.
25 Groves, CE; Hayden, PJ; Lock, EA; et al. (1993) Differential cellular effects in the toxicity of
26 haloalkene and haloalkane cysteine conjugates to rabbit renal proximal tubules. J Biochem
27 Toxicol 8:49-56.
6/22/2009
463
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Hansen J, Raaschou-Nielsen O, Christensen JM, Johansen I, McLaughlin JK, Lipworth L, Blot
2 WJ, Olsen JH. 2001. Cancer incidence among Danish workers exposed to trichloroethylene. J
3 Occup Environ Med 43:133-139.
4 Harrington JM, Whitby H, Gray CN, Reid FJ, Aw TC, Waterhouse JA. 1989. Renal disease and
5 occupational exposure to organic solvents: a case referent approach. Br J Ind Med 46:643-650.
6 Hassall, CD; Brendel, K; Gandolfi, AJ. (1983) Regulation of a S(trans-l,2-dichlorovinyl)-L-
7 cysteine-induced renal tubular toxicity by glutathione. J Appl Toxicol 3:321-325.
8 Henschler, D; Romen, W; Elsasser, HM; et al. (1980) Carcinogenicity study of trichloroethylene
9 by long-term inhalation in three animal species. Arch Toxicol 43:237-248.
10 Henschler, D; Elsasser, H; Romen, W; et al. (1984) Carcinogenicity study of trichloroethylene,
11 with and without epoxide stabilizers, in mice. J Cancer Res Clin Oncol 107:149-156.
12 Henschler D, Vamvakas S, Lammert M, Dekant W, Kraus B, Thomas B, Ulm K. 1995.
13 Increased incidence of renal cell tumors in a cohort of cardboard workers exposed to
14 trichloroethene. Arch Toxicol 69:291-299.
15 Horesovsky G, Ginsler J, Everitt J, Walker C. 1994. Increased susceptibility to in vitro
16 transformation of cells. Carcinogenesis 15:2183-2187.
17 IARC (International Agency for Research on Cancer). 1995. IARC Monographs on the
18 Evaluation of Carcinogenic Risks to Humans. Vol 63, Dry Cleaning, Some Chlorinated solvents
19 and other Industrial Chemicals.
20 Jaffe, DR; Gandolfi, AJ; Nagle, RB. (1984) Chronic toxicity of S-(trans-l,2-dichlorovinyl)-L-
21 cysteine in mice. J Appl Toxicol 4:315-319.
22 Jaffe DR, Hassall CD, Gandolfi AJ, Brendel K. 1985. Production of DNA single strand breaks in
23 rabbit renal tissue after exposure to 1,2-dichlorovinylcysteine. Toxicology 35:25-33.
24 Jemal A, Siegel R, Ward E, Hao Y, Xy J, Murray T, Thun MJ. 2008. cancer Statistics, 2008.
25 CA Cancer J Clin 58:71-96.
26 Jensen OM, Knudsen JB, McLaughlin JK, Sorensen BL. 1988 The Copenhagen case-control
27 study of renal pelvis and ureter cancer: role of smoking and occupational exposures. Int J
28 Cancer 41:557-561.
6/22/2009
464
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Kaneko T, Wang P-Y, Sato A. 1994. Enzymes induced by ethanol differently affect the
2 pharmacokinetics of trichloroethylene and 1,1,1-trichloroethane. Occupational and
3 Environmental Medicine 51:113-119.
4 Kanisawa M, Suzuki S. 1978. Induction of renal and hepatic tumors in mice by ochratoxin A, a
5 mycotoxin. Gann. 69:599-600.
6 Kelsh MA, Weihgart M, Mandel J, Mink P, Alexander D, Basu R, Kalmes R, Goodman M.
7 2005. A meta-analysis of epidemiology studies of occupational TCE exposure and selected
8 cancers. Presented to National Research Committee on Health Risks of Trichloroethylene. June
9 9,2005.
10 Kenck C, Wilhelm M, Bugert P, et al. 1996. Mutation of the VHL gene is associated exclusively
11 with the development of non-papillary renal cell carcinomas. J Pathol 179: 157-161.
12 Kjellstrand, P; Holmquist, B; Mandahl, N; et al. (1983b) Effects of continuous trichloroethylene
13 inhalation on different strains of mice. Acta Pharmacol Toxicol (Copenh) 53:369-374.
14 Korrapati, MC; Lock, EA; Mehendale, HM. (2005) Molecular mechanisms of enhanced renal
15 cell division in protection against S-l,2-dichlorovinyl-L-cysteine-induced acute renal failure and
16 death. Am J Physiol Renal Physiol 289:F175-F185.
17 Krejci, ME; Ridgewell, RE; Koechel, DA. (1991) Acute effects of the D-isomer of S-(l,2-
18 dichlorovinyl)cysteine on renal function and ultrastructure in the pentobarbital-anesthetized dog:
19 site-specific toxicity involving the SI and S2 cells of the proximal tubule. Toxicology
20 69:151-164.
21 Lash LH, Anders MW. 1986. Cytotoxicity of S-(l,2-dichlorovinyl)glutathione and S-(l,2-
22 dichlorovinyl)-L-cysteine in isolated rat kidney cells. Journal of Biol Chem 261:13076-13081.
23 Lash LH, Xu Y, Elfarra AA, Duescher RJ, Parker JC. 1995. Glutathione-dependent metabolism
24 of trichloroethylene in isolated liver and kidney cells of rats and its role in mitochondrial and
25 cellular toxicity. Drug Metabolism and Disposition 23:846-853.
26 Lash LH, Putt DA, Brashear WT, Abbas R, Parker JC, Fisher JW. 1999. Identification of S-(l,2-
27 dichlorovinyl)glutathione in the blood of human volunteers exposed to trichloroethylene. Journal
28 of Toxicology and Environmental Health, Part A 56:1-21.
6/22/2009
465
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Lash, LH; Fisher, JW; Lipscomb, JC; et al. (2000a) Metabolism of trichloroethylene. Environ
2 Health Perspect 108(Suppl. 2): 177-200.
3 Lash, LH; Parker, JC; Scott, CS. (2000b) Modes of action of trichloroethylene for kidney
4 tumorigenesis. Environ Health Perspect 108 Suppl 2:225-240.
5 Lash, LH; Qian, W; Putt, DA; et al. (2001) Renal and hepatic toxicity of trichloroethylene and its
6 glutathione-derived metabolites in rats and mice: sex-, species-, and tissue-dependent
7 differences. J Pharmacol Exp Ther 297:155-164.
8 Lash L, Putt DA, Parker JC. 2006. Metabolism and tissue distribution of orally administered
9 trichloroethylene in male and female rats: identification of glutathione- and cytochrome P-450-
10 derived metabolites in liver, kidney, blood and urine. Journal of Toxicology and Environmental
11 Health, Part A 69:1285-1309.
12 Lash LH, Putt DA, Huang P, Hueni SE, Parker JC. 2007. Modulation of hepatic and renal
13 metabolism and toxicity of trichloroethylene and perchloroethylene by alterations in status of
14 cytochrome P450 and glutathione. Toxicology 235: 11-26.
15 Liu E, Jelinek J, Pastore YD, Guan Y, Prchal JF, Prchal JT. 2003. Discrimination of
16 polycythemias and thrombocytoses by novel, simple, accurate clonality assays and comparison
17 with PRV-1 expression and BFU-E response to erythropoietin. Blood 101:3294-3301.
18 Lock EA, Reed CJ. 2006. Trichloroethylene: Mechanisms of Renal Toxicity and Renal Cancer
19 and Relevance to Risk Assessment. Toxicological Sciences 91:313-331.
20 Lock, EA; Reed, CJ; McMillan, JM; et al. (2007) Lack of formic acid production in rat
21 hepatocytes and human renal proximal tubule cells exposed to chloral hydrate or trichloroacetic
22 acid. Toxicology 230:234-243.
23 Lybarger JA, Lichtveld MY, Amler RW. 1999. Biomedical testing of the kidney for persons
24 exposed to hazardous substances in the environment. Renal Failure 21:263-274.
25 Mally A, Walker CL, Everitt JI, Dekant W, Vamvakas S. 2006. Analysis of renal cell
26 transformation following exposure to trichloroethene in vivo and its metabolite S-
27 (dichlorovinyl)-L-cysteine in vitro. Toxicology 224:108-118.
6/22/2009
466
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Maltoni, C; Lefemine, G; Cotti, G. (1986) Experimental research on trichloroethylene
2 carcinogenesis. In: Maltoni, C; Mehlman MA., eds. Vol. 5. Archives of research on industrial
3 carcinogenesis. Princeton, NJ: Princeton Scientific Publishing;
4 Maltoni, C; Lefemine, G; Cotti, G; et al. (1988) Long-term carcinogenicity bioassays on
5 trichloroethylene administered by inhalation to Sprague-Dawley rats and Swiss and B6C3F1
6 mice. Ann N Y Acad Sci 534:316-342.
7 Mandel JS. 2001. Renal cell cancer correlated with occupational exposure to trichloroethylene.
8 J Cancer Res Clin Oncol 127:265-268.
9 Mandel JS, McLaughlin JK, Schlehofer B, Mellemgaard A, Helmert U, Lindblad P, McCredie
10 M, Adami HO. 1995. Internal renal-cell cancer study. IV. Occupation. Int J Cancer
11 61:601-605.
12 Mather, GG; Exon, JH; Koller, LD. (1990) Subchronic 90 day toxicity of dichloroacetic and
13 trichloroacetic acid in rats. Toxicology 64:71-80.
14 Mattioli S, Truffelli D, Baldasseroni A, Risi A, Marchesini B, Gizcomini C, Bacchini P, Violante
15 FS, Buiatti E. 2002. Occupational risk factors for renal cell cancer: a case-control study in
16 northern Italy. J Occup Environ Med. 44:1028-1036.
17 McCredie M, Stewart JH. 1993. Risk factors for kidney cancer in New South Wales. IV.
18 Occupation. Br J Ind Med. 50:349-54.
19 McLaughlin JK, Blot WJ. 1997. A critical review of epidemiology studies of trichloroethylene
20 and perchloroethylene and risk of renal-cell cancer. Int Arch Occup Environ Health. 70:222-31.
21 McLaughlin JK, Lipworth L, Tarone RE. 2006. Epidemiologic aspects of renal cell carcinoma.
22 Semin Oncol. 33:527-33.
23 Mellemgaard A, Engholm G, McLaughlin JK, Olsen JH. 1994. Occupational risk factors for
24 renal-cell carcinoma in Denmark. Scand J Work Environ Health 20:160-165.
25 Mensing, T; Welge, P; Voss, B; et al. (2002) Renal toxicity after chronic inhalation exposure of
26 rats to trichloroethylene. Toxicol Lett 128:243-247.
27 Moore LE, Wilson RT, Campleman SL. 2005. Lifestyle factors, exposures, genetic
28 susceptibility, and renal cell cancer risk: a review. Cancer Invest 23:240-255.
6/22/2009
467
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Morgan JW, Cassidy RE. 2002. Community cancer assessment in response to long-time
2 exposure to perchlorate and trichloroethylene in drinking water. J Occup Environ Med.
3 44:616-621.
4 Morgan RW, Kelsh MA, Zhao K, Heringer S. 1998. Mortality of aerospace workers exposed to
5 trichloroethylene. Epidemiology 9:424-431.
6 Morton LD, Youssef AF, Lloyd E, Kiorpes AL, Goldsworthy TL, Fort FL. 2002. Evaluation of
7 carcinogenic responses in the Eker rat following short-term exposure to selected nephrotoxins
8 and carcinogens. Toxicologic Pathology 30:559-564.
9 Nagaya T, Ishikawa N, Hata H. 1989. Urinary total protein and P2-Microglobulin in workers
10 exposed to trichloroethylene. Environ Res 50:86-92.
11 NCI (National Cancer Institute). (1976) Carcinogenesis bioassay of trichloroethylene. Division
12 of Cancer Cause and Prevention, National Cancer Institute, U.S. Department of Health,
13 Education, and Welfare, DHEW Publication No. (NIH) 76-802, Technical Report Series No. 2,
14 218 pages; NCI-CG-TR-2; NTIS PB254122.
15 http://ntp.niehs.nih.gov/ntp/htdocs/LT rpts/tr002.pdf.
16 Nickerson ML, Jaeger E, Shi Y, Durocher JA, et al. 2008. Improved identification of von
17 Hippel-Lindau gene alterations in clear cell renal tumors. Clin Cancer Res 14: 4726-4734.
18 NRC (National Research Council). 2006. Assessing the Human Health Risks of
19 Trichloroethylene: Key Scientific Issues. Washington DC: National Academy Press.
20 NRC (National Research Council). 2006. Assessing the Human Health Risks of
21 Trichloroethylene: Key Scientific Issues. Washington DC: National Academy Press.
22 NTP (National Toxicology Program). (1988) Toxicology and carcinogenesis studies of
23 trichloroethylene (CAS no. 79-01-6) in four strains of rats (ACI, August, Marshall, Osborne-
24 Mendel) (gavage studies). Public Health Service, U.S. Department of Health and Human
25 Services; NTP TR-273; NIH Publication No. 88-2529. Available from the National Institute of
26 Environmental Health Sciences, Research Triangle Park, NC, and the National Technical
27 Information Service, Springfield, VA; PB88-218896.
28 http://ntp.niehs.nih.gov/ntp/htdocs/LT_rpts/tr273 .pdf.
29 NTP (National Toxicology Program). (1990) Carcinogenesis studies of trichloroethylene
30 (without epichlorohydrin) in F344/N rats and B6C3F1 mice (gavage studies). Public Health
6/22/2009
468
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Service, U.S. Department of Health and Human Services; NTP TR-243; NIH Publication No. 90-
2 1779. Available from the National Institute of Environmental Health Sciences, Research
3 Triangle Park, NC. http://ntp.niehs.nih.gov/ntp/htdocs/LT rpts/tr243.pdf.
4 Parent ME, HuaY, Siemiatycki. 2000. Occupational risk factors for renal cell carcinoma in
5 Montreal. Am J Ind Med 38:609-618. x 4
6 Partenen T, Heikkila P, Hernberg S, Kauppinen T, Moneta G, Ojajarvi A. 1991. Renal cell
7 cancer and occupational exposure to chemical agents. Scand J Work Environ Health
8 17:231-239.
9 Pesch B, Haerting J, Ranft U, Klimpel A, Oelschlagel B, Schill W. 2000a. Occupational risk
10 factors for renal cell carcinoma: agent-specific results from a case-control study in Germany.
11 MURC Study Group. Multicenter urothelial and renal cancer study. Int J Epidemiol.
12 29:1014-1024.
13 Price RG, Taylor SA, Chivers I, Arce-Tomas M, Crutcher E, Franchini I, et al., 1996.
14 Development and validation of new screening tests for nephrotoxic effects. Human Environ
15 Toxicol 15, Suppl. 1:510-519.
16 Price RG, Wedeen R, Lichtveld MY, Lybarger JA, Poter GA, Fels L, et al., 1999. Urinary
17 biomarkers: roles in risk assessment to environmental and occupational nephrotoxins:
18 monitoring of effects and evaluation of mechanisms of toxicity. Renal Failure 21 :xiii-xviii.
19 Raaschou-Nielsen O, Hansen J, McLaughlin JK, Kolstad H, Christensen JM, Tarone RE, Olsen
20 JH. 2003. Cancer risk among workers at Danish companies using trichloroethylene: a cohort
21 study. Am J Epidemiol 158:1182-1192.
22 Radican L, Wartenberg D, Rhoads, GG, Schneider D, Wedeen R, Stewart P, Blair A. 2006. A
23 retrospective occupational cohort study of end-stage renal disease in aircraft workers exposed to
24 trichloroethylene and other hydrocarbons. J Occup Environ Med 48:1-12.
25 Radican L, Blair A, Stewart P, Wartenberg D. 2008. Mortality follow-up in hydrocarbon-
26 exposed workers. J Occup Environ Med 50:1306-1319.
27 Rasmussen K, Brogren C-H, Sabroe S. 1993. Subclinical affection of liver and kidney function
28 and solvent exposure. Int Arch Occup Environ Health 64:445-448.
6/22/2009
469
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Reznik G, Ward JM, Hardisty JF, & Russfield A (1979) Renal carcinogenic and nephrotoxic
2 effects of the flame retardant tris(2,3-dibromopropyl) phosphate in F344 rats and (C57B1/6N x
3 C3H/HeN)Fl mice. J Natl Cancer Inst, 63(1): 205-212.
4 Rhomberg, LR. (2000) Dose-response analyses of the carcinogenic effects of trichloroethylene in
5 experimental animals. Environ Health Perspect 108 Suppl 2:343-358.
6 Ries LAG, Melbert D, Krapcho M, Stinchcomb DG, Howlader N, Horner MJ, Mariotto A,
7 Miller BA, Feuer EJ, Altekruse SF, Lewis DR, Clegg L, Eisner MP, Reichman M, Edwards BK
8 (eds). SEER Cancer Statistics Review, 1975-2005, National Cancer Institute. Bethesda, MD,
9 http://seer.cancer.gov/csr/1975_2005/, based on November 2007 SEER data submission, posted
10 to the SEER web site, 2008. Access July 9, 2008.
11 RitzB. 1999. Cancer mortality among workers exposed to chemicals during uranium
12 processing. J Occup Environ Med. 41:556-566.
13 Robbiano L, Mereto E, Morando AM, Pastore P, Brambilla G. 1998. Increased frequency of
14 micronucleated kidney cells in rats exposed to halogenated anaesthetics. Mutation Research
15 413:1-6.
16 Robbiano L, Baroni D, Carrozzino R, Mereto E, Brambilla G. 2004. DNA damage and
17 micronuclei induced in rat and human kidney cells by six chemicals carcinogenic to the rat
18 kidney. Toxicology 204:18-195.
19 Salvan A, Stayner L, Steenland K, Smither T. 1995. Selecting an exposure lag period.
20 Epidemiology. 6:387-390.
21 Schlehofer B, Heuer C, Blettner M, Niehoff D, Wahrendorf J. 1995. Occupation, smoking and
22 demographic factors, and renal cell carcinoma in Germany. Int J Epidemiol. 24(1):51—7.
23 Schraml P, Zhou M, Richter J, Briining T, Pommer M, Sauter G, Mihatsch MJ, Moch H. 1999.
24 Analysis of renal tumors in trichloroethylene-exposed workers by comparative genomic
25 hybridization and DNA sequencing analysis Verh Dtsch Ges Pathol 83:218-224.
26 Selden A, Hultberg B, Ulander A, Ahlborg G. 1993. Trichloroethylene exposure in vapour
27 degreasing and the urinary excretion of N-acetyl-P-glucosaminidase. Arch Toxicol 67:224-226.
28 Sharp CR, Rochon JE, Adam JM, Suissa S. 1989. Case-control study of hydrocarbon exposure
29 in patients with renal cell carcinoma. CMAJ 140:1309-1318.
6/22/2009
470
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Shiao, YH; Rice, JM; Anderson, LM; et al. (1998) von Hippel-Lindau gene mutations in N-
2 nitrosodimethylamine-induced rat renal epithelial tumors. J Natl Cancer Inst 90:1720-1723.
3 Shiao YH, Kamata SI, Li LM, Hooth MJ, DeAngelo AB, Anderson LM, Wolf DC. 2002.
4 Mutations in the VHL gene from potassium bromate-induced rat clear cell renal tumors. Cancer
5 Letters 187:207-214.
6 Siemiatycki J, Wacholder S, Dewar R, Cardis E, Greenwood C, Richardson L. 1988. Degree of
7 confounding bias related to smoking, ethnic group, and socioeconomic status in estimates of the
8 associations between occupational and cancer. J Occup Med 30:617-625.
9 Siemiatycki J. 1991. Risk Factors for Cancer in the Workplace. J Siemiatycki, Ed. Baca
10 Raton: CRC Press.
11 Sinks T, Lushniak B, Haussler BJ, Sniezek J, Deng J-F, Roper P, Dill P, Coates R. 1992. Renal
12 cell cancer among paperboard printing workers. Epidemiol 3:483-489.
13 Stemmer K, Ellinger-Ziegelbauer H, Ahr HJ, Dietrich DR. 2007. Carcinogen-specific gene
14 expression profiles in short-term treated Eker and wild-type rats indicative of pathways involved
15 in renal tumorigenesis. Cancer Res. 67(9):4052-68.
16 Stevens JL, Robbins JD, Byrd RA. 1986. A purified cysteine conjugate B-lyase from rat kidney
17 cytosol. Journal of Biol Chem 261:15529-15537.
18 Stewart PA, Lee JS, Marano DE, Spirtas R, Forbes CD, Blair A. 1991. Retrospective cohort
19 mortality study of workers at an aircraft maintenance facility. II. Exposures and their
20 assessment. Br J Ind Med 48:531-537.
21 Sung TI, Chen PC, Jyuhn-Hsiarn Lee L, Lin YP, Hsieh GY, Wang JD. (2007) Increased
22 standardized incidence ratio of breast cancer in female electronics workers. BMC Public Health
23 7:102.
24 Sung TI, Wang JD, Chen PC. 2008. Increased risk of cancer in the offspring of female
25 electronics workers. Reprod Toxicol 25:115-119.
26 Swaen GM. 1995. Increased incidence of renal cell tumours in a cohort of cardboard workers
27 exposed to trichloroethylene. Comments on Henschler D. et al. 1995. Arch Toxicol
28 70:127-128, 131-133.
6/22/2009
471
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Terracini, B; Parker, V. (1965a) A pathological study on the toxicity of s-dichlorovinyl-1-
2 cysteine. Food Cosmet Toxicol 42:67-74.
3 Toma ML, Grosser M, Herr A, Aust DE, Meye A, Hoefling C et al. 2008. Loss of
4 heterozygosity and copy number abnormality in clear cell renal cell carcinoma discovered by
5 high-density affymetrix 10K single nucleotide polymorphism mapping array. Neoplasia 10 (7):
6 634-642.
7 U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. Risk Assessment Forum,
8 Washington, DC; EPA/630/P-03/001B. Available from: .
10 U.S. EPA. (2005b) Supplemental guidance for assessing susceptibility from early-life exposure
11 to carcinogens. Risk Assessment Forum, Washington, DC; EPA/630/R-03/003F. Available from:
12 .
13 Vaidya, VS; Shankar, K; Lock, EA; et al. (2003a) Renal injury and repair following S-l, 2
14 dichlorovinyl-L-cysteine administration to mice. Toxicol Appl Pharmacol 188:110-121.
15 Vaidya, VS; Shankar, K; Lock, EA; et al. (2003b) Role of tissue repair in survival from s-(l,2-
16 dichlorovinyl)-L-cysteine-induced acute renal tubular necrosis in the mouse. Toxicol Sci
17 74:215-227.
18 Vaidya, VS; Shankar, K; Lock, EA; et al. (2003c) Molecular mechanisms of renal tissue repair in
19 survival from acute renal tubule necrosis: role of ERK1/2 pathway. Toxicol Pathol 31:604-618.
20 Vamvakas S, Dekant W, Berthold K, Schmidt S, Wild D, Henschler D. (1987) Enzymatic
21 transformation of metcapturic acids derived from halogenated alkenes to reactive mutagenic
22 intermediates. Biochem Pharmacol. 36(17):2741-8.
23 Vamvakas S, Elfarra AA, Dekant W, Henschler D, Anders MW (1988a) Mutagenicity of amino
24 acid and glutathione S-conjugates in the Ames test. Mutat Res. 206(l):83-90.
25 Vamvakas S, Dekant W, Schiffmann D, Henschler D. (1988b) Induction of unscheduled DNA
26 synthesis and micronucleus formation in Syrian hamster embryo fibroblasts treated with cysteine
27 S-conjugates of chlorinated hydrocarbons. Cell Biol Toxicol. (4):393-403.
28 Vamvakas S, Richter H, Bittner D. (1996) Induction of dedifferentiated clones of LLC-PK1 cells
29 upon long-term exposure to dichlorovinylcysteine. Toxicology. 106(l-3):65-74.
6/22/2009
472
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Vamvakas S, Briining T, Thomasson B, Lammert M, Baumiiller A, Bolt HM et al. 1998. Renal
2 cell cancer correlated with occupational exposure to trichloroethene. J Cancer Res Clin Oncol
3 124:374-382.
4 Van Duuren, BL; Goldschmidt, BM; Loewengart, G; et al. (1979) Carcinogenicity of
5 halogenated olefinic and aliphatic hydrocarbons in mice. J Natl Cancer Inst 63:1433-1439.
6 Vartiainen T, Pukkala E, Rienoja T, Strandman T, Kaksonen K. 1993. Population exposure to
7 tri- and tetrachloroethene and cancer risk: two cases of drinking water pollution. Chemosphere
8 27:1171-1181.
9 Wartenberg D, Reyner D, Scott CS. 2000. Trichloroethylene and Cancer: Epidemiologic
10 Evidence. Environ Health Perspect 108(suppl 2): 161-176.
11 Wiesenhiitter B, Selinski S, Golka K, Briining T, Bolt HM. 2007. Re-assessment of the
12 influence of polymorphisms of phase-II metabolic enzymes on renal cell cancer risk of
13 trichloroethylene-exposed workers. 81:247-251.
14 Wilson RT, Donahue M, Gridley G, Adami J, El Ghormli L, Dosemeci M. 2008. Shared
15 occupational risks for transitional cell cancer of the bladder and renal pelvis among men and
16 women in Sweden. Am J Ind Med. 51:83-99.
17 Wolf DC, Goldsworthy TL, Janszen DB, Harden R, Donner EM, David CS, Everitt JI. 2000.
18 Promotion by sodium barbital induces early development but does not increase the multiplicity
19 of hereditary renal tumors in Eker rats. Carcinogenesis 21:1553—1558.
20 Wolfgang, GH; Gandolfi, AJ; Stevens, JL; et al. (1989a) N-acetyl S-(l,2-dichlorovinyl)-L-
21 cysteine produces a similar toxicity to S-(l,2-dichlorovinyl)-L-cysteine in rabbit renal slices:
22 differential transport and metabolism. Toxicol Appl Pharmacol 101:205-219.
23 Wolfgang, GH; Gandolfi, AJ; Stevens, JL; et al. (1989b) In vitro and in vivo nephrotoxicity of
24 the L and D isomers of S-(l,2-dichlorovinyl)-cysteine. Toxicology 58:33-42.
25 Zandi-Nejad K, Eddy AA, Glassock RJ, Brenner BM. 2004. Why is proteinuria an ominous
26 biomarker of progressive kidney disease? Kidney Int Suppl 92:S76-89.
27 Zanelli U, Puccini P, Acerbi D, Ventura P, Gervasi PG. 1996. Induction of peroxisomal beta-
28 oxidation and P-450 4°-dependent activities by pivalic and trichloroacetic acid in rat liver and
29 kidney. Arch Toxicol 70:145-149.
6/22/2009
473
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Zhang Y, Cantor KP, Lynch CF, Zheng T. 2004. A population-based case-control study of
2 occupation and renal cell carcinoma risk in Iowa. J Occup Environ Med. 46:235-240.
3 Zhao Y, Krishnadasan A, Kennedy N, Morgenstern H, Ritz B. 2005. Estimated effects of
4 solvents and mineral oils on cancer incidence and Mortality in a cohort of aerospace workers.
5 Am J Ind Med 48:249-258.
6
6/22/2009
474
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.4 Liver toxicity and cancer
4.4.1 Liver non-cancer toxicity in humans
The complex of chronic liver disease is a spectrum of effects and comprises nonalcoholic
fatty liver disease (nonalcoholic steatohepatitis) and cirrhosis, more rare anomalies ones such as
autoimmune hepatitis, primary biliary cirrhosis, and primary sclerosing cholangitis, and
hepatocellular and cholangiocarcinoma (intrahepatic bile duct cancer) (Juran and Lazaridis,
2006). Chronic liver disease and cirrhosis, excluding neoplasia, is the 12th leading cause of death
in the United States in 2005 with 27,530 deaths (Kung et al., 2008) with a morality rate of 9.0
per 100,000 (Jemal et al., 2008).
Three studies are suggestive of effects on liver function tests in metal degreasers
occupationally exposed to trichloroethylene (Nagaya et al., 1993; Rasmussen et al., 1993; Xu et
al., 2009). Nagaya et al. (1993) in their study of 148 degreasers in metal parts factories,
semiconductor factors, or other factories, observed total mean serum cholesterol concentration,
mean serum high density lipoprotein-cholesterol (HDL-C) concentrations to increase with
increasing TCE exposure, as defined by urinary excretion of total trichloro-compounds (U-TTC),
although a statistically significant linear trend was not found. Nagaya et al. (1993) estimated
subjects in the low exposure group had TCE exposure to 1 ppm, 6 ppm TCE in the moderate
exposure group, and 210 ppm TCE in the high exposure group. No association was noted
between serum liver function tests and U-TTC, a finding not surprising given individuals with a
history of hepatobiliary disease were excluded from this study. Nagaya et al. (1993) follows 13
workers with higher U-TTC concentrations over a 2-year period; serum HDL-C and two hepatic
function enzymes, gamma glutamyl transferase (GGT) and aspartate aminotrasferase (AST)
concentrations were highest during periods of high level exposure, as indicated from U-TTC
concentrations. Similarly, in a study of 95 degreasers, 70 exposed to trichloroethylene exposure
and 25 to CFC 113 (Rasmussen et al., 1993), mean serum GGT concentration for subjects with
the highest cumulative TCE exposure was above normal reference values and were about 3-fold
higher compared to the lowest exposure group. Rasmussen et al. (1993) estimated mean urinary
TCE concentration in the highest exposure group as 7.7 mg/L with past exposures estimated as
equivalent to 40-60 mg/L. Multivariate regression analysis showed a small statistically non-
significant due to age and a larger effect due to alcohol abuse that reduced but did not eliminate a
TCE exposure affect. Some question exists regarding the presentation of findings from
regression modeling; for example, a negative slope or inverse relationship reported for GGT and
cumulative TCE exposure appears inconsistent with data presented in tables suggesting higher
6/22/2009
475
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
GGT concentrations with higher cumulative TCE exposure. Moreover, the inclusion of CFC113
exposed subjects introduces a downward bias since liver toxicity is not associated with CFC113
exposure (U.S. EPA, 2008) and would underestimate any possible TCE effect. Xu et al. (2009)
reported symptoms and liver function tests of 21 metal degreasers with severe hypersensitivy
dermatitis (see last paragraph in this section for discussion of other liver effects in
hypersensitivity dermatitis cases). TCE concentration of agent used to clean metal parts ranged
from 10.2% to 63.5% with workplace ambient monitoring time-weighted-average TCE
concentrations of 18 mg/m3to 683 mg/m3 (3 to 127 ppm). Exposure was further documented by
urinary TCA levels in 14 of 21 cases above the recommended occupation level of 50 mg/L. The
prevalence of elevated liver enzymes among these subjects was 90% (19 cases) for alanine
aminotrasferase, 86% (18 cases) for asparatate aminotrasferase, and 76% (16 cases) for total
bilirubin (Xu et al., 2009).
Two studies provide evidence of plasma or serum bile acids changes among TCE-
exposed degreasers. Neghab et al. (1997) in a small prevalence study of 10 healthy workers (5
unexposed controls and 5 exposed) observed statistically significantly elevated total serum bile
acids, particularly deoxycholic acid and the subtotal of free bile acids, among TCE subjects at
post-exposure compared to their pre-exposure concentrations and serum bile acid levels
correlated well with TCE exposure (r = 0.94). Total serum bile acid concentration did not
change in control subjects between pre- and postexposure, nor did enzyme markers of liver
function in either unexposed or exposed subjects differ between pre and post-exposure period.
However, the statistical power of this study is quite limited and the prevalence design does not
include subjects who may have left employment because of possible liver problems. The paper
provides minimal details of subject selection and workplace exposure conditions, except that pre-
exposure testing was carried out on the 1st work day of the week (pre-exposure), repeated
sampling after 2 days (post-exposure), and a post-exposure 8-hour time-weighted-average TCE
concentration of 9 ppm for exposed subjects; no exposure information is provided for control
subjects. Driscoll et al. (1992) in a study of 22 subjects (6 unexposed and 16 exposed) employed
at a factory manufacturing small appliances reported statistically significant group differences in
logistic regression analyses controlling for age and alcohol consumption in mean fasting plasma
bile acid concentrations. Other indicators of liver function such as plasma enzyme levels were
statistically significant different between exposed and unexposed subjects. Laboratory samples
were obtained at the start of subject's work shift. Exposure data are not available on the 22
subjects and assignment of exposed and unexposed was based on work duties. Limited personal
monitoring from other non-participating workers at this facility indicated TCE exposure as low,
less than 5 ppm, with occasional peaks over 250 ppm although details are lacking whether these
data represent exposures of study subjects.
6/22/2009
476
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Davis et al. (2006) in their analysis of subjects from the TCE subregistry of ATSDR's
National Exposure Registry examined the prevalence of subjects reporting liver problems
(defined as seeking treatment for the problem from a physician within the past year) using rates
for the equivalent health condition from the National Health Interview Survey (a nationwide
multi-purpose health survey conducted by the National Center for Health Statistics (NCHS),
Centers for Disease Control and Prevention (CDC)). The TCE subregistry is a cohort of exposed
persons from 15 sites in 5 states. The shortest time interval from inclusion in the exposure
registry and last follow-up was 5 years for one site and 10 years for seven sites. Excess in past-
year liver disorders relative to the general population persisted for much of the lifetime of
follow-up. Standardized morbidity ratios (SMRs) for liver problems were 3rd follow-up,
SMR = 2.23 (99% CI: 1.13, 3.92); 4th follow-up, SMR = 3.25 (99% CI: 1.82, 5.32); and, 5th
follow-up, SMR = 2.82 (99% CI: 1.46, 4.89). Examination by TCE exposure, duration or
cumulative exposure to multiple organic solvents did not show exposure-response patterns.
Overall, these observations are suggestive of liver disorders as associated with potential TCE
exposure, but whether TCE caused these conditions is not possible to determine given the
study's limitations. These limitations include a potential for misclassification bias, the direction
of which could dampen observations in a negative direction, and lack of adjustment in statistical
analyses for alcohol consumption, which could bias observations in a positive direction.
Evaluation in epidemiologic studies of risk factors for cirrhosis other than alcohol
consumption and Hepatitis A, B, and C is quite limited. NRC (2006) cited a case report of
cirrhosis developing in an individual exposed occupationally to TCE for 5 years from a hot-
process degreaser and to 1, 1, 1-trichloroethane for 3 months thereafter (Thiele et al., 1982). One
cohort study on cirrhosis deaths in California between 1979 and 1981 and occupational risk
factors as assessed using job title observed elevated risks with occupational titles of sheet metal
workers and metalworkers and cirrhosis among white males who comprised the majority of
deaths (Leigh and Jiang, 1993). This analysis lacks information on alcohol patterns by
occupational title in addition to specific chemical exposures. Few deaths attributable to cirrhosis
are reported for nonwhite male and for both white and nonwhite female metalworkers with
analyses examining these individuals limited by low statistical power. Some but not all
trichloroethylene mortality studies report risk ratios for cirrhosis (Table 4.4.2). A statistically
significant deficit in cirrhosis mortality is observed in three studies (Morgan et al., 1998; Boice
et al., 1999, 2006) and with risk ratios including a risk of 1.0 in the remaining studies (Garabrant
et al., 1988; Blair et al., 1998; Ritz, 1999; ATSDR, 2004). These results do not rule out an effect
of TCE on liver cirrhosis since disease misclassification may partly explain observations.
Available studies are based on death certificates where a high degree of underreporting, up to
50%), is know to occur (Blake et al., 1988).
6/22/2009
477
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
A number of case reports exist of liver toxicity including hepatitis accompanying
immune-related generalized skin diseases described as a variation of erythema multiforme,
Stevens-Johnson syndrome, toxic epiderma necrolysis patients, and hypersensitivity syndrome
(Section 4.5.1.2. describes these disorders and evidence on TCE) (Kamijima et al., 2007).
Kamijima et al. (2007) reported hepatitis was seen in 92%-94% of cases presenting with an
immune-related generalized skin diseases of variation of erythema multiforme, Stevens-Johnson
syndrome, and toxic epiderma necrolysis patients, but the estimates within the hypersensitivity
syndrome group were more variable (46%-94%). Many cases developed with a short time after
initial exposure and presented with jaundice, hepatomegaly or hepatosplenomegaly, in addition,
to hepatitis. Hepatitis development was of a non-viral etiology, as antibody titers for Hepatitis
A, B, and C viruses were not detectable, and not associated with alcohol consumption (Huang et
al., 2002; Kamijima et al., 2007). Liver failure was moreover a leading cause of death among
these subjects. Kamijima et al. (2007) note the similarities between specific skin manifestations
and accompanying hepatic toxicity and case presentations of TCE-related generalized skin
diseases and conditions that have been linked to specific medications (e.g., carbamezepine,
allupurinol, antibacterial sulfonamides), possibly in conjunction with reactivation of specific
latent viruses. However, neither cytomegalovirus or Epstein-Barr viruses are implicated in the
few reports which did include examination of viral antibodies.
4.4.2 Liver cancer in humans
Primary hepatocellular carcinoma and cholangiocarcinoma (intrahepatic and extrahepatic
bile ducts) are the most common primary hepatic neoplasms (El-Serag, 2007; Blehacz and
Gores, 2008). Primary hepatocellular carcinoma is the 5th most common of cancer deaths in
males and 9th in females (Jemal et al., 2008). Age-adjusted incidence rates of hepatocellular
carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC) are increasing, with a 2-fold
increase in HCC over the past 20 years. This increase has not attributable to an expanded
definition of liver cancer to include primary or secondary neoplasms since ICD-9, incorrect
classification of hilar cholangiocarcinomas in ICD-0 as ICC, or to improved detection methods
(Welzel et al., 2006; El-Serag, 2007). It is estimated that 21,370 Americans will be diagnosed in
2008 with liver and intrahepatic bile cancer; age-adjusted incidence rates for liver and
intrahepatic bile duct cancer for all races are 9.9 per 100,000 for males and 3.5 per 100,000 for
females (Ries et al., 2008). Survival for liver and biliary tract cancers remains poor and age-
adjusted mortality rates are just slightly lower than incidence rates. While hepatitis B and C
viruses and heavy alcohol consumption are believed major risk factors for HCC and intrahepatic
cholangiocarcinoma, these risk factors cannot fully account for roughly 10% and 20% of HCC
6/22/2009
478
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
cases (Kulkarni et al., 2004). Cirrhosis is considered a premalignant condition for HCC,
however, cirrhosis is not a sufficient cause for HCC since 10% to 25% of HCC cases lack
evidence of cirrhosis at time of detection (Chiesa et al., 2000; Fattovich et al., 2004; Kumar et
al., 2007). Nonalcoholic steatohepatitis reflecting obesity and metabolic syndrome is recently
suggested as contributing to liver cancer risk (El-Serag, 2007). Few data exist on extrahepatic
cholangiocarcinoma (ECC)-related incidence and mortality other than ECC may account for
50% of the estimated 5,000 new cases diagnosed annually (Shaib and El-Serag, 2004).
All cohort studies, except Zhao et al. (2005), present risk ratios (SIRs or SMRs) for liver
and biliary tract cancer. More rarely reported in cohort studies are risk ratios for primary liver
cancer (hepatocellular carcinoma or HCC) or for gallbladder and extrahepatic bile duct cancer.
Four community studies also presented risk ratios for liver and biliary tract cancer including a
case-control study of primary liver cancer of residents of Taiwanese community with solvent-
contaminated drinking water wells (Vartiainen et al., 1993; Morgan and Cassidy, 2002; Lee et
al., 2003; ATSDR, 2006). Several population case-control studies examine liver cancer and
organic solvents or occupational job titles with possible TCE usage (Stemhagen et al., 1983;
Hardell et al., 1984; Hernberg et al., 1984, 1988; Austin et al., 1987; Dossing et al., 1997;
Heinemann et al., 2000; Porru et al., 2001; Weiderpass et al., 2003; Ji and Hemminki, 2005;
Kvam et al., 2005; Lindbohm et al., 2009); however, the lack of exposure assessment to TCE,
specifically, or, too few exposed cases and controls in those studies that do present some
information limits their usefulness for evaluating hepatobiliary or gall bladder cancer and
trichloroethylene exposure. Table 4.4.3 presents observations from cohort, case-control, and
community studies on liver and biliary tract cancer, primary liver, and gallbladder and
extrahepatic bile duct cancer and trichloroethylene.
Excess liver cancer incidence is observed in most high quality studies (Axelson et al.,
1994; Anttila et al., 1995; Hansen et al., 2001; Raaschou-Nielsen et al., 2003) as is mortality in
studies which assess TCE exposure by job exposure matrix approaches (Blair et al., 1998;
Morgan et al., 1998; Ritz, 1999; ATSDR, 2004; Boice et al., 2006; Radican et al., 2008). Risks
for primary liver cancer and for gallbladder and biliary tract cancers in females were statistically
significantly elevated only in Raaschou-Nielsen et al. (2003), the study with the largest number
of observed cases, although without suggestion of exposure duration-response patterns. Cohort
studies with more uncertain exposure assessment approaches, e.g., studies of all subjects working
at a factory (Garabrant et al., 1998; Costa et al., 1989; Chang et al., 2003, 2005), do not show
association but are quite limited given their lacking attribution of who may have higher or lower
exposure potentials. Ritz (1999), the exception, found evidence of an exposure-response
relationship; mortality from hepatobiliary cancer was found to increase with degree and duration
of exposure and time since first exposure with a statistically significant but imprecise liver
6/22/2009
479
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
cancer risk for those with the highest exposure and longest time since first exposure. This
observation is consistent with association with TCE, but with uncertainty given one TCE
exposed case in the highest exposure group and correlation between TCE, cutting fluids, and
radiation exposures.
Observations in these studies provide some evidence of susceptibility of liver, gallbladder
and biliary tract; observations consistent with pharmacokinetic processing of TCE and the
extensive intra- and extra-hepatic recirculation of metabolites. Magnitude of risk of gallbladder
and biliary tract cancer is slightly higher than that for primary liver cancer in Raaschou-Nielsen
et al. (2003), the study with the most cases. Observations in Blair et al. (1998), Hansen et al.
(2001), and Radican et al. (2008), three smaller studies, suggest slightly larger risk ratios for
primary liver cancer compared to gallbladder and biliary tract cancer. Overall, these studies are
not highly informative for cross-organ comparison of relative magnitude of susceptibility.
The largest geographic studies (Morgan and Cassidy, 2002; Lee et al., 2003) are also
suggestive of association with the risk ratio (mortality odds ratio) in Lee et al. (2003) as
statistically significantly elevated. The geographic studies do not include a characterization of
TCE exposure to individual subjects other than residency in a community with groundwater
contamination by TCE with potential for exposure misclassification bias dampening
observations; these studies lack characterization of TCE concentrations in drinking water and
exposure characteristics such as individual consumption patterns. For this reason, observations
in Morgan and Cassidy (2002) and Lee et al. (2003) are noteworthy, particularly if positive bias
leading to false positive finding is considered minimal, and the lack of association with liver
cancer in the two other community studies (Vartiainen et al., 1993; ATSDR, 2006) does not
detract from Morgan and Cassidy (2002) or Lee et al. (2003). Lee et al. (2003), however, do not
address possible confounding related to hepatitis viral infection status, a risk factor for liver
cancer, or potential misclassification due to the inclusion of secondary liver cancer among the
case series, factors which may amplify observed association.
Meta-analysis is adopted as a tool for examining the body of epidemiologic evidence on
liver cancer and TCE exposure and to identify possible sources of heterogeneity. The meta-
analyses of the overall effect of TCE exposure on liver (and gall bladder/biliary passages) cancer
suggest a small, statistically significant increase in risk. The pooled estimate from the primary
random effects meta-analysis of the 9 (all cohort) studies is 1.36 (95% CI 1.10, 1.67). The study
of Raaschou-Nielsen et al. (2003) contributes almost 60% of the weight; its removal from the
analysis does not noticeably change the RRp estimate, but the estimate is no longer statistically
significant (RRp = 1.36; 95% CI 0.98, 1.89). The pooled estimate was not overly influenced by
any other single study, nor was it overly sensitive to individual RR estimate selections. There is
6/22/2009
480
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
no evidence of publication bias in this dataset, and no observable heterogeneity across the study
results.
Examination of sites individually (i.e., primary liver and intrahepatic bile ducts separate
from the combined liver and gallbladder/biliary passage grouping) resulted in the RRp estimate
for liver cancer alone (for the 3 studies for which the data are available; for the other studies,
results for the combined grouping were used) slightly lower than the one based entirely on
results from the combined cancer categories (1.32; 95% CI 1.02, 1.70). This result is driven by
the fact that the risk ratio estimate from the large Raaschou-Nielsen et al. (2003) study decreased
from 1.35 for liver and gall bladder/biliary passage cancers combined to 1.28 for liver cancer
alone.
The RRp estimate from the random effects meta-analysis of liver cancer in the highest
exposure groups in the 6 studies which provide risk estimates associated with highest exposure
primary liver cancer is 1.25 (95% CI 0.87, 1.79), slightly lower than the RRp estimate for liver
and gallbladder/biliary cancer and any TCE exposure of 1.34 (95% CI 1.09, 1.65), and not
statistically significant. Again, the RRp estimate of the highest-exposure groups is dominated by
one study (Raaschou-Nielsen et al., 2003). Two studies lacking reporting of liver cancer risk
associated with highest exposure and consideration of reporting bias in alternative meta-analyses
is similar to the estimated in the more restricted set of studies presenting risk ratios association
with highest exposure groups in published papers, 1.22 (95% CI; 0.87, 1.71).
Different exposure metrics are used in the various studies, and the purpose of combining
results across the different highest exposure groups is not to estimate a RRp associated with
some level of exposure, but rather to examine impacts of combining RR estimates that should be
less affected by exposure misclassification. In other words, the highest exposure category is
more likely to represent a greater differential TCE exposure compared to people in the referent
group than the exposure differential for the overall (typically any versus none) exposure
comparison. Thus, if TCE exposure increases the risk of liver and gallbladder/biliary cancer, the
effects should be more apparent in the highest exposure groups. The findings of a lower RRp
associated with highest exposure group reflects observations in Blair et al. (1998) and Raaschou-
Nielsen et al. (2003), the study contributing greatest weight to the meta-analysis, that RR
estimates for the highest-exposure groups, although greater than 1.0, are less than the RR
estimates with any TCE exposure.
Thus, while the finding of an elevated and statistically significant RRp for liver and
gallbladder/biliary cancer and any TCE exposure provides evidence of association, the statistical
significance of the pooled estimates is dependent on one study, which provides the majority of
the weight in the meta-analyses. Furthermore, combining results from the highest-exposure
groups yields lower RRp estimates than for an overall effect. These results do not rule out an
6/22/2009
481
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
effect of TCE on liver cancer, because the liver cancer results are relatively underpowered with
respect to numbers of studies and number of cases; overall, the meta-analysis provides only
minimal support for association between TCE exposure and liver and gallbladder/biliary cancer.
NRC (2006) deliberations on trichloroethylene commented on two prominent evaluations
of the then-current TCE epidemiologic literature using meta-analysis techniques, Wartenberg et
al. (2000) and Kelsh et al. (2005), submitted by Exponent-Health Sciences to NRC during their
deliberations and published afterwards in the open literature as Alexander et al. (2007) with the
substitution of the recently published study of Boice et al. (2006) for Ritz (1999) which Kelsh et
al. (2005) included in their NRC presentation. NRC (2006) found weaknesses in the techniques
used in Wartenberg et al. (2000) and the Exponent analyses. EPA staff conducted their analysis
according to NRC (2006) suggestions for transparency, systematic review criteria, and
examination of both cohort and case-control studies. The EPA analysis of liver cancer
considered a similar set of studies as Alexander et al. (2007) although treatment of these studies
differs between analyses. Alexander et al. (2007) present many pooled relative risk estimates,
grouping of studies with differing exposure potentials, for example, including the large cohort of
Boice et al. (1999) of 45,323 subject identified with TCE exposure with biomarker studies
(Axelson et al., 1994; Anttila et al., 1995; Hansen et al., 2001) in one analysis; yet, in other
analyses, including the TCE subcohort (2,267 subjects or 3% of the larger cohort) of Boice et al.
(1999) with the biomarker studies. Additionally, Alexander et al. (2007) lacks quantitative
examination of liver cancer risk in the higher TCE exposure groups even though a meta-analysis
of NHL of the same studies as analyzed by Alexander et al. (2007) and from the same group of
investigators, Mandel et al. (2006), contains such an analysis. Alexander et al. (2007) lacks
discussion of their rationale for different treatment of subjects from a same study and their basis
for grouping studies with subjects of different exposure potentials. The inclusion of subjects
with little to no TCE exposure over background levels has the potential to introduce
misclassification bias and dampen observed risk ratios. Another difference between the EPA and
previous meta-analyses is their inclusion of Ritz (1999), included in Wartenberg et al. (2000) and
Kelsh et al. (2005). Despite the weaknesses in past meta-analyses, pooled liver and gall
bladder/biliary tract cancer risk estimates for overall TCE exposure for TCE subcohorts is of a
similar magnitude as that observed in EPA's updated and expanded analysis, Kelsh et al. (2005),
1.32 (95% CI: 1.05, 1.66) and Alexander et al. (2007), 1.30 (95% CI: 1.09-1.55).
6/22/2009
482
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Table 4.4.1. Summary of human liver toxicity studies
Subjects
Effect
Exposure
Reference
148 male metal degreasers in
metal parts, semiconductor
and other factories
Serum liver function
enzyme (HDL-C, AST and
GGT) concentrations did
not correlated with TCE
exposure assesses in a
prevalence study but did
correlate with TCE
concentration over a 2-year
follow-up period
U-TTC levels obtained from
spot urine sample obtained
during working hours used to
assign exposure category:
High: 209 + 99 mg/g Cr
Medium: 35 +27 mg/g Cr
Low: 5 + 2 mg/g Cr
Note: this study does not
include an unexposed
referent group
Nagaya et al.,
1993
95 workers (70 TCE exposed,
25 CFC 113 exposed) selected
from a cohort of 240 workers
at 72 factors engaged in metal
degreasing with chlorinated
solvents
Increased serum GGT
concentration with
increasing cumulative
exposure
4 groups (cumulative number
of years exposed over a
working life):
I: 0.6 (0-0.99)
II: 1.9 (1-2.8)
III: 4.4 (2.9-6.7)
IV: 14.4 (6.8-35.6)
Rasmussen et al.,
1993
21 metal degreasers with
severe hypersensitivity
dermatitis
High prevalence of serum
liver function enzymes
above normal levels: ALT,
19 or 21 cases; AST, 18 of
21 cases, and T-Bili, 16 of
21 cases.
TWA mean ambient TCE
concentration occupational
setting of cases, 18 mg/m3to
683 mg/m3
14 of 21 cases with U-TCE
above recommended
occupational level of 50
mg/L
Xu et al., 2009
5 healthy workers engaged in
decreasing activities in steel
industry and 5 healthy
workers from clerical section
of same company
Total serum bile acid
concentration increased
between pre- and post-
exposure (2-day period)
8-hour TWA mean personal
air: 8.9 + 3.2 ppmpost-
exposure
Neghab et al.,
1997
22 workers at a factory
manufacturing small
appliances
Increased in several bile
acids
Regular exposure to <5 ppm
TCE; peak exposure for 2
workers to >250 00m
Driscoll et al.,
1992
4,489 males and female
residents from 15 Superfund
site and identified from
ATSDR Trichloroethylene
Liver problems diagnosed
with past year
Residency in community
with Superfund site identified
with TCE and other
chemicals
Davis et al., 2006
6/22/2009
483
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Exposure Subregistry
Case reports from 8 countries
of individuals with
idiosyncratic generalized skin
disorders
Hepatitis in 46% to 94% of
cases; other liver effects
includes hepatomegaly and
elevated liver function
enzymes; and in rare cases,
acute liver failure
If reported, TCE, from < 50
mg/m3 to more than 4,000
mg/m3.
Symptoms developed within
2-5 weeks of initial
exposure, with some
intervals up to 3 months.
Kamijima et al.,
2007
Deaths in California between
1979-1981 due to cirrhosis
SMR of 211 (95% CI: 136,
287) for white male sheet
metal workers and
SMR = 174 (95% CI:
150-197) for metal workers
Occupational title on death
certificate
Leigh and Jiang,
1993
Table 4.4.2. Selected Results from Epidemiologic Studies of TCE Exposure and Cirrhosis
Study No. obs.
Population Exposure Group Relative Risk (95% CI) events Reference
Cohort-Mortality
Aerospace workers (Rocketdyne)
Any TCE (utility/eng flush) 0.39 (0.16, 0.80) 7
Boice et al., 2006
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
Not reported
Zhao et al., 2005
View-Master workers
Electronic workers (Taiwan)
Males 0.76(0.16,2.22)
Females 1.51 (0.72,2.78)
Primary Liver, males Not reported
Primary Liver, females Not reported
3
10
ATSDR, 2003, 2004
Chang et al., 2005,
2003
US Uranium-processing workers
Any TCE exposure 0.91 (0.63,1.28)
Light TCE exposure, >2 years duration Not reported
Mod TCE exposure, >2 years duration Not reported
33
Ritz, 1999
Aerospace workers (Lockheed)
TCE Routine Exp
TCE Routine-Intermittent
0.61 (0.39,0.91)
0 years 1.001
23
22
Boice et al., 1999
6/22/2009
484
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Study
No. obs.
Population Exposure Group
Relative Risk (95% CI)
events
Reference
Cohort-Mortality
Any exposure
Not reported
13
Aerospace workers (Hughes)
TCE Subcohort
0.55 (0.30, 0.93)
14
Morgan etal., 1998,
Low Intensity (<50 ppm)5
0.95 (0.43, 1.80)
9
2000
High Intensity (>50 ppm)5
0.32(0.10,0.74)
5
Aircraft maintenance workers (Hill AFB, Utah)
TCE Subcohort
1.1 (0.6, 1.9)1
44
Blair etal., 1998
Males, Cumulative exp
0
1.01
< 5 ppm-yr
0.6 (0.2, 1.3)
10
5-25 ppm-yr
0.8 (0.3, 1.9)
9
>25 ppm-yr
1.2 (0.6, 2.4)
17
Females, Cumulative exp
0
1.01
< 5 ppm-yr
2.4(1.4, 13.7)
6
5-25 ppm-yr
1.8 (0.2, 15.0)
1
>25 ppm-yr
0.6 (0.1,4.8)
1
TCE Subcohort
1.04 (0.56, 1.93)1 2
37
Radican et al., 2008
Males, Cumulative exp
0.87 (0.43, 1.73)
31
0
1.01'2
< 5 ppm-yr
0.56 (0.23, 1.40)
8
5-25 ppm-yr
1.07 (0.45, 2.53)
10
>25 ppm-yr
1.06 (0.48, 2.38)
13
Females, Cumulative exp
1.79 (0.54, 5.93)
6
0
1.01'2
< 5 ppm-yr
3.30 (0.88, 12.41)
4
5-25 ppm-yr
2.20 (0.26, 18.89)
1
>25 ppm-yr
0.59 (0.97, 5.10)
1
Deaths reported to GE pension fund (Pittsfield, MA)
Not reported
Greenland et al., 1994
Aircraft manufacturing plant employees (Italy)
Costa etal., 1989
All subjects
Not reported
Aircraft manufacturing plant employees (San Diego,
CA)
Garabrant et al., 1988
All subjects
0.86 (0.67, 1.11)
63
6/22/2009
485
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 1 Referent group are subjects from the same plant or company, or internal referents.
2 2 Numbers of cirrhosis deaths in Radican et al. (2009) are fewer than Blair et al. (1998) because Radican et al.
3 (2008) excluded cirrhosis deaths due to alcohol.
6/22/2009
486
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.4.3: Selected Results from Epidemiologic Studies of TCE Exposure and Liver Cancer
Study
Relative Risk (95%
No. obs.
Relative Risk (95%
No. obs.
Relative Risk (95%
No. obs.
Population Exposure Group
CI)
events
CI)
events
CI)
events
Reference
Liver and Intrahepatic Bile Ducts
Primary Liver
Gallbladder and Extrahepatic Bile Ducts
Cohort Studies - Incidence
Aerospace workers (Rocketdyne)
Low cum TCE score
Not reported
Zhao et al..
Med cum TCE score
Not reported
High TCE score
Not reported
p for trend
Raaschou-
Danish blue-collar workers w/ TCE exposure
2003
Males + Females
1.3 (1.0, 1.6)1
82
Males + Females
1.4(1.0, 1.8)2
57
Males, Any exposure
1.1 (0.8, 1.5)2
41
1.1 (0.7, 1.6)
27
1.1 (0.6, 1.9)
14
<1 year employment duration
1.2 (0.7, 2.1)2
13
1.3 (0.6, 2.5)
9
1.1 (0.3,2.9)
4
1-4.9 years employment duration
0.9 (0.5, 1.6)2
13
1.0 (0.5, 1.9)
9
0.8(0.2,2.1)
4
>5 years employment duration
1.1 (0.6, 1.7)2
15
1.1 (0.5,2.1)
9
1.4(0.5,3.1)
6
Females, Any exposure
2.8 (1.6, 4.6)2
16
2.8(1.1,5.8)
7
2.8(1.3,5.3)
9
< 1 year employment duration
2.5 (0.7, 6.5)2
4
2.8 (0.3, 10.0)
2
2.3 (0.3, 8.4)
2
1-4.9 years employment duration
4.5 (2.2, 8.3)2
10
4.1 (1.1, 10.5)
4
4.8(1.7, 10.4)
6
> 5 years employment duration
1.1 (0.1, 3.8)2
2
1.3 (0.0,7.1)
1
0.9 (0.0, 5.2)
1
Biologically-monitored Danish workers
Hansen et;
Males + Females
2.1 (0.7, 5.0)2
5
1.7 ((0.2, 6.0)
2
2.5(0.5,7.3)
3
Males
2.6 (0.8, 6.0)2
5
1.8 (0.2, 6.6)
2
3.3 (0.7, 9.7)
3
0 (0.3
Females
0 (0.4 exp)
0 (0.1 exp)
exp)
6/22/2009
487
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Study
Population
Relative Risk (95% No. obs.
Exposure Group CI)
Cumulative exp (Ikeda) Not reported
<17 ppm-yr
>17 ppm-yr
Mean concentration (Ikeda) Not reported
<4 ppm
4+ ppm
Employment duration Not reported
<6.25 yr
>6.25
events
Aircraft maintenance workers from Hill Air Force Base
TCE Subcohort
Males, Cumulative exp
Not reported
Females, Cumulative exp
0
1.03
< 5 ppm-yr
0.6 (0.1, 3.1)2
3
5-25 ppm-yr
0.6 (0.1,3.8)
2
>25 ppm-yr
1.1 (0.2,4.8)
4
0
1.89 (0.86, 3.59)'
Biologically-monitored Finnish workers
All subjects
Mean air-TCE (Ikeda extrapolation from U-TCA)
< 6 ppm Not reported
6+ ppm
Biologically-monitored Swedish workers
Males 1.41 (0.38, 3.60)2
Females Not reported
6/22/2009
Relative Risk
CI)
(95%
No. obs. Relative Risk (95%
events CI)
No. obs.
events Reference
Blair etal., 1998
Not reported
1.03
1.2(0.1,2.1) 2
1.0(0.1,16.7) 1
2.6(0.3,25.0) 3
0
Anttila et al., 1995
2.27 (0.74,5.29) 5 1.56 (0.43,4.00) 4
1.64 (0.20, 5.92) 2
2.74 (0.33, 9.88) 2
Axelson et al., 1994
488
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Study
Population
Exposure Group
Relative Risk (95% No. obs.
CI) events
Relative Risk (95% No. obs.
CI) events
Relative Risk (95% No. obs.
CI) events Reference
Cohort-Mortality
Computer manufacturing workers (IBM), NY
Not reported
Clapp and Hoffman, 20(
Aerospace workers (Rocketdyne)
Any TCE (utility/eng flush)
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
1.28 (0.35, 3.27)
Not reported
Boice et al., 2006
Zhao et al., 2005
View-Master workers
Males 2.45 (0.50, 7.12)4
Females
1.01 (0.03, 5.63)
0 (2.61 exp)
1 8.41 (1.01,30.4)4 2 ATSDR, 2003,2004
0 (0.95
0(1.66 exp) exp)
Electronic workers (Taiwan)
Primary Liver, males
Primary Liver, females
Not reported
Not reported
0 (0.69 exp)
0 (0.57 exp)
Chang et al., 2005, 2003
US Uranium-processing workers
Any TCE exposure Not reported
Light TCE exposure, >2 years duration 0.93 (0.19, 4.53)5 3
Mod TCE exposure, >2 years duration 4.97 (0.48, 51.1)5 1
Light TCE exposure, >5 years duration 2.86 (0.48, 17.3)6 3
Mod TCE exposure, >5 years duration 12.1 (1.03, 144)6 1
Ritz, 1999
Aerospace workers (Lockheed)
6/22/2009
489
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Study
Population
Exposure Group
TCE Routine Exp
TCE Routine-Intermittent
Relative Risk (95% No. obs.
CI) events
0.54 (0.15, 1.38) 4
0 years
1.003
22
Any exposure
Not reported
13
< 1 year
0.53 (0.18, 1.60)
4
1-4 years
0.52 (0.15, 1.79)
3
> 5 years
0.94 (0.36, 2.46)
6
p for trend
>0.20
Aerospace workers (Hughes)
TCE Subcohort
0.98 (0.36,2.13)
6
Low Intensity (<50 ppm)5
1.32 (0.27, 3.85)
3
High Intensity (>50 ppm)5
0.78 (0.16,2.28)
3
TCE Subcohort (Cox Analysis)
Never exposed
1.003
14
Ever exposed
1.48 (0.56, 3.91)7 8
6
Cumulative
Low
2.12 (0.59, 7.66)8
3
High
1.19 (0.34, 4.16)8
3
Peak
No/Low
1.003
17
Med/Hi
0.98 (0.29, 3.35)8
3
Aircraft maintenance workers (Hill AFB, Utah)
TCE Subcohort
1.3 (0.5, 3.4)3
15
Males, Cumulative exp
0
1.03
< 5 ppm-yr
1.1 (0.3,4.1)
6
5-25 ppm-yr
0.9 (0.2, 4.3)
3
6/22/2009
Relative Risk (95% No. obs.
CI) events
Relative Risk (95% No. obs.
CI) events Reference
Boice et al., 1999
1.7 (0.2, 16.2)3
Morgan et al., 1998, 20C
Blair et al., 1998
490
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Study Relative Risk (95% No. obs.
Population Exposure Group CI) events
>25 ppm-yr 0.7(0.2,3.2) 3
Females, Cumulative exp
0
1.03
< 5 ppm-yr
1.6 (0.2, 18.2)
1
5-25 ppm-yr
0
>25 ppm-yr
2.3 (0.3, 16.7)
2
TCE Subcohort
1.12(0.57,2.19)3'9
31
Males, Cumulative exp
1.36 (0.59, 3.II)3
28
0
1.03
< 5 ppm-yr
1.17 (0.45,3.09)
10
5-25 ppm-yr
1.16 (0.39,3.46)
6
>25 ppm-yr
1.72 (0.68, 4.38)
12
Females, Cumulative exp
0.74 (0.18, 2.97)3
3
0
1.03
< 5 ppm-yr
0.69 (0.08, 5.74)
1
5-25 ppm-yr
0
>25 ppm-yr
0.98 (0.20, 4.90)
2
Deaths reported to GE pension fund (Pittsfield, MA)
0.54 (0.11, 2.63)10
9
Aircraft manufacturing plant employees (Italy)
All subjects 0.70(0.23,1.64)
Aircraft manufacturing plant employees (San Diego,
CA)
All subjects 0.94(0.40,1.86)
Community Studies
6/22/2009
Relative Risk (95%
CI)
No. obs.
events
Relative Risk (95% No. obs.
CI) events Reference
1.25 (0.31, 4.97)3 9 8 Radican et al., 2008
2.72 (0.34, 21.88)3 8
1.03
3.28 (0.37, 29,45) 4
0
4.05 (0.45, 36.41) 4
0
Greenland et al., 1994
Costa et al., 1989
Garabrant et al., 1988
491
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Study
Population Exposure Group
Residents in two study areas in Endicott, NY
Relative Risk (95% No. obs.
CI) events
0.71 (0.09,2.56) <6
Relative Risk (95% No. obs.
CI) events
Relative Risk (95% No. obs.
CI) events Reference
ATSDR, 2006
Residents of community with contaminated drinking water (Taiwan)
Village of residency, males
Upstream 1.00
Downstream 2.57 (1.21, 5.46)
26
Lee et al., 2003
Residents in 13 census tracts in Redland, CA
1.29 (0.74, 2.05) 28
Morgan and Cassidy, 20
Finnish residents
Residents of Hausjarvi
Residents of Huttula
0.7 6(0.3, 1.4)
0.6 (0.2, 1.3)
Vartiainen et al., 1993
1 ICD-7, 155 and 156; Primary liver (155.0), gallbladder and biliary passages (155.1), and liver secondary and unspecified (156)
2 ICD-7, 155; Primary liver, gallbladder and biliary passages
3 Internal referents, workers without TCE exposure
4 Proportional mortality ratio (PMR)
5 Logistic regression analysis with a 0-year lag for TCE exposure
6 Logistic regression analysis with a 15-year lag for TCE exposure
7 Risk ratio from Cox Proportional Hazard Analysis, stratified by age, sex and decade (Environmental Health Strategies, 1997)
8 Morgen et al. (1998) do not identify if SIR is for liver and biliary passage or primary liver cancer; identified as primary live in NRC (2006)
9 Radican et al. (2008) provide results for TCE exposure for follow-up through 1990, comparing the Poisson model rate ratios as reported by Blair et al. (1998) with Cox model hazard ratios.
Relative risk from Cox model adjusted for age and gender for liver and intrahepatic bile duct cancer was 1.2 (95% CI: 0.5, 3.4) and for primary liver cancer was 1.3 (95% CI: 0.1, 12.0).
10 Odds ratio
11 99% Confidence Intervals
6/22/2009
492
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE and liver cancer
Study name
Statistics for each study
Risk
Lower
Upper
ratio
limit
limit
p-Value
Anttila 1995
1.890
0.983
3.632
0.056
Axelson 1994
1.410
0.529
3.757
0.492
Blair 1998
1.300
0.499
3.390
0.592
Boice 1999
0.540
0.203
1.439
0.218
Boice 2006
1.280
0.480
3.410
0.622
Greenland 1994
0.540
0.110
2.640
0.447
Hansen 2001
2.100
0.874
5.045
0.097
Morgan 1998 unpub RR
1.481
0.561
3.909
0.428
Raaschou-Nielsen 2003
1.350
1.030
1.770
0.030
1.355
1.100
1.670
0.004
Risk ratio and 95% CI
0.1 0.2
0.5
10
random effects model: same for fixed
Figure 4.4.1. Meta-analysis of liver and biliary tract cancer and overall TCE exposure (The pooled estimate is in the bottom row. Symbol sizes reflect
relative weights of the studies. The horizontal midpoint of the bottom diamond represents the pooled RR estimate and the horizontal extremes depict the 95%
CI limits.)
6/22/2009
493
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.4.3 Experimental studies of TCE in rodents - introduction
The previous sections have described available human data for TCE-induced non-cancer
effects (e.g., disturbances in bile production) and whether an increased risk of liver cancer in
humans has been established from analysis of the epidemiological literature. A primary concern
for effects on the liver comes from a large database in rodents indicating that, not only TCE, but
a number of its metabolites are capable of inducing hepatocellular adenomas and carcinomas in
rodent species. Thus, many of rodent bioassays have focused on the study of liver cancer for
TCE and its metabolites and possible early effects specifically that may be related to tumor
induction.
This section describes the hazard data for TCE effects in the rodent liver and inferences
from studies of its metabolites. For more detailed descriptions of the issues providing context for
this data in terms the state of the science of liver physiology (Section 1), cancer (Section 3), liver
cancer (Section 3), and the MOA of liver cancer and other TCE-induced effects (Section 3.4),
please see Appendix E. A more comprehensive review of individual studies of TCE-induced
liver effects in laboratory animals is also provided in Section 2 of Appendix E that includes
detailed analyses of the strengths and the limitations of these studies. Issues have been raised
regarding the relevance of mouse liver tumor data to human liver cancer risk that are addressed
in sections 3.2 and 3.3 in Appendix E. Given that activation of the PPARa receptor has received
great attention as a potential MOA for TCE induced liver tumors, the current status of that
hypothesis is reviewed in Section 3.4.1 of Appendix E. Finally, comparative studies of TCE
metabolites and the similarities and differences of such study results are described in summary
sections of Appendix E (i.e., Section 2.4) as well as discussions of proposed MOAs for TCE-
induced liver cancer (i.e., Sections 2.4 and 3.4.2).
A number of acute and subchronic studies have been undertaken to describe the early
changes in the rodent liver after TCE administration with the majority using the oral gavage
route of administration. Several key issues affect the interpretation of this data. The few
drinking water studies available for TCE have recorded significant loss of TCE through
volatilization in drinking water solutions and thus this route of administration is generally not
used. Some short-term studies of TCE have included detailed examinations while others have
reported primarily liver weight changes as a marker of TCE response. The matching and
recording of age, but especially initial and final body weight, for control and treatment groups is
of particular importance for studies using liver weight gain as a measure of TCE response as
differences in these parameters affect TCE-induced liver weight gain. Most data are for TCE
exposures of at least 10 days to 42 days. For many of the subchronic inhalation studies
6/22/2009
494
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(Kjellstrand et al., 1981, 1983a, b), issues associated with whole body exposures make
determination of dose levels more difficult. The focus of the long-term studies of TCE is
primarily detection and characterization of liver tumor formation.
For gavage experiments, death due to gavage errors and specifically from use of this
route of administration, especially at higher TCE exposure concentrations, has been a recurring
problem, especially in rats. Unlike inhalation exposures, the effects of vehicle can also be an
issue for background liver effects in gavage studies. Concerns regarding effects of oil vehicles,
especially corn oil, have been raised (Kim et al., 1990; Charbonneau et al., 1991). Several oral
studies in particular document that use of corn oil as the vehicle for TCE gavage dosing induces
a different pattern of toxicity, especially in male rodents (see Merrick et al., 1989, Section 2.2.1.
of Appendix E). Several studies also report the effects of corn oil on hepatocellular DNA
synthesis and indices of lipid peroxidation (Channel et al., 1998; Rusyn et al., 1999). For
example, Rusyn et al. (1999) report that a single dose of dietary corn oil increases hepatocyte
DNA synthesis 24 hours after treatment by -3.5 fold of control, activates of NF-kB to a similar
extent ~ 2 hours after treatment almost exclusively in Kupffer cells, and induces a -3-4 fold
increase of control NF-kB in hepatocytes after 8 hours and an increase in TNFa mRNA between
8 and 24 hours after a single dose in female rats.
In regard to studies that have used the i.p. route of administration, as noted by Kawamoto
et al. (1988), injection of TCE may result in paralytic ileus and peritonitis and that subcutaneous
treatment paradigm will result in TCE not immediately being metabolized but retained in the
fatty tissue. Wang and Stacey (1990) state that "intraperitoneal injection is not particularly
relevant to humans" and suggest that intestinal interactions require consideration in responses
such as increase serum bile acid.
While studies of TCE metabolites have been almost exclusively conducted via drinking
water, and thus have avoided vehicle effects and gavage error, they have issues of palatability at
high doses and decreased drinking water consumption as a result that not only raises issues of the
resulting internal dose of the agent but also of effects of drinking water reduction.
Although there are data for both mice and rats for TCE exposure and studies of its
metabolites, the majority of the available information has been conducted in mice. This is
especially the case for long-term studies of DCA and TCA in rats. There is currently one study
each available for TCA and DCA in rats and both were conducted with such few numbers of
animals that the ability to detect and discern whether there was a treatment-related effect are very
limited (DeAngelo et al., 1997, 1996; Richmond et al., 1995).
With regard to the sensitivity of studies used to detect a response, there are issues
regarding not only the number of animals used but also the strain and weight of the animals. For
some studies of TCE strains were used that have less background rate of liver tumor
6/22/2009
495
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
development and carcinogenic response. As for the B6C3F1 mouse, the strain most used in the
bioassays of TCE metabolites, the susceptibility of the B6C3F1 to hepatocarcinogenicity has
made the strain a sensitive biomarker for a variety of hepatocarcinogens. Moreover, Leakey et
al. (2003b) demonstrated that increased body weight at 45 weeks of life is an accurate predictor
of large background tumor rates. Unfortunately a 2-year study of chloral hydrate (George et al.,
2000) and the only available 2-year study of TCA (DeAngelo et al., 2008), which used the same
control animals, were both conducted in B6C3F1 mice that grew very large (-50 g) and prone to
liver cancer (64% background incidence of hepatocellular adenomas and carcinomas) and
premature mortality. Thus, these bioassays are of limited value for determination of the dose-
response for carcinogenicity.
Finally, as discussed below, the administration of TCE to laboratory animals as well as
environmental exposure of TCE in humans are effectively co-exposure studies. TCE is
metabolized to a number of hepatoactive as well as hepatocarcinogenic agents. A greater
variability of response is expected than from exposure to a single agent making it particularly
important to look at the TCE database in a holistic fashion rather than the results of a single
study, especially for quantitative inferences. This approach is particularly useful given that the
number of animals in treatment groups in a variety of TCE and TCE metabolite studies have
been variable and small for control and treatment groups. Thus, their statistical power was not
only limited for detection of statistically significant changes but also in many cases to be able to
determine whether there is not a treatment related effect (i.e. Type II error for power calculation).
Section 2.4.2 of Appendix E provides detailed analyses of the database for liver weight induction
by TCE and its metabolites in mice and the results of those analyses are described below.
Specifically, the relationship of liver weight induction, but also other endpoints such as
peroxisomal enzyme activation and increases in DNA synthesis to liver tumor responses are also
addressed as well.
4.4.4 TCE-induced liver non-cancer effects
A number of effects have been studied as indicators of TCE effects on the liver but also
as proposed events whose sequellae could be associated with resultant liver tumors after chronic
TCE exposure in rodents. Similar effects have been studied in rodents exposed to TCE
metabolites which may be useful for not only determining whether such effects are associated
with liver tumors induced by these metabolites but also if they are similar to what has been
observed for TCE.
6/22/2009
496
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.4.4.1 Liver weight
Increases in liver weight in mice, rats, and gerbils have been reported as a result of acute
and short-term, and sub-chronic TCE treatment by inhalation and oral routes of exposure (Nunes
et al., 2001; Tao et al., 2000, Tucker et al., 1982; Goldsworthy and Popp, 1987; Elcombe et al.,
1985; Dees and Travis, 1993; Nakajima et al., 2000; Berman et al., 1995; Melnick et al., 1987;
Laughter et al., 2004; Merrick et al., 1989; Goel et al., 1992; Kjellstrand et al., 1981, 1983a, b;
Buben and O'Flaherty, 1985). The extent of TCE-induced liver weight gain is dependent on
species, strain, gender, nutrition status, duration of exposure, route of administration, vehicle
used in oral studies, and the concentration of TCE administered. Of great importance to the
determination of the magnitude of response is whether the dose of TCE administered also affects
whole body weight, and thus liver weight and the % liver/body weight ratio. Therefore studies
which employed high enough doses to induce whole body weight loss generally showed a
corresponding decrease in percent liver/body weight at such doses and "flattening" of the dose-
response curve, while studies which did not show systemic toxicity reported liver/body weight
ratios generally proportional to dose. Chronic studies, carried out for longer durations, that
examine liver weight are few and often confounded by the presence of preneoplastic foci or
tumors that also affect liver weight after an extended period of TCE exposure. The number of
studies that examine liver weight changes in the rat are much fewer than for mouse. Overall, the
database for mice provides data for examination of the differences in TCE-induced effects from
differing exposure levels, durations of exposure, vehicle, strain, and gender. One study provided
a limited examination of TCE-induced liver weight changes in gerbils.
TCE-induced increases in liver weight have been reported to occur quickly. Kjellstrand
et al. (1981) reported liver weight increases after 2 days inhalation exposure in NMRI mice,
Laughter et al. (2004) reported increased liver weight in SV129 mice in their 3-days study (see
below), and Tao et al. (2000) reported a increased in % liver/body weight ratio in female
B6C3Flmice for after 5 days. Elcombe et al. (1985) and Dees and Travis (1993) reported
gavage results in mice and rats after 10 days exposure to TCE which showed TCE-induced
increases in liver weight. Tucker et al. (1982) reported that 14 days of exposure to 24 mg/kg and
240 mg/kg TCE via gavage to induce a dose-related increase in liver weight in male CD-I mice
but did not show the data.
For mice, the inhalation studies of Kjellstrand et al provided the most information on the
affect of duration of exposure, dose of exposure, strain tested, gender, initial weight, and
variability in response between experiments on TCE-induced liver weight increases. These
experiments also provided results that were independent of vehicle effect. Although the
determination of the exact magnitude of response is limited by experimental design, Kjellstrand
et al. (1981) reported that in NMRI mice, continuous TCE inhalation exposure induced increased
6/22/2009
497
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
% liver/body weight by 2 days and that by 30 days (the last recorded data point) the highest %
liver/body weight ratio was reported (~ 1.75-fold over controls) in both male and female mice.
Kjellstrand et al. (1983b) exposed seven different strains of mice (Wild, C57BL, DBA, B6CBA,
A/sn, NZB, NMRI) to 150 ppm TCE for 30 days and demonstrated that strain, gender, and
toxicity, as reflected by changes in whole body weight, affected the % liver/body weight ratios
induced by 30 days of continuous TCE exposure. In general for the 7 strains of mice examined,
female mice had the less variable increases in TCE-induce liver weight gain across duplicate
experiments than male mice. For instance, in strains that did not exhibit changes in body weight
(reflecting systemic toxicity) in either gender (wild-type and DBA), 150 ppm TCE exposure for
30 days induced 1.74-fold to 1.87-fold of control % liver/body weight ratios in female mice and
1.45-fold to 2.00-fold of control % liver/body weight ratios in male mice. The strain with the
largest TCE-induced increase in % liver/body weight increase was the NZB strain (~ 2.08-fold of
control for females and 2.34- to 3.57-fold of control for males). Kjellstrand et al. (1983b)
provided dose-response information for the NMRI strain of mice (A Swiss-derived strain) that
indicated dose-related increases in % liver/body weight ratios between 37 and 300 ppm TCE
exposure for 30 days. The 150-ppm dose was reported to induce a 1.66-fold and 1.69-fold
increases in % liver/body weight ratios in male and female mice, respectively. Interestingly,
they also reported similar liver weight increases among groups with the same cumulative
exposure, but with different daily exposure durations (1 hr/day at 3,600 ppm to 24 hr/day at
150 ppm for 30 days).
Not only have most gavage experiments have been carried out in male mice, which
Kjellstrand et al. (1983a) had demonstrated to have more variability in response than females,
but also vehicle effects were noted to occur in experiments that examined them. Merrick et al.
(1989) reported that corn oil induced a similar increase in % liver/body weight ratios in female
mice fed TCE in emulphor and corn oil 4 weeks, male mice TCE administered in the corn oil
vehicle induced a greater increase in liver weight than emulphor but less mortality at a high does.
Buben and O'Flaherty (1985) treated male, outbred Swiss-Cox mice for 6 weeks at doses
ranging from 100 to 3,200 mg/kg-d, and reported increased liver/body-weight ratios at all tested
doses (1.12- to 1.75-fold of controls). Given the large strain differences observed by Kjellstrand
et al. (1983b), the use of predominantly male mice, and the effects of vehicle in gavage studies,
inter-study variability in dose-response relationships is not surprising.
Dependence of PPARa activation for TCE-liver weight gain has been investigated in
PPARa null mice by both Nakajima et al. (2000) and Laughter et al. (2004). Nakajima et al.
(2000) reported that after 2 weeks of 750 mg/kg TCE exposure to carefully matched SV129 wild
type or PPARa-null male and female mice (n = 6 group), there was a reported 1.50-fold increase
in wild-type and 1.26-fold of control % liver/body weight ratio in PPARa-null male mice. For
6/22/2009
498
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
female mice, there was ~ 1.25-fold of control % liver/body weight ratios for both wild-type and
PPARa-null mice. Thus, TCE-induced liver weight gain was not dependent on a functional
PPARa receptor in female mice and some portion of it may have been in male mice. Both wild-
type male and female mice were reported to have similar increases in the number of peroxisome
in the pericentral area of the liver and TCE exposure and, although increased 2-fold, were still
only ~ 4% of cytoplasmic volume. Female wild type mice were reported to have less TCE-
induced elevation of very long chain acyl-CoA synthetase, D-type peroxisomal bifunctional
protein, mitochondrial trifunctional protein a subunits a and P, and cytochrome P450 4A1 than
males mice, even though peroxisomal volume was similarly elevated in male and female mice.
The induction of PPARa protein by TCE treatment was also reported to be slightly less in female
than male wild-type mice (2.17-fold vs. 1.44-fold of control induction, respectively). Thus,
differences between genders in this study were for increased liver weight were not associated
with differences in peroxisomal volume in the hepatocytes but there was a gender-related
difference in induction of enzymes and proteins associated with PPARa.
The study of Laughter et al. (2004) used SV129 wild type and PPARa-null male mice
treated with 3 daily doses of TCE in 0.1% methyl cellulose for either 3 days (1,500 mg/kg TCE)
or 3 weeks (0, 10, 50, 125, 500, 1,000, or 1,500 mg/kg TCE 5 days a week). However, the
paradigm is not strictly comparable to other gavage paradigms due to the different dose vehicle
and the documented impacts of vehicles such as corn oil on TCE-induced effects. In addition, no
initial or final body weights of the mice were reported and thus the influence of differences in
initial body weight on % liver/body weight determinations could not be ascertained. While
control wild type and PPARa-null mice were reported to have similar % liver/body weight ratios
(i.e., ~ 4.5%) at the end of the 3-day study, at the end of the 3-week experiment the % liver/body
weight ratios were reported to be larger in the control PPARa-null male mice (5.1%). TCE
treatment for 3 days was reported for % liver/body weight ratio to be 1.4-fold of control in the
wild type mice and 1.07-fold of control in the null mice. After 3 weeks of TCE exposure at
varying concentrations, wild-type mice were reported to have % liver/body weight ratios that
were within ~ 2% of control values with the exception of the 1,000 mg/kg and 1,500 mg/kg
treatment groups (~ 1.18- fold and 1.30- fold of control, respectively). For the PPARa-null
mice the variability in % liver/body weight ratios were reported to be greater than that of the
wild-type mice in most of the TCE groups and the baseline levels of % liver/body weight ratio
for control mice 1.16-fold of that of wild-type mice. TCE exposure was apparently more toxic in
the PPARa-null mice. Decreased survival at the 1,500 mg/kg TCE exposure level resulted in the
prevention of recording of % liver/body weight ratios for this group. At 1,000 mg/kg TCE
exposure level, there was a reported 1.10-fold of control % liver/body weight ratio in the
PPARa-null mice. None of the increases in % liver/body weight in the null mice were reported
6/22/2009
499
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
to be statistically significant by Laughter et al. (2004). However, the power of the study was
limited due to low numbers of animals and increased variability in the null mice groups. The %
liver/body weight ratio after TCE treatment reported in this study was actually greater in the
PPARa-null mice than the wild-type male mice at the 1,000 mg/kg TCE exposure level (5.6 ±
0.4 % vs. 5.2 ± 0.5%, for PPARa-null and wild-type mice, respectively) resulting in a 1.18-fold
of wild-type and 1.10-fold of PPARa-null mice. Although the results reported in Laughter et al.
(2004) for DC A and TCA were not conducted in experiments that used the same paradigm, the
TCE-induced increase in % liver/body weight more closely resembled the dose-response pattern
for DCA than for DCA wild-type SV129 and PPARa-null mice.
No study examined strain differences among rats, and cross-study comparisons are
confounded by heterogeneity in the age of animals, dosing regimen, and other design
characteristics that may affect the degree of response. For rats, TCE-induced % liver/body
weight ratios were reported to range from 1.16-fold to 1.46-fold of control values depending on
the study paradigm. The studies which employed the largest range of exposure concentrations
(Melnick et al., 1987; Berman et al., 1995) examined 4 doses in the rat. In general, there was a
dose-related increase in % liver/body weight in the rat, especially at doses that did not cause
concurrent decreased survival or significant body weight loss. For gerbils, Kjellstrand et al.
(1981) reported a similar value of - 1.25 fold of control % liver/body weight as for Sprague-
Dawley (CD) rats exposed to 150 ppm TCE continuously for 30 days. Woolhiser et al. (2006)
also reported inhalation TCE exposure to increase the % liver/body weight ratios in female
Sprague-Dawley rats although this strain appeared to be less responsive that others tested for
induction of hepatomegaly from TCA exposure and to also be less prone to spontaneous liver
cancer.
The size of the liver is under tight control and after cessation of a mitogenic stimulus or
one inducing hepatomegaly, the liver will return to its preprogrammed size (see Appendix E).
The increase in liver weight from TCE-exposure also appears to be reversible. Kjellstrand et al.
(1981) reported a reduction in liver weight gain increases after cessation of TCE exposure for 5
or 30 days in male and female mice. However, experimental design limitations precluded
discernment of the magnitude of decrease. Kjellstrand et al. (1983b) reported that mice exposed
to 150 ppm TCE for 30 days and then examined 120 days after the cessation of exposure, had
liver weights were 1.09-fold of control for TCE-exposed female mice and the same as controls
for TCE-exposed male mice. However, the livers were not the same as untreated liver in terms
of histopathology. The authors reported that "after exposure to 150 ppm for 30 days, followed
by 120 days of rehabilitation, the morphological picture was similar to that of the air-exposure
controls except for changes in cellular and nuclear sizes." Qualitatively, the reduction in liver
weight after treatment cessation is consistent with the report of Elcombe et al. (1985) in Alderly
6/22/2009
500
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Park mice. The authors report that the reversibility of liver effects after the administration of
TCE to Alderly Park mice for 10 consecutive days. Effects upon liver weight, DNA
concentration, and tritiated thymidine incorporation 24 and 48 hours after the last dose of TCE
were reported to still be apparent. However, 6 days following the last dose of TCE, all of these
parameters were reported to return to control values with the authors not showing the data to
support this assertion. Thus, cessation of TCE exposure would have resulted in a 75% reduction
in liver weight by 4 days in mice exposed to the highest TCE concentration. Quantitative
comparisons are not possible because Elcombe et al. (1985) did not report data for these results
(e.g., how many animals, what treatment doses, and differences in baseline body weights) and
such a large decrease in such a short period of time needs to be verified.
4.4.4.2 Cytotoxicity
Acute exposure to TCE appears to induce low cytotoxicity below sub-chronically lethal
doses. Relatively high doses of TCE appear necessary to induce cytotoxicity after a single
exposure with two available studies reported in rats. Okino et al. (1991) reported small increases
in the incidence of hepatocellular necrosis in male Wistar rats exposed to 2,000 ppm (8 hr) and
8,000 ppm (2 hr), but not at lower exposures. In addition, "swollen" hepatocytes were noted at
the higher exposure when rats were pre-treated with ethanol or Phenobarbital. Serum
transaminases increased only marginally at the 8,000-ppm exposure, with greater increases with
pre-treatments. Berman et al. (1995) reported hepatocellular necrosis, but not changes in serum
markers of necrosis, after single gavage doses of 1,500 and 5,000 mg/kg TCE in female F344
rats. However, they did not report any indications of necrosis after 14 days of treatment at
50-1,500 mg/kg/day nor the extent of necrosis.
At acute and sub-chronic exposure periods to multiple doses, the induction of
cytotoxicity, though usually mild, appears to differ depending on rodent species, strain, dosing
vehicle and duration of exposure, and the extent of reporting to vary between studies. For
instance, Elcombe et al. (1985) and Dees and Travis(1993), which used the B6C3F1 mouse
strain and corn oil vehicle, reported only slight or mild necrosis after 10 days of treatment with
TCE at doses up to 1,500 mg/kg/day. Elcombe et al. (1985) also reported cell hypertrophy in the
centrilobular region. Dees and Travis (1993) reported some loss of vacuolization in hepatocytes
of mice treated at 1,000 mg/kg/day. Laughter et al. (2004) reported that "wild-type" SV129 mice
exposed to 1,500 mg/kg TCE exposure for 3 weeks exhibited mild granuloma formation with
calcification or mild hepatocyte degeneration but gave not other details or quantitative
information as to the extent of the lesions or what parts of the liver lobule were affected. The
authors noted that "wild-type mice administered 1,000 and 1,500 mg/kg exhibited centrilobular
hypertrophy" and that "the mice in the other groups did not exhibit any gross pathological
6/22/2009
501
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
changes" after TCE exposure. Channel et al. (1998) reported no necrosis in B6C3F1 mice
treated by 400-1,200 mg/kg-d TCE by corn oil gavage for 2 days to 8 weeks.
However, as stated above, Merrick et al. (1989) reported that corn oil resulted in more
hepatocellular necrosis, as described by small focal areas of 3-5 hepatocytes, in male B6C3F1
mice than use of emulphor as a vehicle for 4-week TCE gavage exposures. Necrotic hepatocytes
were described as surrounded by macrophages and polymorphonuclear cells. The authors
reported that visible necrosis was observed in 30-40% of male mice administered TCE in corn
oil but not that there did not appear to be a dose-response. For female mice, the extent of
necrosis was reported to be 0 for all control and TCE treatment groups using either vehicle.
Serum enzyme activities for ALT, AST and LDH (markers of liver toxicity) showed that there
was no difference between vehicle groups at comparable TCE exposure levels for male or female
mice. Except for LDH levels in male mice exposed to TCE in corn oil there was not a
correlation with the extent of necrosis and the patterns of increases in ALT and AST enzyme
levels.
Ramdhan et al. (2008) assessed TCE-induced hepatotoxicity by measuring plasma ALT
and AST activities and histopathology in Sv/129 mice treated by inhalation exposure, which are
not confounded by vehicle effects. Despite high variability and only six animals per dose group,
all three measures showed statistically significant increases at the high dose of 2,000 ppm (8 hr/d
for 7 d), although a non-statistically significant elevation is evident at the low dose of 1,000 ppm.
Even at the highest dose, cytotoxicity was not severe, with ALT and AST measures increased 2-
fold or less and an average histological score less than 2 (range 0-4).
Kjellstrand et al. (1983b) exposed male and female NRMI mice to 150 ppm for 30 to 120
days. Kjellstrand et al. (1983b) reported more detailed light microscopic findings from their
study and stated that:
"After 150 ppm exposure for 30 days, the normal trabecular arrangement of the liver
cells remained. However, the liver cells were generally larger and often displayed a fine
vacuolization of the cytoplasm. The nucleoli varied slightly to moderately in size and
shape and had a finer, granular chromatin with a varying basophilic staining intensity.
The Kupffer cells of the sinusoid were increased in cellular and nuclear size. The
intralobular connective tissue was infiltrated by inflammatory cells. There was not sign
of bile stasis. Exposure to TCE in higher or lower concentrations during the 30 days
produced a similar morphologic picture. After intermittent exposure for 30 days to a time
weighted average concentration of 150 ppm or continuous exposure for 120 days, the
trabecular cellular arrangement was less well preserved. The cells had increased in size
and the variations in size and shape of the cells were much greater. The nuclei also
displayed a greater variation in basophilic staining intensity, and often had one or two
6/22/2009
502
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
enlarged nucleoli. Mitosis was also more frequent in the groups exposed for longer
intervals. The vacuolization of the cytoplasm was also much more pronounced.
Inflammatory cell infiltration in the interlobular connective tissue was more prominent.
After exposure to 150 ppm for 30 days, followed by 120 days of rehabilitation, the
morphological picture was similar to that of the air-exposure controls except for changes
in cellular and nuclear sizes."
Although not reporting comparisons between male and female mice in the results section
of the paper for TCE-induced histopathological changes, the authors stated in the discussion
section that "However, liver mass increase and the changes in liver cell morphology were similar
in TCE-exposed male and female mice." Kjellstrand et al. (1983b) did not present any
quantitative data on the lesions they describe, especially in terms of dose-response. Most of the
qualitative description presented was for the 150-ppm exposure level and the authors suggest that
lower concentrations of TCE give a similar pathology as those at the 150 ppm level, but do not
present data to support that conclusion. Although stating that Kupffer cells were reported to be
increased in cellular and nuclear size, no differential staining was applied light microscopy
sections to distinguish Kupffer from endothelial cells lining the hepatic sinusoid in this study.
Without differential staining such a determination is difficult at the light microscopic level.
Indeed, Goel et al. (1992) describe proliferation of "sinusoidal endothelial cells" after
1,000 mg/kg/day and 2,000 mg/kg/d TCE exposure for 28 days in male Swiss mice. They
reported that histologically, "the liver exhibits swelling, vacuolization, widespread
degeneration/necrosis of hepatocytes as well as marked proliferation of endothelial cells of
hepatic sinusoids at 1,000 and 2,000 mg/kg TCE doses." Only one figure is given, at the light
microscopic level, in which it is impossible to distinguish endothelial cells from Kupffer cells
and no quantitative measures or proliferation were examined or reported to support the
conclusion that endothelial cells are proliferating in response to TCE treatment. Similarly, no
quantitative analysis regarding the extent or location of hepatocellular necrosis was given. The
presence or absence of inflammatory cells were not noted by the authors as well. In terms of
white blood cell count, the authors note that it is slightly increased at 500 mg/kg/day but
decreased at 1,000 and 2,000 mg/kg/day TCE, perhaps indicating macrophage recruitment from
blood to liver and kidney, which was also noted to have pathology at these concentrations of
TCE.
The inflammatory cell infiltrates described in the Kjellstrand et al. (1983b) study are
consistent with invasion of macrophages and well as polymophorphonuclear cells into the liver,
which could activate resident Kupffer cells. Although not specifically describing the changes as
consistent with increased polyploidization of hepatocytes, the changes in cell size and especially
the continued change in cell size and nuclear staining characteristics after 120 days of cessation
6/22/2009
503
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
of exposure are consistent with changes in polyploidization induced by TCE. Of note is that in
the histological description provided by the authors, although vacuolization is reported and
consistent with hepatotoxicity or lipid accumulation, which is lost during routine histological
slide preparation, there is no mention of focal necrosis or apoptosis resulting from these
exposures to TCE.
Buben and O'Flaherty (1985) reported liver degeneration "as swollen hepatocytes" and to
be common with treatment of TCE to Male Swiss-Cox mice after 6 weeks. They reported that
"Cells had indistinct borders; their cytoplasm was clumped and a vesicular pattern was apparent.
The swelling was not simply due to edema, as wet weight/dry weight ratios did not increase."
Karyorrhexis (the disintegration of the nucleus) was reported to be present in nearly all
specimens and suggestive of impending cell death. No Karyorrhexis, necrosis, or polyploidy
was reported in controls, but a low score Karyorrhexis was given for 400 mg/kg TCE and a
slightly higher one given for 1,600 mg/kg TCE. Central lobular necrosis reported to be present
only at the 1,600 mg/kg TCE exposure level and assigned a low score. Polyploidy was described
as characteristic in the central lobular region but with low score for both 400 mg/kg and 1,600
mg/kg TCE exposures. The authors reported that "hepatic cells had two or more nuclei or had
enlarged nuclei containing increased amounts of chromatin, suggesting that a regenerative
process was ongoing" and that there were no fine lipid droplets in TCE exposed animals. The
finding of "no polyploidy" in control mouse liver in the study of Buben and O'Flaherty (1985) is
unexpected given that binucleate and polyploid hepatocytes are a common finding in the mature
mouse liver. It is possible that the authors were referring to unusually high instances of
"polyploidy" in comparison to what would be expected for the mature mouse. The score given
by the authors for polyploidy did not indicate a difference between the two TCE exposure
treatments and that it was of the lowest level of severity or occurrence. No score was given for
centrolobular hypertrophy although the DNA content and liver weight changes suggested a dose-
response. The "Karyrrhexis" described in this study could have been a sign of cell death
associated with increased liver cell number or dying of maturing hepatocytes associated with the
increased ploidy, and suggests that TCE treatment was inducing polyploidization. Consistent
with enzyme analyses, centrilobular necrosis was only seen at the highest dose and with the
lowest qualitative score, indicating that even at the highest dose there was little toxicity.
At high doses, Kaneko et al. (2000) reported sporadic necrosis in male Mrl-lpr/lpr mice,
which are "genetically liable to autoimmune disease", exposed to 500 to 2,000 ppm, 4 h/d, 6
d/week, for 8 weeks (n = 5). Dose-dependent mild inflammation and associated changes were
reported to be found in the liver. The effects on hepatocytes were reported to be minimal by the
authors with 500-ppm TCE inducing sporadic necrosis in the hepatic lobule. Slight mobilization
6/22/2009
504
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
and activation of sinusoid lining cells were also noted. These pathological features were
reported to increase with dose.
NTP (1990), which used the B6C3F1 mouse strain, reported centrilobular necrosis in
6/10 male and 1/10 female B6C3F1 mice treated at a dose of 6,000 mg/kg/day for up to 13
weeks (all the male mice and 8 of the 10 female mice died in the first week of treatment). At
3,000 mg/kg/day exposure level, although centrilobular necrosis was not observed, 2/10 males
had multifocal areas of calcification in their livers, which the authors suggest is indicative of
earlier hepatocellular necrosis. However, only 3/10 male mice at this dose survived to the end of
the 13 week study.
For the NTP (1990) 2-year study, B6C3F1 mice were reported to have no treatment-
related increase in necrosis in the liver. A slight increase in the incidence of focal necrosis was
noted TCE-exposed male mice (8% vs. 2%) with a slight reduction in fatty metamorphosis in
treated male mice (0 treated vs. 2 control animals) and in female mice a slight increase in focal
inflammation (29% vs. 19% of animals) and no other changes. Therefore this study did not show
concurrent evidence of liver toxicity with TCE-induced neoplasia after 2 years of TCE exposure
in mice.
For the more limited database in rats, there appears to be variability in reported TCE
induced cytotoxicity and pathology. Nunes et al. (2001) reported no gross pathological changes
in rats gavaged with corn oil or with corn oil plus 200 mg/kg TCE for 7 days. Goldsworthy and
Popp (1987) gave no descriptions of liver histology given in this report for TCE-exposed animals
or corn-oil controls. Kjellstrand et al. (1981) gave also did not give histological descriptions for
livers of rats in their inhalation study.
Elcombe et al. (1985) provided a description of the histopathology at the light
microscopy level in Osborne-Mendel rats, and Alderly Park rats exposed to TCE via gavage for
10 days. However, they did not provide a quantitative analysis or specific information regarding
the variability of response between animals within group and there was no indication by the
authors regarding how many rats were examined by light microscopy. H& E sections from
Osborne-Mendel rats were reported to show that:
"Livers from control rats contained large quantities of glycogen and isolated
inflammatory foci, but were otherwise normal. The majority of rats receiving 1500
mg/kg body weight TCE showed slight changes in centrilobular hepatocytes. The
hepatocytes were more eosinophilic and contained little glycogen. At lower doses these
effects were less marked and were restricted to fewer animals. No evidence of treatment-
related hepatotoxicity (as exemplified by single cell or focal necrosis) was seen in any rat
receiving TCE. H& E sections from Alderly Park Rats showed no signs of treatment-
6/22/2009
505
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
related hepatotoxicity after administration of TCE. However, some signs of dose-related
increase in centrilobular eosinophilia were noted."
Thus both mice and rats were reported to exhibit pericentral hypertrophy and eosinophilia
as noted from the histopathological examination in Elcombe et al. (1985).
Berman et al. (1995) reported that for female rats exposed to TCE for 14 days
hepatocellular necrosis was noted to occur in the 1,500 and 5,000 mg/kg groups in 6/7 and 6/8
female rats, respectively but not to occur in lower doses. The extent of necrosis was not noted by
the authors for the two groups exhibiting a response after 1 day of exposure. Serum enzyme
levels, indicative of liver necrosis, were not presented and because only positive results were
presented in the paper, presumed to be negative. Therefore, the extent of necrosis was not of a
magnitude to affect serum enzyme markers of cellular leakage.
Melnick et al. (1987) reported that the only treatment-related lesion observed
microscopically in rats from either dosed-feed or gavage groups was individual cell necrosis of
the liver with the frequency and severity of this lesion similar at each dosage levels of TCE
microencapsulated in the feed or administered in corn oil. The severity for necrosis was only
mild at the 2.2 and 4.8 g/kg feed groups and for the 6 animals in the 2.8 g/kg group corn oil
group. The individual cell necrosis was reported to be randomly distributed throughout the liver
lobule with the change to not be accompanied by an inflammatory response. The authors also
reported that there was no histologic evidence of cellular hypertrophy or edema in hepatic
parenchymal cells. Thus, although there appeared to be TCE-treatment related increases in focal
necrosis after 14 days of exposure, the extent was mild even at the highest doses and involved
few hepatocytes.
For the 13-week NTP study (1990), only control and high dose F344/N rats were
examined histologically. Pathological results were reported to reveal that 6/10 males and 6/10
female rats had pulmonary vasculitis at the highest concentration of TCE. This change was also
reported to have occurred in 1/10 control male and female rats. Most of those animals were also
reported to have had mild interstitial pneumonitis. The authors report that viral titers were
positive during this study for Sendai virus.
Kumar et al. (2001) reported that male Wistar rats exposed to 376 ppm, 4 hr/day, 5
days/wk for 8-24 weeks showed evidence of hepatic toxicity. The authors stated that, "after 8
weeks of exposure enlarged hepatocytes, with uniform presence of fat vacuoles were found in all
of the hepatocytes affecting the periportal, midzonal, and centriolobular areas, and fat vacuoles
pushing the pyknosed nuclei to one side of hepatocytes. Moreover, congestion was not
significant. After exposure of 12 and 24 weeks, the fatty changes became more progressive with
marked necrosis, uniformly distributed in the entire organ." No other description of pathology
was provided in this report. In regard to the description of fatty change, the authors only did
6/22/2009
506
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
conventional H&E staining of sections with no precautions to preserve or stain lipids in their
sections. However, as noted below, the NCI study also reports long-term TCE exposure in rats
to result in hepatocellular fatty metamorphosis. The authors provided a table with histological
scoring of simply + or - for minimal, mild or moderate effects and do not define the criteria for
that scoring. There is also no quantitative information given as to the extent, nature, or location
of hepatocellular necrosis. The authors report "no change was observed in glutamic oxoacetate
transaminase and glutamic pyruvated transaminase levels of liver in all the three groups. The
GSH level was significantly decreased while "total sulphydryl" level was significantly increased
during 8, 12, and 24 weeks of TCE exposure. The acid and alkaline phosphatases were
significantly increased during 8, 12, and 24 weeks of TCE exposure." The authors present a
series of figures that are poor in quality to demonstrate histopathological TCE-induced changes.
No mortality was observed from TCE exposure in any group despite the presence of liver
necrosis.
Thus in this limited database that spans durations of exposure from days to 24 weeks and
uses differing routes of administration, generally high doses for long durations of exposure are
required to induce hepatotoxicity from TCE exposure in the rat. The focus of 2-year bioassays in
rats has been the detection of a cancer response with little or no reporting of noncancer pathology
in most studies. Henschler et al. (1984) and Fukuda et al. (1983) do not report non-cancer
histopathology, but do both report rare biliary cell derived tumors in rats in relatively insensitive
assays. For male rats, noncancer pathology in the NCI (1976) study was reported to include
increased fatty metamorphosis after TCE exposure and angiectasis or abnormally enlarged blood
vessels. Angiectasis can be manifested by hyperproliferation of endothelial cells and dilatation
of sinusoidal spaces. For the NTP (1990) study there was little reporting of non-neoplastic
pathology or toxicity and no report of liver weight at termination of the study. In the NTP
(1988) study, the 2 year study of TCE exposure reported no evidence of TCE-induced liver
toxicity described as non-neoplastic changes in ACI, August, Marshal, and Osborne-Mendel rats.
Interestingly, for the control animals of these four strains there was, in general, a low background
level of focal necrosis in the liver of both genders. Obviously, the negative results in this
bioassay for cancer are confounded by the killing of a large portion of the animals accidently by
experimental error but TCE-induced overt liver toxicity was not reported.
In sum, the cytotoxic effects in the liver of TCE treatment appear include little or no
necrosis in the rodent liver, but rather, a number of histological changes such as mild focal
hepatocyte degeneration at high doses, cellular "swelling" or hypertrophy, and enlarged nuclei.
Histological changes consistent with increased polyploidization and specific descriptions of
TCE-induced polyploidization have been noted in several experiments. Several studies note
proliferation of nonparenchymal cells after TCE exposure as well. These results are more
6/22/2009
507
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
consistently reported in mice, but also have been reported in some studies at high doses in rats,
for which fewer studies are available. In addition, the increase in cellular and nuclear sizes
appeared to persist after cessation of TCE treatment. In neither rats nor mice is there evidence
that TCE treatment results in marked necrosis leading to regenerative hyperplasia.
4.4.4.3 Measures of DNA synthesis, cellular proliferation, and apoptosis
The increased liver weight observed in rodents after TCE exposure may result from either
increased numbers of cells in the liver, increased size of cells in the liver, or a combination of
both. Studies of TCE in rodents have studied whole liver DNA content of TCE-treated animals
to determine whether the concentration of DNA per gram of liver decreases as an indication of
hepatocellular hypertrophy (Buben and O'Flaherty, 1985; Dees and Travis, 1993; Elcombe et al.,
1985). While the slight decreases observed in some studies are consistent with hypertrophy, the
large variability in controls and lack of dose-response limits the conclusions that can be drawn
from these data. In addition, multiple factors beyond hypertrophy affect DNA concentration in
whole-liver homogenates, including changes in ploidy and the number of hepatocytes and non-
parenchymal cells.
The incorporation of tritiated thymidine or BrdU has also been analyzed in whole liver
DNA and in individual hepatocytes as a measure of DNA synthesis. Such DNA synthesis can
occur from either increased numbers of hepatocytes in the liver or by increased polyploidization.
Sections 1.1 of Appendix E describe polyploidization in human and rodent liver and its impacts
on liver function, while Sections 3.1.2. and 3.3.1.discuss issues of target cell identification for
liver cancer and changes in ploidy as a key even in liver cancer using animals models,
respectively. Along with changes in cell size (hypertrophy), cell number (cellular proliferation),
and the DNA content per cell (cell ploidy), the rate of apoptosis has also been noted or
specifically examined in some studies of TCE and its metabolites. All of these phenomena have
been identified in proposed hypotheses as key events possibly related to carcinogenicity. In
particular, changes in cell proliferation and apoptosis have been postulated to be part of the
MOA for PPARa-agonists by Klaunig et al. (2003) (see Section 3.4 of Appendix E).
In regard to early changes in DNA synthesis, the data for TCE is very limited Mirsalis et
al. (1989) reported measurements of in vivo-in vitro hepatocyte DNA repair and S-phase DNA
synthesis in primary hepatocytes from male Fischer-344 rats and male and female B6C3F1 mice
administered single doses of TCE by gavage in corn oil. They reported negative results 2-12
hours after treatment from 50-1,000 mg/kg TCE in rats and mice (male and female) for
unscheduled DNA synthesis and repair using 3 animals per group. After 24 and 48 hours of 200
or 1,000 mg/kg TCE in male mice (n = 3) and after 48 hours of 200 (n = 3) or 1,000 (n = 4)
mg/kg TCE in female mice, similar values of 0.30 to 0.69% of hepatocytes were reported as
6/22/2009
508
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
undergoing DNA synthesis in primary culture. Only the 1,000 mg/kg TCE dose in male mice at
48 hours was reported to give a result considered to be positive (~ 2.2% of hepatocytes) but no
statistical analyses were performed on these measurements. These results are limited by both the
number of animals examined and the relevance of the paradigm.
As noted above, TCE treatment in rodents has been reported to result in hepatocellular
hypertrophy and increased centrilobular eosinophilia. Elcombe et al. (1985) reported a small
decrease in DNA content with TCE treatment (consistent with hepatocellular hypertrophy) that
was not dose-related, increased tritiated thymidine incorporation in whole mouse liver DNA that
was that was treatment but not dose-related (i.e., a 2-, 2-, and 5-fold of control in mice treated
with 500, 1,000, and 1,500 mg/kg TCE), and slightly increased numbers of mitotic figures that
were treatment but not dose-related and not correlated with DNA synthesis as measured by
thymidine incorporation. Elcombe et al., reported no difference in response between 500 and
1,000 mg/kg TCE treatments for tritiated thymidine incorporation. Dees and Travis (1993) also
reported that incorporation of tritiated thymidine in DNA from mouse liver was elevated after
TCE treatment with the mean peak level of tritiated thymidine incorporation occurred at 250
mg/kg TCE treatment level and remaining constant for the 500 and 1,000 mg/kg treated groups.
Dees and Travis (1993) specifically report that mitotic figures, although very rare, were more
frequently observed after TCE treatment, found most often in the intermediate zone, and found in
cells resembling mature hepatocytes. They reported that there was little tritiated thymidine
incorporation in areas near the bile duct epithelia or close to the portal triad in liver sections from
both male and female mice. Channel et al. (1998) reported PCNA positive cells, a measure of
cells that have undergone DNA synthesis, was elevated only on day 10 (out of the 21 studied)
and only in the 1,200 mg/kg/day TCE exposed group with a mean of - 60 positive nuclei per
1,000 nuclei for 6 mice (~ 6%). Given that there was little difference in PCNA positive cells at
the other TCE doses or time points studied, the small number of affected cells in the liver could
not account for the increase in liver size reported in other experimental paradigms at these doses.
The PCNA positive cells as well as "mitotic figures" were reported to be present in centrilobular,
midzonal, and periportal regions with no observed predilection for a particular lobular
distribution. No data was shown regarding any quantitative estimates of mitotic figures and
whether they correlated with PCNA results. Thus, whether the DNA synthesis phases of the cell
cycle indicated by PCNA staining were indentifying polyploidization or increased cell number
cannot be determined.
For both rats and mice, the data from Elcombe et al. (1985) showed that tritiated
thymidine incorporation in total liver DNA observed after TCE exposure did not correlate with
mitotic index activity in hepatocytes. Both Elcombe et al. (1985) and Dees and Travis (1993)
reported a small mitotic indexes and evidence of periportal hepatocellular hypertrophy from TCE
6/22/2009
509
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
exposure. Neither mitotic index or tritiated thymidine incorporation data support a correlation
with TCE-induced liver weight increase in the mouse, but rather the increase to be most likely
due to hepatocellular hypertrophy. If higher levels of hepatocyte replication had occurred
earlier, such levels were not sustained by 10 days of TCE exposure. These data suggest that
increased tritiated thymidine levels were targeted to mature hepatocytes and in areas of the liver
where greater levels of polyploidization occur (see Section 1.1 of Appendix E). Both Elcombe et
al. (1985) and Dees and Travis (1993) show that tritiated thymidine incorporation in the liver
was ~ 2-fold greater than controls between 250-1,000 mg/kg TCE, a result consistent with a
doubling of DNA. Thus, given the normally quiescent state of the liver, the magnitude of this
increase over control levels, even if a result of proliferation rather than polyploidization, would
be confined to a very small population of cells in the liver after 10 days of TCE exposure.
Laughter et al. (2004) reported that there was an increase in DNA synthesis after aqueous
gavage exposure to 500 and 1,000 mg/kg TCE given as 3 boluses a day for 3 weeks with BrdU
given for the last week of treatment. An examination of DNA synthesis in individual
hepatocytes was reported to show that 1% and 4.5% of hepatocytes had undergone DNA
synthesis in the last week of treatment for the 500 and 1,000 mg/kg doses, respectively. Again,
this level of DNA synthesis is reported for a small percentage of the total hepatocytes in the liver
and not reported to be a result of regenerative hyperplasia.
Finally, Dees and Travis (1993) and Channel et al. (1998) reported evaluating changes in
apoptosis with TCE treatment. Dees and Travis (1993) enumerated identified by either
hematoxylin and eosin or feulgen staining in male and female mice after 10 days of TCE
treatment by. Only 0 or 1 apoptosis was observed per 100 high power (400x) fields in controls
and all dose groups except for those given 1,000 mg/kg-d, in which 8 or 9 apoptoses per 100
fields were reported. None of the apoptoses were in the intermediate zones where mitotic figures
were observed, and all were located near the central veins. This is the same region where one
would expect endogenous apoptoses as hepatocytes "stream" from the portal triad toward the
central vein (Schwartz-Arad, 1989). In addition, this is the same region where Buben and
O'Flaherty (1985) noted necrosis and polyploidy. By contrast Channel et al. (1998) reported no
significant differences in apoptosis at any treatment dose (400 to 1,200 mg/kg-d) examined after
any time from 2 days to 4 weeks.
4.4.4.4 Peroxisomal proliferation and related effects
Numerous studies have reported that TCE administered to mice and rats by gavage leads
to proliferation of peroxisomes in hepatocytes. Some studies have measured changes in the
volume and number of peroxisomes as measures of peroxisome proliferation while others have
measured peroxisomal enzyme activity such catalase and cyanide-insensitive palmitoyl-CoA
6/22/2009
510
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
oxidation (PCO). Like liver weight, the determination of a baseline level of peroxisomal
volume, number, or enzyme activity can be variable and have great effect on the ability to
determine the magnitude of a treatment-related effect.
Elcombe et al. (1985) reported increases in the percent of the cytoplasm occupied by
peroxisomes in B6C3F1 and Alderley Park mice treated for 10 days at 500 to 1,500 mg/kg-d.
Although the increase over controls appeared larger in the B6C3F1 strain, this is largely due to
the 2-fold smaller control levels in that strain, as the absolute percentage of peroxisomal volume
was similar between strains after treatment. All these results showed high variability, as
evidenced from the reported standard deviations. Channel et al. (1998) found a similar absolute
percentage of peroxisomal volume after 10 days treatment in the B6C3F1 mouse at 1,200 mg/kg-
d TCE but with the percentage in vehicle controls similar to the Alderley-Park mice in the
Elcombe et al. (1985) study. Interestingly, Channel et al. (1998) found that the increase in
peroxisomes peaked at 10 days, with lower values after 6 and 14 days of treatment.
Furthermore, the vehicle control levels also varied almost 2-fold depending on the number of
days of treatment. Nakajima et al. (2000), who treated male wild-type SV129 mice at 750
mg/kg-d for 14 days, found even higher baseline values for the percentage of peroxisomal
volume, but with an absolute level after treatment similar to that reported by Channel et al.
(1998) in B6C3F1 mice treated at 1,200 mg/kg/day TCE for 14 days. Nakajima et al. (2000) also
noted that the treatment-related increases were smaller for female wild-type mice, and that there
were no increases in peroxisomal volume in male or female PPARa-null mice, although vehicle
control levels were slightly elevated (not statistically significant). Only Elcombe et al. (1985)
examined peroxisomal volume in rats, and reported smaller treatment-related increases in two
strains (OM and AP), but higher baseline levels. In particular, at 1,000 mg/kg-d, after 10 days
treatment, the percent peroxisomal volume was similar in OM and AP rats, with similar control
levels as well. While the differences from treatment were not statistically significant, only five
animals were used in each group, and variability, as can be seen by the standard deviations, was
high, particularly in the treated animals.
The activities of a number of different hepatic enzymes have also been as markers for
peroxisome proliferation and/or activation of PPARa. The most common of these are catalase
and cyanide-insensitive palmitoyl-CoA oxidation (PCO). In various strains of mice (B6C3F1,
Swiss albino, SV129 wild-type) treated at doses of 500 to 2,000 mg/kg-d for 10 to 28 days,
increases in catalase activity have tended to be more modest (1.3- to 1.6-fold of control) as
compared to increases in PCO (1.4 to 7.9-fold of control) (Elcombe et al., 1985; Goel et al.,
1992; Goldsworthy and Popp, 1987; Laughter et al., 2004; Nakajima et al., 2000; Watanabe and
Fukui, 2000). In rats, Elcombe et al. (1985) reported no increases in catalase or PCO activity in
Alderley-Park rats treated at 1,000 mg/kg/day TCE for 10 days. In F344 rats, Goldsworthy and
6/22/2009
511
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Popp (1987) and Melnick et al. (1987) reported increases of up to 2-fold in catalase and 4.1-fold
in PCO relative to controls treated at 600 to 4,800 mg/kg-d for 10 to 14 days. The changes in
catalase were similar to those in mice at similar treatment levels, with 1.1- to 1.5-fold of control
enzyme activities at doses of 1,000 to 1,300 mg/kg-d (Elcombe et al., 1985; Melnick et al.,
1987). However, the changes in PCO were smaller, with 1.1- to 1.8-fold of control activity at
these doses, as compared to 6.3 to 7.9-fold of control in mice (Goldsworthy and Popp, 1987;
Melnick et al., 1987).
In SV129 mice, Nakajima et al. (2000) and Laughter et al. (2004) investigated the
dependence of these changes on PPARa by using a null mouse. Nakajima et al. reported that
neither male nor female wild-type or PPARa null mice had significant increases in catalase after
14 days of treatment at 750 mg/kg-d. However, given the small number of animals (4 per group)
and the relatively small changes in catalase observed in other (wild-type) strains of mice, this
study had limited power to detect such changes. Several other markers of peroxisome
proliferation, including acyl-CoA oxidase and CYP4A1 (PCO was not investigated), were
induced by TCE in male wild-type mice, but not in male null mice or female mice of either type.
Unfortunately, none of these markers have been investigated using TCE in female mice of any
other strain, so it is unclear whether the lack of response is characteristic of female mice in
general, or just in this strain. Interestingly, as noted above, liver/body weight ratio increases
were observed in both sexes of the null mice in this study. Laughter et al. (2004) only quantified
activity of the peroxisome proliferation marker PCO in their study, and found in null mice a
slight decrease (0.8-fold of control) at 500 mg/kg/day TCE and an increase (1.5-fold of control)
at 1,500 mg/kg/day TCE after 3 weeks of treatment, with neither statistically significant (4-5
mice per group). However, baseline levels of PCO were almost 2-fold higher in the null mice,
and the treated wild-type and null mice differed in PCO activity by only about 1.5-fold.
In sum, oral administration of TCE for up to 28 days causes proliferation of peroxisomes
in hepatocytes along with associated increases in peroxisomal enzyme activities in both mice and
rats. Male mice tend to be more sensitive in that at comparable doses, rats and female mice tend
to exhibit smaller responses. For example, for peroxisomal volume and PCO, the fold-increase
in rats appears to be lower by 3 to 6-fold than that in mice, but, for catalase, the changes were
similar between mice in F344 rats. No inhalation or longer-term studies were located, and only
one study examined these changes at more than one time-point. Therefore, little is known about
the route-dependence, time course, and persistence of these changes. Finally, two studies in
PPARa-null mice (Laughter et al., 2004; Nakajima et al., 2000) found diminished responses in
terms of increased peroxisomal volume and peroxisomal enzyme activities as compared to wild-
type mice, although there was some confounding due to baseline differences between null and
wild-type control mice in several measures.
6/22/2009
512
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.4.4.5 Oxidative stress
Several studies have attempted to study the possible effects of "oxidative stress" and
DNA damage resulting from TCE exposures. The effects of induction of metabolism by TCE, as
well as through co-exposure to ethanol, have been hypothesized to in itself increase levels of
"oxidative stress" as a common effect for both exposures (see Sections 3.4.2.3 and 4.2.4. of
Appendix E). Oxidative stress has been hypothesized to be a key event or MOA for peroxisome
proliferators as well, but has been found to neither be correlated with cell proliferation nor
carcinogenic potency of peroxisome proliferators (see Section 3.4.1.1 of Appendix E). As a
MOA, it is not defined or specific as the term "oxidative stress" is implicated as part of the
pathophysiologic events in a multitude of disease processes and is part of the normal physiologic
function of the cell and cell signaling.
In regard to measures of oxidative stress, Rusyn et al. (2006) noted that although an
overwhelming number of studies draw a conclusion between chemical exposure, DNA damage,
and cancer based on detection of 8-hydroxy-2' deoxyguanosine (8-OHdG), a highly mutagenic
lesion, in DNA isolated from organs of in vivo treated animals, a concern exists as to whether
increases in 8-OHdG represent damage to genomic DNA, a confounding contamination with
mitochondrial DNA, or an experimental artifact. As noted in Sections 2.1.1. and 2.2.11. of
Appendix E, studies of TCE which employ the i.p. route of administration can be affected by
inflammatory reactions resulting from that routes of administration and subsequent toxicity that
can involve oxygen radical formation from inflammatory cells. Finally, as described in Section
2.2.8 of Appendix E, the study by Channel et al. (1998) demonstrated that corn oil as vehicle had
significant effects on measures of "oxidative stress" such as thiobarbiturate acid-reactive
substances (TBARS).
The TBARS results presented by Channel et al. (1988) indicate suppression of TBARS
with increasing time of exposure to corn oil alone with data presented in such a way for 8-OHdG
and total free radical changes that the pattern of corn oil administration was obscured. It was not
apparent from that study that TCE exposure induced oxidative damage in the liver.
Toraason et al. (1999) measured 8-OHdG and a "free radical-catalyzed isomer of
arachidonic acid and marker of oxidative damage to cell membranes, 8-Epi-prostaglandin F2a
(8epiPGF)", excretion in the urine and TBARS (as an assessment of malondialdehyde and
marker of lipid peroxidation) in the liver and kidney of male Fischer rats exposed to single i.p.
injections in of TCE in Alkamuls vehicle. Using this paradigm, 500 mg/kg TCE was reported to
induce stage II anesthesia and a 1,000 mg/kg TCE to induce level III or IV (absence of reflex
response) anesthesia and burgundy colored urine with 2/6 rats at 24 hours comatose and
hypothermic. The animals were sacrificed before they could die and the authors suggested that
they would not have survived another 24 hours. Thus, using this paradigm there was significant
6/22/2009
513
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
toxicity and additional issues related to route of exposure. Urine volume declined significantly
during the first 12 hours of treatment and while water consumption was not measured, it was
suggested by the authors to be decreased due to the moribundity of the rats. Given that this study
examined urinary markers of "oxidative stress" the effects on urine volume and water
consumption, as well as the profound toxicity induced by this exposure paradigm, limit the
interpretation of the study. The issues of bias in selection of the data for this analysis, as well as
the issues stated above for this paradigm limit interpretation of this data while the authors
suggest that evidence of oxidative damage was equivocal.
4.4.4.6Bile production
Effects of TCE exposure in humans and in experimental animals is presented in Section
2.6 of Appendix E. Serum bile acids (SBA) have been suggested as a sensitive indicator of
hepatotoxicity to a variety of halogenated solvents with an advantage of increased sensitivity and
specificity over conventional liver enzyme tests that primarily reflect the acute perturbation of
hepatocyte membrane integrity and "cell leakage" rather than liver functional capacity (i.e.,
uptake, metabolism, storage, and excretion functions of the liver) (Bai et al., 1992b; Neghab et
al., 1997). While some studies have reported negative results, a number of studies have reported
elevated SBA in organic solvent-exposed workers in the absence of any alterations in normal
liver function tests. These variations in results have been suggested to arise from failure of some
methods to detect some of the more significantly elevated SBA and the short-lived and reversible
nature of the effect (Neghab et al., 1997). Neghab et al. (1997) have reported that occupational
exposure to l,l,2-trichloro-l,2,2-trifluoroethane and trichloroethylene has resulted in elevated
SBA and that several studies have reported elevated SBA in experimental animals to chlorinated
solvents such as carbon tetrachloride, chloroform, hexachlorobutadiene, tetrachloroethylene, 1,1,
1-trichloroethane, and trichloroethylene at levels that do not induce hepatotoxicity (Bai et al.,
1992a, b; Hamdan and Stacey, 1993; Wang and Stacey, 1990). Toluene, a non-halogenated
solvent, has also been reported to increase SBA in the absence of changes in other hepatobiliary
functions (Neghab and Stacey, 1997). Thus, disturbance in SBA appears to be a generalized
effect of exposure to chlorinated solvents and nonchlorinated solvents and not specific to TCE
exposure.
Wang and Stacey (1990) administered TCE in corn oil via i.p. injection to male Sprague-
Dawley rats with liver enzymes and SBA examined 4 hours after the last TCE treatment. The
limitations of i.p injection experiments have already been discussed. While reporting no overt
liver toxicity there was, generally, a reported dose-related increase in cholic acid,
chenodeoxycholic acid, deoxycholic acid, taurocholic acid, tauroursodeoxycholic acid with
cholic acid and taurochlolic acid increased at the lowest dose. The authors report that
6/22/2009
514
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
"examination of liver sections under light microscopy yielded no consistent effects that could be
ascribed to trichloroethylene." In the same study a rats were also exposed to TCE via and using
this paradigm, cholic acid and taurocholic acid were also significantly elevated but the large
variability in responses between rats and the low number of rats tested in this paradigm limit its
ability to determine quantitative differences between groups. Nevertheless, without the
complications associated with i.p. exposure, inhalation exposure of TCE at relatively low
exposure levels that were not associated with other measures of toxicity were associated with
increased SB A level.
Hamdan et al. (1993) administered TCE in corn oil (1 mmol/kg) in male Sprague Dawley
rats and followed the time-course of SBA elevation, TCE concentration, and trichloroethanol in
the blood up to 16 hours. Liver and blood concentration of TCE were reported to peak at 4 hours
while those of trichloroethanol peaked at 8 hours after dosing. TCE levels were not detectable
by 16 hours in either blood or liver while those of trichloroethanol were still elevated.
Elevations of SBA were reported to parallel those of TCE with cholic acid and taurochloate acid
reported to show the highest levels of bile acids. The authors state that liver injury parameters
were checked and found unaffected by TCE exposure but did not show the data. Thus, it was
TCE concentration and not that of its metabolite that was most closely related to changes in SBA
and after a single exposure and the effect appeared to be reversible. In an in vitro study by Bai
and Stacey (1993), TCE was studied in isolated rat hepatocytes with TCE reported to cause a
dose-related suppression of initial rates of cholic acid and taurocholic acid but with no significant
effects on enzyme leakage and intracellular calcium contents, further supporting a role for the
parent compound in this effect.
4 A A.1 Summary: TCE-induced non-cancer effects in laboratory animals
In laboratory animals, TCE leads to a number of structural changes in the liver, including
increased liver weight, small transient increases in DNA synthesis, cytomegaly in the form of
"swollen" or enlarged hepatocytes, increased nuclear size probably reflecting polyploidization,
and proliferation of peroxisomes. Liver weight increases proportional to TCE dose are
consistently reported across numerous studies, and appear to be accompanied by periportal
hepatocellular hypertrophy. There is also evidence of increased DNA synthesis in a small
portion of hepatocytes at around 10 days in vivo exposure. The lack of correlation of
hepatocellular mitotic figures with whole liver DNA synthesis or DNA synthesis observed in
individual hepatocytes supports the conclusion that cellular proliferation is not the predominant
cause of increased DNA synthesis. The lack of correlation of whole liver DNA synthesis and
those reported for individual hepatocytes suggests that nonparenchymal cells also contribute to
such synthesis. Indeed, nonparenchymal cell activation or proliferation has been noted in several
6/22/2009
515
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
studies. Moreover, the histological descriptions of TCE exposed liver are consistent with and in
some cases specifically note increased polyploidy after TCE exposure. Interestingly, changes in
TCE-induced hepatocellular ploidy, as indicated by histological changes in nuclei, have been
noted to remain after the cessation of exposure. In regard to apoptosis, TCE has been reported to
either not change apoptosis or to cause a slight increase at high doses. Some studies have also
noted effects from dosing vehicle alone (such as corn oil in particular) not only on liver
pathology, but also on DNA synthesis.
Available data also suggest that TCE is does not induce substantial cytotoxicity, necrosis,
or regenerative hyperplasia, as only isolated, focal necroses and mild to moderate changes in
serum and liver enzyme toxicity markers having been reported. Data on peroxisome
proliferation, along with increases in a number of associated biochemical markers, shows effects
in both mice and rats. These effects are consistently observed across rodent species and strains,
although the degree of response at a given mg/kg/d dose appears to be highly variability across
strains, with mice on average appearing to be more sensitive.
In addition, like humans, laboratory animals exposed to TCE have been observed to have
increased serum bile acids, though the toxicologic importance of these effects is unclear.
4.4.5 TCE-induced liver cancer in laboratory animals
For 2-year or lifetime studies of TCE exposure a consistent hepatocarcinogenic response
has been observed using mice of differing strains and genders and from differing routes of
exposure. However, some rat studies have been confounded by mortality from gavage error or
the toxicity of the dose of TCE administered. In some studies, a relative insensitive strain of rat
has been used. However, in general it appears that the mouse is more sensitive than the rat to
TCE-induced liver cancer. Three studies give results the authors consider to be negative for
TCE-induced liver cancer in mice, but have either design and/or reporting limitations, or are in
strains and paradigms with apparent low ability for liver cancer induction or detection.
4.4.5.1 Negative or inconclusive studies of mice and rats
Fukuda et al. (1983) reported a 104-week inhalation bioassay in female Crj:CD-l (ICR)
mice and female Crj:CD (SD) rats exposed to 0, 50, 150 and 450 ppm TCE (n = 50). There were
no reported incidences of mice or rats with liver tumors for controls indicative of relatively
insensitive strains and gender used in the study for liver effects. While TCE was reported to
induce a number of other tumors in mice and rats in this study, the incidence of liver tumors was
less than 2% after TCE exposure. Of note is the report of cystic cholangioma reported in 1 group
of rats.
6/22/2009
516
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Henschler et al. (1980) exposed NMRI mice and WIST random bred rats to 0, 100 and
500 ppm TCE for 18 months (n = 30). Control male mice were reported to have one
hepatocellular carcinoma and 1 hepatocellular adenoma with the incidence rate unknown. In the
100 ppm TCE exposed group, 2 hepatocellular adenomas and 1 mesenchymal liver tumor were
reported. No liver tumors were reported at any dose of TCE in female mice or controls. For
male rats, only 1 hepatocellular adenomas at 100 ppm was reported. For female rats no liver
tumors were reported in controls, but 1 adenoma and 1 cholangiocarcinoma was reported at 100
ppm TCE and at 500 ppm TCE, 2 cholangioadenomas, a relatively rare biliary tumor, was
reported. The difference in survival in mice, did not affect the power to detect a response, as was
the case for rats. However, the low number of animals studied, abbreviated exposure duration,
low survival in rats, and apparently low sensitivity of this paradigm (i.e., no background
response in controls) suggests a study of limited ability to detect a TCE carcinogenic liver
response. Of note is that both Fukuda et al. (1983) and Henschler et al. (1980) report rare biliary
cell derived tumors in rats in relatively insensitive assays.
Van Duuren et al. (1979), exposed mice to 0.5 mg/mouse to TCE via gavage once a week
in 0.1 mL trioctanion (n = 30). Inadequate design and reporting of this study limit that ability to
use the results as an indicator of TCE carcinogenicity.
The NCI (1976) study of TCE was initiated in 1972 and involved the exposure of
Osborn-Mendel rats to varying concentrations of TCE. A low incidence of liver tumors was
reported for controls and carbon tetrachloride positive controls in rats from this study. The
authors concluded that due to mortality, "the test is inconclusive in rats." They note the
insensitivity of the rat strain used to the positive control of carbon tetrachloride exposure.
The NTP (1990) study of TCE exposure in male and female F344/N rats, and B6C3F1
mice (500 and 1,000 mg/kg for rats) is limited in the ability to demonstrate a dose-response for
hepatocarcinogenicity. For rats, the NTP (1990) study reported no treatment-related non-
neoplastic liver lesions in males and a decrease in basophilic cytological change reported from
TCE-exposure in female rats. The results for detecting a carcinogenic response in rats were
considered to be equivocal because both groups receiving TCE showed significantly reduced
survival compared to vehicle controls and because of a high rate (e.g., 20% of the animals in the
high-dose group) of death by gavage error.
The NTP (1988) study of TCE exposure in four strains of rats to "diisopropylamine-
stabilized TCE" was also considered inadequate for either comparing or assessing TCE-induced
liver carcinogenesis in these strains of rats because of chemically induced toxicity, reduced
survival, and incomplete documentation of experimental data. TCE gavage exposures of 0, 500
or 1,000 mg/kg per day (5 days per week, for 103 weeks) male and female rats was also marked
6/22/2009
517
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
by a large number of accidental deaths (e.g., for high-dose male Marshal rats 25 animals were
accidentally killed.
Maltoni et al. (1986) reported the results of several studies of TCE via inhalation and
gavage in mice and rats. A large number of animals were used in the treatment groups but the
focus of the study was detection of a neoplastic response with only a generalized description of
tumor pathology phenotype given and limited reporting of nonneoplastic changes in the liver.
Accidental death by gavage error was reported not to occur in this study. In regards to effects of
TCE exposure on rat survival, "a nonsignificant excess in mortality correlated to TCE treatment
was observed only in female rats (treated by ingestion with the compound.
For rats, Maltoni et al. (1986) reported 4 liver angiosarcomas (1 in a control male rat, 1
both in a TCE-exposed male and female at 600 ppm TCE for 8 weeks, and 1 in a female rat
exposed to 600 ppm TCE for 104 weeks), but the specific results for incidences of hepatocellular
"hepatomas" in treated and control rats were not given. Although the Maltoni et al. (1986)
concluded that the small number was not treatment related, the findings were brought forward
because of the extreme rarity of this tumor in control Sprague-Dawley rats, untreated or treated
with vehicle materials. In rats treated for 104 weeks, there was no report of a TCE treatment-
related increase in liver cancer in rats. This study only presented data for positive findings so it
did not give the background or treatment-related findings in rats for liver tumors in this study.
Thus, the extent of background tumors and sensitivity for this endpoint cannot be determined.
Of note is that the Sprague-Dawley strain used in this study was also noted in the Fukuda et al.
(1983) study to be relatively insensitive for spontaneous liver cancer and to also be negative for
TCE-induced hepatocellular liver cancer induction in rats. However, like Fukuda et al. (1983)
and Henschler et al. (1980), that reported rare biliary tumors in insensitive strains of rat for
hepatocellular tumors, Maltoni et al. (1986) reported a relatively rare tumor type, angiosarcoma,
after TCE exposure in a relatively insensitive strain for "hepatomas." As noted above, many of
the rat studies were limited by premature mortality due to gavage error or premature mortality
(Henschler et al., 1980; NCI, 1976; NTP, 1990, 1988), which was reported not occur in Maltoni
et al. (1986).
4.4.5.2 Positive TCE studies of mice
In the NCI (1976) study of TCE exposure in B6C3F1 mice, TCE was reported to increase
incidence of hepatocellular carcinomas in both doses and both genders of mice (~ 1,170 and
2,340 mg/kg for males and 870 and 1,740 mg/kg for female mice). Hepatocellular carcinoma
diagnosis was based on histologic appearance and metastasis to the lung. The tumors were
described in detail and to be heterogeneous "as described in the literature" and similar in
appearance to tumors generated by carbon tetrachloride. The description of liver tumors in this
6/22/2009
518
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
study and tendency to metastasize to the lung are similar to descriptions provided by Maltoni et
al. (1986) for TCE-induced liver tumors in mice via inhalation exposure.
The NTP (1990) study of TCE exposure in male and female B6C3F1 mice (1,000 mg/kg
for mice) reported decreased latency of liver tumors, with animals first showing carcinomas at 57
weeks for TCE-exposed animals and 75 weeks for control male mice. The administration of
TCE was also associated with increased incidence of hepatocellular carcinoma (tumors with
markedly abnormal cytology and architecture) in male and female mice. Hepatocellular
adenomas were described as circumscribed areas of distinctive hepatic parenchymal cells with a
perimeter of normal appearing parenchyma in which there were areas that appeared to be
undergoing compression from expansion of the tumor. Mitotic figures were sparse or absent but
the tumors lacked typical lobular organization. Hepatocellular carcinomas had markedly
abnormal cytology and architecture with abnormalities in cytology cited as including increased
cell size, decreased cell size, cytoplasmic eosinophilia, cytoplasmic basophilia, cytoplasmic
vacuolization, cytoplasmic hyaline bodies, and variations in nuclear appearance. Furthermore, in
many instances several or all of the abnormalities were present in different areas of the tumor
and variations in architecture with some of the hepatocellular carcinomas having areas of
trabecular organization. Mitosis was variable in amount and location. Therefore the phenotype
of tumors reported from TCE exposure was heterogeneous in appearance between and within
tumors. However, because it consisted of a single dose group in addition to controls, this study
is limited of limited utility for analyzing the dose-response for hepatocarcinogenicity. There was
also little reporting of non-neoplastic pathology or toxicity and no report of liver weight at
termination of the study.
Maltoni et al. (1986) reported the results of several studies of TCE in mice. A large
number of animals were used in the treatment groups but the focus of the study was detection of
a neoplastic response with only a generalized description of tumor pathology phenotype given
and limited reporting of nonneoplastic changes in the liver. There was no accidental death by
gavage error reported to occur in mice but, a "nonsignificant" excess in mortality correlated to
TCE treatment was observed in male B6C3F1 mice. TCE-induced effects on body weight were
reported to be absent in mice except for one experiment (BT 306 bis) in which a slight non-dose
correlated decrease was found in exposed animals. "Hepatoma" was the term used to describe
all malignant tumors of hepatic cells, of different sub-histotypes, and of various degrees of
malignancy and were reported to be unique or multiple, and have different sizes (usually
detected grossly at necropsy) from TCE exposure. In regard to phenotype tumors were described
as usual type observed in Swiss and B6C3F1 mice, as well as in other mouse strains, either
untreated or treated with hepatocarcinogens and to frequently have medullary (solid), trabecular,
and pleomorphic (usually anaplastic) patterns. Swiss mice from this laboratory were reported to
6/22/2009
519
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
have a low incidence of hepatomas without treatment (1%). The relatively larger number of
animals used in this bioassay (n = 90 to 100), in comparison to NTP standard assays, allows for a
greater power to detect a response.
TCE exposure for 8 weeks via inhalation at 100 ppm or 600 ppm may have been
associated with a small increase in liver tumors in male mice in comparison to concurrent
controls during the life span of the animals. In Swiss mice exposed to TCE via inhalation for 78
weeks, there a reported increase in hepatomas associated with TCE treatment that was dose-
related in male but not female Swiss mice. In B6C3F1 mice exposed via inhalation to TCE for
78 weeks, increases in hepatomas were reported in both males and females. However, the
experiment in males was repeated with B6C3F1 mice from a different source, since in the first
experiment more than half of the mice died prematurely due to excessive fighting. Although the
mice in the two experiments in males were of the same strain, the background level of liver
cancer was significantly different between mice from the different sources (1/90 versus 19/90),
though the early mortality may have led to some censoring. The finding of differences in
response in animals of the same strain but from differing sources has also been reported in other
studies for other endpoints. However, for both groups of male B6C3F1 mice the background
rate of liver tumors over the lifetime of the mice was no greater than about 20%.
There were other reports of TCE carcinogenicity in mice from chronic exposures that
were focused primarily on detection of liver tumors with limited reporting of tumor phenotype or
non-neoplastic pathology. Herren-Freund et al. (1987) reported that male B6C3 F1 mice given
40 mg/L TCE in drinking water had increased tumor response after 61 weeks of exposure.
However, concentrations of TCE fell by about V2 at this dose of TCE during the twice a week
change in drinking water solution so the actual dose of TCE the animals received was less than
40 mg/L. The % liver /body weight was reported to be similar for control and TCE -exposed
mice at the end of treatment. However, despite difficulties in establishing accurately the dose
received, an increase in adenomas per animal and an increase in the number of animals with
hepatocellular carcinomas were reported to be associated with TCE exposure after 61 weeks of
exposure and without apparent hepatomegaly. Anna et al. (1994) reported tumor incidences for
male B6C3F1 mice receiving 800 mg/kg/day TCE via gavage (5 days/week for 76 weeks). All
TCE-treated mice were reported to be alive after 76 weeks of treatment. Although the control
group contained a mixture of exposure durations (76-134 weeks) and concurrent controls had a
very small number of animals, TCE-treatment appeared to increase the number of animals with
adenomas, the mean number of adeonomas and carcinomas, but with no concurrent TCE-induced
cytotoxicity.
6/22/2009
520
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.4.5.3Summary: TCE-induced cancer in laboratory animals
Chronic TCE bioassays have consistently reported increased liver tumor incidences in
both sexes of B6C3F1 mice treated by inhalation and gavage exposure in a number of bioassays.
The only inhalation study of TCE in Swiss mice also showed an effect in males. Data in the rat,
while not reporting statistically significantly increased risks, are not entirely adequate due to low
numbers of animals, inadequate reporting, use of insensitive bioassays, increased systemic
toxicity, and/or increased mortality. Notably, several studies in rats noted a few very rare types
of liver or biliary tumors (cystic cholangioma, cholangiocarcinoma, or angiosarcomas) in treated
animals.
4.4.6 Role of metabolism in liver toxicity and cancer
It is generally thought that TCE oxidation by CYP450s is necessary for induction of
hepatotoxicity and hepatocarcinogenicity (Bull, 2000). Direct evidence for this hypothesis is
limited, e.g., the potentiation of hepatotoxicity by pretreatment with P450 inducers such as
ethanol and phenobarbital (Nakajima et al., 1988; Okino et al., 1991). Rather the presumption
that P450-mediated oxidation is necessary for TCE hepatotoxicity and hepatocarcinogenicity is
largely based on similar effects (e.g., increases in liver weight, peroxisome proliferation, and
hepatocarcinogenicity) having been observed with TCE's oxidative metabolites. The discussion
below focuses the similarities and differences between the major effects in the liver of TCE and
of the oxidative metabolites CH, TCA, and DCA. In addition, CH is largely converted to TCOH,
TCA, and possibly DCA.
4.4.6.1 Pharmacokinetics of CH, TCA, and DCA from TCE exposure
As discussed in Chapter 3, in vivo data confirm that CH and TCA, are oxidative
metabolites of TCE. In addition, there is indirect data suggesting the formation of DCA.
However, direct in vivo evidence of the formation of DCA is confounded by its rapid clearance
at low concentrations, and analytical artifacts in its detection in vivo that have yet to be entirely
resolved. PBPK modeling (Section 3.5) predicts that the proportions of TCE metabolized to CH
and TCA varies considerably in mice (ranging from 15-97% and 4-38%, respectively) and rats
(ranging 7—75% and 0.5-22%, respectively). Therefore, a range of smaller concentrations of
TCA or CH may be relevant for comparisons with TCE-induced liver effects. For example, for
1,000 mg/kg-d oral doses of TCE, the relevant comparisons would be approximately 0.25-1.5
g/L in drinking water for TCA and CH. For DCA a corresponding range is harder to determine
and has been suggested to be an upper limit of about 12% (Barton et al., 1999).
6/22/2009
521
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.4.6.2 Comparisons between TCE and TCA, DCA, and CH non-cancer effects
4.4.6.2.1 Hepatomegaly — qualitative and quantitative comparisons
As discussed above, TCE causes hepatomegaly in rats, mice, and gerbils under both acute
and chronic dosing. Data from a few available studies suggest that oxidative metabolism is
important for mediating these effects. Buben and O'Flaherty (1985) collected limited
pharmacokinetic data in a sample of the same animals for which liver weight changes were being
assessed. While liver weight increases had similarly strong correlations with applied dose and
urinary metabolites for doses up to 1,600 mg/kg-d (R-squared of 0.97 for both), above that dose,
the linear relationship was maintained with urinary metabolites but not with applied dose.
Ramdhan et al. (2008) conducted parallel experiments at TCE 1,000 and 2,000 ppm (8 hr/d, 7 d)
in wild-type and cyp2el-null mice, which did not exhibit increased liver/body weight ratios with
TCE treatment and excreted 2-fold lower amounts of oxidative metabolites TCA and TCOH in
urine as compared to wild-type mice. However, among control mice, those with the null
genotype had 1.32-fold higher absolute liver weights and 1.18-fold higher liver/body weight
ratios than wild-type mice, reducing the sensitivity of the experiment, particularly with only 6
mice per dose group.
With respect to oxidative metabolites themselves, data from CH studies are not
informative—either because data were not shown (Sanders et al., 1982) or, because at the time
points measured, liver weight increases are substantially confounded by foci and carcinogenic
lesions (Leakey et al., 2003a). TCA and DCA have both been found to cause hepatomegaly in
mice and rats, with mice being more sensitive to this effect. DCA also increases liver/body
weight ratios in dogs, but TCE and TCA have not been tested in this species (Cicmanec et al.,
1991).
As noted above, TCE-induced changes in liver weight appear to be proportional to the
exposure concentration across route of administration, gender and rodent species. As an
indication of the potential contribution of TCE metabolites to this effect, a quantitative
comparison of the shape of the dose-response curves for liver weight induction for TCE and its
metabolites is informative. The analysis below was reported in Evans et al. (2009).
A number of short-term (<4 wk) studies of TCA and DCA in drinking water have
attempted to measure changes in liver weight induction, with the majority of these studies being
performed in male B6C3F1 mice. Studies conducted from 14 to 30 days show a consistent
increase in % liver/body weight induction by TCA or DCA. However, as stated in many of the
discussions of individual studies (Appendix E), there is a limited ability to detect a statistically
significant change in liver weight change in experiments that use a relatively small number of
animals or do not match control and treatment groups for age and weight. The experiments of
Buben and O'Flaherty used 12-14 mice per group giving it a greater ability to detect a TCE-
6/22/2009
522
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
induced dose response. However, many experiments have been conducted with 4-6 mice per
dose group. For example, the data from DeAngelo et al. (2008) for TCA-induced % liver/body
weight ratio increases in male B6C3F1 mice were only derived from 5 animals per treatment
group after 4 weeks of exposure. The 0.05 g/L and 0.5 g/L exposure concentrations were
reported to give a 1.09-fold and 1.16-fold of control % liver/body weight ratios which were
consistent with the increases noted in the cross-study database above. However, a power
calculation shows that the type II error (which should be > 50% and thus greater than the chances
of "flipping a coin") was only a 6% and 7% and therefore the designed experiment could accept
a false null hypothesis. In addition, some experiments took greater care to age and weight match
the control and treatment groups before the start of treatment.
Therefore, given these limitations and the fact that many studies used a limited range of
doses, an examination of the combined data from multiple studies (Parrish et al., 1996; Sanchez
and Bull, 1990; Carter et al., 1995; Kato-Weinstein et al., 2001; DeAngelo et al., 1989, 2008) can
best inform/ discern differences in DCA and TCA dose-response relationships for liver weight
induction (described in more detail in Section 2.4.2 of Appendix E). The dose-response curves
for similar concentrations of DCA and TCA are presented in Figure 4.4.1 for durations of
exposure from 14-28 days in the male B6C3F1 mouse, which was the most common sex and
strain used. As noted in Appendix E, there appears to be a linear correlation between dose in
drinking water and liver weight induction up to 2 g/L of DCA. However, the shape of the dose-
response curve for TCA appears to be quite different. Lower concentrations of TCA induce
larger increase that does DCA, but the TCE response reaches an apparent plateau while that of
DCA continues to increase the response. TCA studies did not show significant duration-
dependent difference in liver weight induction in this duration range. Short duration studies
(10-42 days) were selected because (i) in chronic studies, liver weight increases are confounded
by tumor burden, (ii) multiple studies are available, and (iii) TCA studies do not show significant
duration-dependent differences in this duration range.
6/22/2009
523
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Male B6C3F1 mice liver weight for TCA and DCA in drinking water - days 14-30
2.0
DCA
TCA
1.8
1.6
1.4
1.2
1.0
0.0
0.5
2.0
2.5
Concentration of DCA or TCA (g/l)
Figure 4.4.1. Comparison of average fold-changes in relative liver weight to control and
exposure concentrations of 2 g/L or less in drinking water for TCA and DCA in male B6C3F1
mice for 14-30 days (Parrish et al.,1996; Sanchez and Bull, 1990; Carter et al., 1995; Kato-
Weinstein et al., 2001; DeAngelo et al., 1989, 2008).
Of interest is the issue of how the dose-response curves for TCA and DCA compare to
that of TCE in a similar model and dose range. Since TCA and DCA have strikingly different
dose-response curves, which one if either best fits that of TCE and thus can give insight as to
which is causative agent for TCE's effects in the liver? The carcinogenicity of chronic TCE
exposure has been predominantly studies in two mouse strains, Swiss and B6C3F1, both of
which reportedly developed liver tumors. Rather than administered in drinking water, oral TCE
studies have been conducted via oral gavage and generally in corn oil for 5 days of exposure per
week. Factors adding to the increased difficulty in establishing the dose-response relationship
for TCE across studies and for comparisons to the DCA and TCA database include vehicle
effects, the difference between daily and weekly exposures, the dependence of TCE effects in the
liver on its metabolism to a variety of agents capable inducing effects in the liver, differences in
response between strains, and the inherent increased variability in use of the male mouse model.
Despite difference in exposure route, etc, a consistent pattern of dose-response emerges from
combining the available TCE data. The effects of oral exposure to TCE from 10-42 days on
6/22/2009
524
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
liver weight induction is shown below in Figure 4.4.2 using the data of Elcombe et al. (1985),
Dees and Travis (1993), Goel et al. (1992), Merrick et al. (1987), Goldsworthy and Popp (1987),
and Buben and O'Flaherty (1985). Oral TCE administration in male B6C3F1 and Swiss mice
appeared to induce a dose-related increase in % liver/body weight that was generally
proportional to the increase in magnitude of dose, though as expected, with more variability than
observed for a similar exercise for DCA or TCA in drinking water. Some of the variability is
due to the inclusion of the 10 day studies, since as discussed in Section 2.4.2. of Appendix E,
there was a greater increase in TCE-induced liver weight at 28-42 days of exposure Swiss mice
than the 10-day data in B6C3F1 mice, and Kjellstrand et al. (1981) noted that TCE-induced liver
weight increases are still increasing at 10 days inhalation exposure. A strain difference is not
evident between the Swiss and B6C3F1 males, as both the combined TCE data and that for only
B6C3F1 mice show similar correlation with the magnitude of dose and magnitude of %
liver/body weight increase. The correlation coefficients for the linear regressions presented for
the B6C3F1 data is R2 = 0.861 and for the combined data sets is R2 = 0.712. Comparisons of the
slopes of the dose-response curves suggest a greater consistency between TCE and DCA than
between TCE and TCA. There did not appear to be evidence of a plateau with higher TCE
doses, and the degree of fold-increase rises to higher levels with TCE than with TCA in the same
strain of mouse.
Male mice liver weight for TCE oral gavage - days 10-42
2.0
• B6C3F1
Regression
O)
0
500
1000
1500
2000
2500
3000
Concentration of TCE (mg/kg/day)
6/22/2009
525
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Male mice liver weight forTCE oral gavage - days 10-42
2.0
• B6C3F1 and Swiss
Plot 2 Regr
O)
0
500
1000
1500
2000
2500
3000
Concentration of TCE (mg/kg/day)
Figure 4.4.2 Comparisons of fold-changes in average relative liver weight and gavage dose of
(top panel) male B6C3F1 mice for 10-28 days of exposure (Merrick et al., 1989; Elcombe et al.,
1985; Goldsworthy and Popp, 1987; Dees and Travis, 1993) and (bottom panel) in male B6C3F1
and Swiss mice
A more direct comparison would be on the basis of dose rather than drinking water
concentration. The estimations of internal dose of DCA or TCA from drinking water studies,
while varying considerably (DeAngelo et al., 1989, 2008), nonetheless suggest that the doses of
TCE used in the gavage experiments were much higher than those of DCA or TCA. However,
only a fraction of ingested TCE is metabolized to DCA or TCA, as, in addition to oxidative
metabolism, TCE is also cleared by GSH conjugation and by exhalation. While DCA dosimetry
is highly uncertain (Sections 3.3 and 3.5), the mouse PBPK model, described in Section 3.5 was
calibrated using extensive in vivo data on TCA blood, plasma, liver, and urinary excretion data
from inhalation and gavage TCE exposures, and makes robust predictions of the rate of TCA
production. If TCA were predominantly responsible for TCE-induced liver weight increases,
then replacing administered TCE dose (e.g., mg TCE/kg/day) by the rate of TCA produced from
TCE (mg TCA/kg/day) should lead to dose-response curves for increased liver weight consistent
with those from directly administered TCA. Figure 4.4.3 shows this comparison using the PBPK
model-based estimates of TCA production for 4 TCE studies from 28-42 days in the male
NMRI, Swiss, and B6C3F1 mice (Kjellstrand et al., 1983b; Buben and O'Flaherty, 1985;
Merrick et al., 1989; Goel et al., 1992) and 4 oral TCA studies in B6C3F1 male mice at 2 g/L or
lower drinking water exposure (DeAngelo et al., 1989, 2008; Parrish et al., 1996; Kato-
6/22/2009
526
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Weinstein et al., 2001) from 14-28 days of exposure. The selection of the 28-42 day data for
TCE was intended to address the decreased opportunity for full expression of response at 10
days. PBPK modeling predictions of daily internal doses of TCA in terms of mg/kg/day via
produced via TCE metabolism would be are indeed lower than the TCE concentrations in terms
of mg/kg/day given orally by gavage. The predicted internal dose of TCA from TCE exposure
studies are of a comparable range to those predicted from TCA drinking water studies at
exposure concentrations in which palpability has not been an issue for estimation of internal
dose. Thus although the TCE data are for higher exposure concentrations, they are predicted to
produce comparable levels of TCA internal dose estimated from direct TCA administration in
drinking water.
Figure 4.4.3 clearly shows that for a given amount of TCA produced from TCE, but
going through intermediate metabolic pathways, the liver weight increases are substantially
greater than, and highly inconsistent with, that expected based on direct TCA administration. In
particular, the response from direct TCA administration appears to "saturate" with increasing
TCA dose at a level of about 1.4-fold, while the response from TCE administration continues to
increase with dose to 1.75-fold at the highest dose administered orally in Buben and O'Flaherty
1985) and over 2-fold in the inhalation study of Kjellstrand et al. (1983b).
6/22/2009
527
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
2.5
~ TCE Studies [28-42 d]
O TCA Studies [14-28 d]
— - Linear (TCA Studies [14-28 d])
Linear (TCE Studies [28-42 d])
O)
¦o
O)
~ ~
0 100 200 300 400 500
mg TCA/kg-d
(produced [TCE studies] or administered [TCA studies])
Figure 4.4.3. Comparison of fold-changes in relative liver weight for datasets in male B6C3F1,
Swiss, and NRMI mice between TCE studies (Kjellstrand et al., 1983b; Buben and O'Flaherty,
1985; Merrick et al., 1989; Goel et al., 1992) [duration 28-42 days] and studies of direct oral
TCA administration to B6C3 F1 mice (DeAngelo et al., 1989; Parrish et al., 1996; Kato-
Weinstein et al., 2001; DeAngelo et al., 2008) [duration 14-28 days]. Abscissa for TCE studies
consists of the median estimates of the internal dose of TCA predicted from metabolism of TCE
using the PBPK model described in Section 3.5 of the TCE risk assessment. Lines show linear
regression with intercept fixed at unity. All data were reported fold-change in mean liver
weight/body weight ratios, except for Kjellstrand et al. (1983b), with were the fold-change in the
ratio of mean liver weight to mean body weight. In addition, in Kjellstrand et al. (1983b), some
systemic toxicity as evidence by decreased total body weight was reported in the highest dose
group.
6/22/2009
528
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Furthermore, while as noted previously, oral studies appear to report a linear relationship
between TCE exposure concentration and liver weight induction, the inclusion of inhalation
studies on the basis of internal dose led to a highly consistent dose-response curve for among
TCE study. Therefore, it is unlikely that differing routes of exposure can explain the
inconsistencies in dose-response.
Additional analyses do, however, support a role for oxidative metabolism in TCE-
induced liver weight increases, and that the parent compound TCE is not the likely active moiety
(suggested previously by Buben and O'Flaherty [1985]). In particular, the same studies are
shown in Figure 4.4.4 using PBPK-model based predictions of the area-under-the-curve (AUC)
of TCE in blood and total oxidative metabolism, which produces chloral, TCOH, DCA, and
other metabolites in addition to TCA. The dose-response relationship between TCE blood levels
and liver weight increase, while still having a significant trend, shows substantial scatter and a
low R2 of 0.43. On the other hand, using total oxidative metabolism as the dose metric leads to
substantially more consistency dose-response across studies, and a much tighter linear trend with
an R2 of 0.90 (Figure 4.4.4). A similar consistency is observed using liver-only oxidative
metabolism as the dose metric, with R2 of 0.86 (not shown). Thus while the slope is similar
between liver weight increase and TCE concentration in the blood and liver weight increase and
rate of total oxidative metabolism, the data are a much better fit for total oxidative metabolism.
6/22/2009
529
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1
03
CO
CO
03
i_
o
c
R =0.426
i
o
U_
LO
0
100 200 300 400 500
R =0.8955
03
CO
CO
03
o
c
I
o
U_
LO
0
500
1000
1500
2 Daily AUC TCE in Blood (mg-hr/l) Daily TCE Oxidized (mg/kg-d)
3 Figure 4.4.4. Fold-changes in relative liver weight for data sets in male B6C3F1, Swiss, and
4 NRMI mice reported by TCE studies of duration 28-42 days (Kjellstrand et al., 1983b; Buben
5 and O'Flaherty, 1985; Merrick et al., 1989; Goel et al., 1992) using internal dose metrics
6 predicted by the PBPK model described in section 3.5: (A) dose metric is the median estimate of
7 the daily AUC of TCE in blood, (B) dose metric is the median estimate of the total daily rate of
8 TCE oxidation. Lines show linear regression. Use of liver oxidative metabolism as a dose
9 metric gives results qualitatively similar to (B), with R2 = 0.86.
6/22/2009
530
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Although the qualitative similarity to the linear dose-response relationship between DCA
and liver weight increases is suggestive of DCA being the predominant metabolite responsible
for TCE liver weight increases, due to the highly uncertain dosimetry of DCA derived from
TCE, this hypothesis cannot be tested on the basis of internal dose. Similarly, another TCE
metabolite, chloral hydrate, has also been reported to induce liver tumors in mice, however, there
are no adequate comparative data to assess the nature of liver weight increases induced by this
TCE metabolite (see Section 2.5. of Appendix E and Section 4.4.1.2.4 below). Whether its
formation in the liver after TCE exposure correlates with TCE-induced liver weight changes
cannot be determined.
4.4.6.2.2 Cytotoxicity
As discussed above, TCE has sometimes been reported to cause minimal/mild focal
hepatocellular necrosis or other signs of hepatic injury, albeit of low frequency and mostly at
doses > 1,000 mg/kg-d (Dees and Travis, 1993; Elcombe et al., 1985) or at exposures > 1,000
ppm in air (Ramdhan et al., 2008) from 7-10 days of exposure. Data from available studies are
supportive of a role for oxidative metabolism in TCE-induced cytotoxicity in the liver, though
they are not informative as to the actual active moiety(ies). Buben and O'Flaherty (1985) noted
a strong correlation (R-squared of between G6P inhibition and total urinary oxidative
metabolites. Ramdhan et al. (2008) conducted parallel experiments at TCE 1,000 and 2,000 ppm
(8 hr/d, 7 d) in wild-type and cyp2el-null mice, the latter of which did not exhibit hepatotoxicity
(assessed by serum ALT, AST, and histopathology) and excreted 2-fold lower amounts of
oxidative metabolites TCA and TCOH in urine as compared to wild-type mice. In addition,
urinary TCA and TCOH excretion was correlated with serum ALT and AST measures, though
the R-squared values (square of the reported correlation coefficients) were relatively low (0.54
and 0.67 for TCOH and TCA, respectively).
With respect to CH (166 mg/kg/d) and DCA (-90 mg/kg/d), Daniel et al. (1992) reported
that after drinking water treatment, hepatocellular necrosis and chronic active inflammation were
reported to be mildly increased in both prevalence and severity in all treated groups after 104
weeks of exposure. The histological findings, from interim sacrifices (n = 5), were considered
by the authors to be unremarkable and were not reported. TCA has not been reported to induce
necrosis in the liver under the conditions tested. Relatively high doses of DCA (> lg/L in
drinking water) appear to result in mild focal necrosis with attendant reparative proliferation at
lesion sites, but no such effects were reported at lower doses (< 0.5 g/L in drinking water) more
relevant for comparison with TCE (DeAngelo et al., 1999; Sanchez and Bull, 1990; Stauber et
al., 1998). Enlarged nuclei and changes consistent with increased ploidy, are further discussed
below in the context of DNA synthesis.
6/22/2009
531
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.4.6.2.3 DNA synthesis andpolyploidization
The effects on DNA synthesis and polyploidization observed with TCE treatment have
similarly been observed with TCA and DCA. With respect to CH, George et al. (2000) reported
that CH exposure did not alter DNA synthesis in rats and mice at any of the time periods
monitored (all well past 2 weeks), with the exception of 0.58 g/L chloral hydrate at 26 weeks
slightly increasing hepatocyte labeling (~ 2-3 fold of controls) in rats and mice but the %
labeling still representing 3% or less of hepatocytes.
In terms of whole liver or hepatocyte label incorporation, the most comparable exposure
duration between TCE, TCA, and DCA studies is the 10- and 14-day period. Several studies
have reported that in this time period, peak label incorporation into individual hepatocytes and
whole liver for TCA and DCA have already passed (Styles et al., 1991; Sanchez and Bull, 1990;
Pereira, 1996; Carter et al., 1995). A direct time-course comparison is difficult, since data at
earlier times for TCE are more limited.
There are conflicting reports of DNA synthesis induction in individual hepatocytes for up
to 14 days of DCA or TCA exposure. In particular, Sanchez and Bull (1990) reported tritiated
thymidine incorporation in individual hepatocytes up to 2 g/L exposure to DCA or TCA induced
little increase in DNA synthesis except in instances and in close proximity to areas of
proliferation/necrosis for DCA treatment after 14 days of exposure in male mice. The largest
percentage of hepatocytes undergoing DNA synthesis for any treatment group was less than 1%
of hepatocytes. However, they reported treatment- and exposure duration-changes in hepatic
DNA incorporation of tritiated thymidine for DCA and TCA. For TCA treatment, the largest
increases over control levels for hepatic DNA incorporation (at the highest dose) was a 3-fold
increase after 5 days of treatment and a 2-fold increase over controls after 14 days of treatment.
For DCA whole-liver tritiated thymidine incorporation was only slightly elevated at necrogenic
concentrations and decreased at the 0.3 g/L non-necrogenic level after 14 days of treatment. In
contrast to Sanchez and Bull (1990), Stauber and Bull (1997) reported increased tritiated
thymidine incorporation for individual hepatocytes after 14 days of treatment with 2 g/L DCA or
TCA in male mice. They used a more extended period of tritiated thymidine exposure of 3-5
days and so these results represent aggregate DNA synthesis occurring over a more extended
period of time. A "1-day labeling index" was reported as less than 1% for the highest level of
increased incorporation. However, after 14 days, the labeling index was reported to be increased
by ~3.5-fold for TCA and -5.5 fold for DCA over control values. After 28 days, the labeling
index was reported to be decreased ~ 2.3 fold by DCA and increased -2.5 fold after treatment
with TCA. Pereira (1996) reported that for female B6C3F1 mice, 5-day incorporation of BrDU,
as a measure of DNA synthesis, was increased at 0.86 g/L and 2.58 g/L DCA treatment for 5
days (~ 2-fold at the highest dose) but that by day 12 and 33 levels had fallen to those of
6/22/2009
532
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
controls. For TCA exposures, 0.33 g/L, 1.10 g/L and 3.27 g/L TCA all gave a similar ~3-fold
increase in BrdU incorporation by 5 days, but that by 12 and 33 days were not changed from
controls. Nonetheless, what is consistent is that these data report that, similar to TCE-exposed
mice at 10 days of exposure, cells undergoing DNA synthesis in DCA- or TCA-exposed mice for
up to 14 days of exposure to be confined to a very small population of cells in the liver. Thus,
these data are consistent with hypertrophy being primarily responsible for liver weight gains as
opposed to increases in cell number in mice.
Interestingly, a lack of correlation between whole liver label incorporation and that in
individual hepatocytes has been reported by several studies of DCA (Sanchez and Bull, 1990;
Carter et al., 1995). For example, Carter et al. (1995) reported no increase in labeling of
hepatocytes in comparison to controls for any DCA treatment group from 5 to 30 days of DCA
exposure. Rather than increase hepatocyte labeling, DCA induced no change from days 5 though
15 but significantly decreased levels between days 20 and 30 for 0.5 g/L that were similar to
those observed for the 5 g/L exposures. However for whole liver DNA tritiated thymidine
incorporation, Carter et al. (1995) reported 0.5g/L DCA treatments to show trends of initial
inhibition of DNA tritiated thymidine incorporation followed by enhancement of labeling that
was not statistically significant from 5 to 30 days of exposure. Examination of individual
hepatocytes does not include the contribution of nonparenchymal cell DNA synthesis that would
be detected in whole liver DNA. As noted above, proliferation of the nonparenchymal cell
compartment of the liver has been noted in several studies of TCE in rodents, and thus this is one
possible reason for the reported discrepancy.
Another possible reason for this inconsistency with DCA treatment is polyploidization, as
was suggested above for TCE. Although this was not examined for DCA or TCA exposure by
Sanchez and Bull (1990), Carter et al. (1995) reported that hepatocytes from both 0.5 and 5 g/L
DCA treatment groups had enlarged, presumably polyploidy nuclei, with some hepatocyte nuclei
labeled in the mid-zonal area. There were statistically significant changes in cellularity, nuclear
size, and multinucleated cells during 30 days exposure to DCA. The percentage of
mononucleated cells hepatocytes was reported to be similar between control and DCA treatment
groups at 5- and 10-day exposure. However, at 15 days and beyond DCA treatments were
reported to induce increases in mononucleated hepatocytes with later time periods to also
showing DCA-induced increases nuclear area, consistent with increased polyploidization without
mitosis. The consistent reporting of an increasing number of mononucleated cells between 15
and 30 days could be associated with clearance of mature hepatocytes as suggested by the report
of DCA-induced loss of cell nuclei. The reported decrease in the numbers of binucleate cells in
favor of mononucleate cells is not typical of any stage of normal liver growth (Brodsky and
Uryvaeva, 1977). The pattern of consistent increase in % liver/body weight induced by 0.5 g/L
6/22/2009
533
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
DCA treatment from days 5 though 30 was not consistent with the increased numbers of
mononucleate cells and increase nuclear area reported from day 20 onward. Specifically, the
large differences in liver weight induction between the 0.5 g/L treatment group and the 5 g/L
treatment groups at all times studied also did not correlate with changes in nuclear size and % of
mononucleate cells. Thus, increased liver weight was not a function of cellular proliferation, but
probably included both aspects of hypertrophy associated with polyploidization and increased
glycogen deposition (see below) induced by DCA. Carter et al. (1995) suggested that although
there is evidence of DCA-induced cytotoxicity (e.g., loss of cell membranes and apparent
apoptosis), the 0.5 g/L exposure concentration has been shown to increase hepatocellular lesions
after 100 weeks of treatment without concurrent peroxisome proliferation or cytotoxicity
(DeAngelo et al., 1999).
In sum, the observation of TCE-treatment related changes in DNA content, label
incorporation, and mitotic figures are generally consistent with patterns observed for both TCA
and DCA. In all cases, hepatocellular proliferation is confined to a very small fraction of
hepatocytes, and hepatomegaly observed with all three treatments probably largely reflects
cytomegaly rather than cell proliferation. Moreover, label incorporation likely largely reflects
polyploidization rather than hepatocellular proliferation, with a possible contribution from non-
parenchymal cell proliferation. As with TCE, histological changes in nuclear sizes and number
also suggest a significant degree of treatment-related polyploidization, particularly for DCA.
4.4.6.2.4 Apoptosis
As for apoptosis, Both Elcombe et al. (1985) and Dees and Travis (1993) reported no
changes in apoptosis other than increased apoptosis only at a treatment level of 1,000 mg/kg
TCE. Dees and Travis (1993) reported that increased apoptoses from TCE exposure "did not
appear to be in proportion to the applied TCE dose given to male or female mice." Channel et al.
(1998) reported that there was no significant difference in apoptosis between TCE treatment and
control groups with data not shown. However, the extent of apoptosis in any of the treatment
groups, or which groups and timepoints were studied for this effect cannot be determined. While
these data are quite limited, it is notable that peroxisome proliferators have been suggested
inhibit, rather than increase, apoptosis as part of their carcinogenic MOA (Klaunig et al., 2003).
However, for TCE metabolites, DCA has been most studied, though it is clear that age
and species affect background rates of apoptosis. Snyder et al. (1995), in their study of DCA,
report that control mice were reported to exhibit apoptotic frequencies ranging from ~ 0.04 to
0.085%, that over the 30-day period of their study the frequency rate of apoptosis declined, and
suggest that this pattern is consistent with reports of the livers of young animals undergoing
rapid changes in cell death and proliferation. They reported rat liver to have a greater the
estimated frequency of spontaneous apoptosis (~ 0.1%) and therefore greater than that of the
6/22/2009
534
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
mouse. Carter et al. (1995) reported that after 25 days of 0.5 g/L DCA treatment apoptotic
bodies were reported as well as fewer nuclei in the pericentral zone and larger nuclei in central
and midzonal areas. This would indicate an increase in the apoptosis associated with potential
increases in polyploidization and cell maturation. However, Snyder et al. (1995) report that mice
treated with 0.5 g/L DCA over a 30-day period had a similar trend as control mice of decreasing
apoptosis with age. The percentage of apoptotic hepatocytes decreased in DCA-treated mice at
the earliest time point studied and remained statistically significantly decreased from controls
from 5 to 30 days of exposure. Although the rate of apoptosis was very low in controls,
treatment with 0.5g/L DCA reduced it further (-30-40% reduction) during the 30-day study
period. The results of this study not only provide a baseline of apoptosis in the mouse liver,
which is very low, but also to show the importance of taking into account the effects of age on
such determinations. The significance of the DCA-induced reduction in apoptosis reported in
this study, from a level that is already inherently low in the mouse, for the MOA for induction of
DCA-induce liver cancer is difficult to discern.
4.4.6.2.5 Glycogen accumulation
As discussed in Sections 3.2 and 3.4.2.1 of Appendix E, glycogen accumulation has been
described to be present in foci in both humans and animals as a result from exposure to a wide
variety of carcinogenic agents and predisposing conditions in animals and humans. The data
from Elcombe et al. (1985) included reports of TCE-induced pericentral hypertrophy and
eosinophilia for both rats and mice but with "fewer animals affected at lower doses." In terms of
glycogen deposition, Elcombe report "somewhat" less glycogen pericentrally in the livers of rats
treated with TCE at 1,500 mg/kg than controls with less marked changes at lower doses
restricted to fewer animals. They do not comment on changes in glycogen in mice. Dees and
Travis (1993) reported TCE-induced changes to "include an increase in eosinophilic cytoplasmic
staining of hepatocytes located near central veins, accompanied by loss of cytoplasmic
vacuolization." Since glycogen is removed using conventional tissue processing and staining
techniques, an increase in glycogen deposition would be expected to increase vacuolization and
thus the report from Dees and Travis is consistent with less not more glycogen deposition.
Neither study produced a quantitative analysis of glycogen deposition changes from TCE
exposure. Although not explicitly discussing liver glycogen content or examining it
quantitatively in mice, these studies suggest that TCE-induced liver weight increases did not
appear to be due to glycogen deposition after 10 days of exposure and any decreases in glycogen
were not necessarily correlated with the magnitude of liver weight gain either.
For TCE and TCA 500 mg/kg treatments in mice for 10 days, changes in glycogen were
not reported in the general descriptions of histopathological changes (Elcombe et al., 1985;
Styles et al., 1991; Dees and Travis, 1993) or were specifically described by the authors as being
6/22/2009
535
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
similar to controls (Nelson et al., 1989). However for DC A, glycogen deposition was
specifically noted to be increased with treatment, although no quantitative analyses was
presented that could give information as to the nature of the dose-response (Nelson et al., 1989).
In regard to cell size, although increased glycogen deposition with DCA exposure was
noted by Sanchez and Bull (1990) to occur to a similar extent in B6C3F1 and Swiss Webster
male mice despite differences in DCA-induced liver weight gain. Lack of quantitative analyses
of that accumulation in this study precludes comparison with DCA-induced liver weight gain.
Carter et al. (1995) reported that in control mice there was a large variation in apparent glycogen
content and also did not perform a quantitative analysis of glycogen deposition. The variability
of this parameter in untreated animals and the extraction of glycogen during normal tissue
processing for light microscopy make quantitative analyses for dose-response difficult unless
specific methodologies are employed to quantitatively assess liver glycogen levels as was done
by Kato-Weinstein et al. (2001) and Pereira et al. (2004).
Bull et al. (1990) reported that glycogen deposition was uniformly increased from 2 g/L
DCA exposure with photographs of TCA exposure showing slightly less glycogen staining than
controls. However, the abstract and statements in the paper suggest that there was increased
PAS positive material from TCA treatment that has caused confusion in the literature in this
regard. Kato-Weinstein et al. (2001) reported that in male B6C3F1 mice exposed to DCA and
TCA, the DCA treatment increased glycogen and TCA decreased glycogen content of the liver
by using both chemical measurement of glycogen in liver homogenates and by using ethanol-
fixed sections stained with PAS, a procedure designed to minimize glycogen loss.
Kato-Weinstein et al. (2001) reported that glycogen rich and poor cells were scattered
without zonal distribution in male B6C3F1 mice exposed to 2 g/L DCA for 8 weeks. For TCA
treatments they reported centrilobular decreases in glycogen and ~ 25% decreases in whole liver
by 3 g/L TCA. Kato-Weinstein et al. (2001) reported whole liver glycogen to be increased
~1.50-fold of control (90 vs. 60 mg glycogen/g liver) by 2 g/L DCA after 8 weeks exposure male
B6C3F1 mice with a maximal level of glycogen accumulation occurring after 4 weeks of DCA
exposure. Pereira et al. (2004) reported that after 8 weeks of exposure to 3.2 g/L DCA liver
glycogen content was 2.20-fold of control levels (155.7 vs. 52.4. mg glycogen/g liver) in female
B6C3F1 mice. Thus, the baseline level of glycogen content reported by (~ 60 mg/g) and the
increase in glycogen after DCA exposure was consistent between Kato-Weinstein et al. (2001)
and Pereira et al. (2004). However, the increase in liver weight reported by Kato-Weinstein et al.
(2001) of 1.60-fold of control % liver/body weight cannot be accounted for by the 1.50-fold of
control glycogen content. Glycogen content only accounts for 5% of liver mass so that 50%
increase in glycogen cannot account for the 60% increase liver mass induced by 2 g/L DCA
exposure for 8 weeks reported by Kato-Weinstein (2001). Thus, DCA-induced increases in liver
6/22/2009
536
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
weight are occurring from other processes as well. Carter et al. (2003) and DeAngelo et al.
(1999) reported increased glycogen after DC A treatment at much lower doses after longer
periods of exposure (100 weeks). Carter reported increased glycogen at 0.5 g/L DCA and
DeAngelo et al. (1999) at 0.03 g/L DCA in mice. However there is no quantitation of that
increase.
4.4.6.2.6 Peroxisome proliferation and related effects
TCA and DCA have both been reported to induce peroxisome proliferation or increase in
related enzyme markers in rodent hepatocytes (DeAngelo et al., 1989, 1997; Mather et al., 1990;
Parrish et al., 1996). Between TCA and DCA, both induce peroxisome proliferation in various
strains of mice, but it clear that TCA and DCA are weak PPARa agonists and that DCA is
weaker than TCA in this regard (Nelson et al., 1989) using a similar paradigm.
George et al. (2000) reported that CH exposure did not hepatic PCO activity in rats and
mice at any of the time periods monitored. It is notable that the only time at which DNA
synthesis index was (slightly) increased, at 26 weeks, there remained a lack of induction of PCO.
A number of measures that may be related to peroxisome proliferation were investigated in
Leakey et al. (2003a). Of the enzymes associated with PPARa agonism (total CYP, CYP2B
isoform, CYP4A, or lauric acid P-hydroxylase activity), only CYP4A and lauric acid P-
hydroxylase activity were significantly increased at 15 months of exposure in the dietary-
restricted group administered the highest dose (100 mg/kg CH) with no other groups reported
showing a statistically significant increased response (n = 12/group). There is an issue of
interpretation of peroxisomal enzyme activities and other enzymes associated with PPARa
receptor activation to be a relevant event in liver cancer induction at a time period in which
tumors or foci are already present. Although not statistically significant, the 100 mg/kg CH
exposure group of ad libitum-fed mice also had an increase in CH-induced increases of CYP4A
and lauric acid P-hydroxylase activity. Seng et al. (2003) described CH toxicokinetics and
peroxisome proliferation-associated enzymes in mice at doses up to 1,000 mg/kg/day for 2 weeks
with dietary control or caloric restriction. Lauric acid P-hydroxylase and PCO activities were
reported to be induced only at doses >100 mg/kg in all groups, with dietary-restricted mice
showing the greatest induction. Differences in serum levels of TCA, the major metabolite
remaining 24 hr after dosing, were reported not to correlate with hepatic lauric acid P-
hydroxylase activities across groups.
Direct quantitative inferences regarding the magnitude of response in these studies in
comparison to TCE, however, are limited by possible variability and confounding. In particular,
many studies used cyanide-insensitive palmitoyl CoA oxidase activity (PCO) as a surrogate for
peroxisome proliferation, but the utility of this marker may be limited for a number of reasons.
6/22/2009
537
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
First, several studies have shown that this activity is not well correlated with the volume or
number of peroxisomes that are increased as a result of exposure to TCE or it metabolites
(Nakajima et al., 2000; Elcombe et al., 1985; Nelson et al., 1989). In addition, this activity
appears to be highly variable both as a baseline measure and in response to chemical exposures.
Laughter et al. (2004) presented data showing WY-14,643 induced increases in PCO activity that
varied up to 6-fold between different experiments in wild-type mice. They also showed that, in
some instances, PCO activity in untreated PPARa-null mice was up to 6-fold greater than that in
wild type mice. Parrish et al. (1996) noted that control values between experiments varied as
much as a factor of 2-fold for PCO activity and thus their data were presented as percent of
concurrent controls. Furthermore, Melnick et al. (1987) reported that corn oil administration
alone can elevate PCO (as well as catalase) activity, and corn oil has also been reported to
potentiate the induction of PCO activity of TCA in male mice (DeAngelo et al., 1989). Thus,
quantitative inferences regarding the magnitude of response in these studies are limited by a
number of factors. For example, in the studies reported in DeAngelo et al. (2008) a small
number of animals was studied for PCO activity at interim sacrifices (n = 5). PCO activity
varied 2.7-fold as baseline controls. Although there was a 10-fold difference in TCA exposure
concentration, the increase in PCO activity at 4 weeks was 1.3-fold, 2.4-fold, and 5.3-fold of
control. More information on the relationship of PCO enzyme activity and its relationship to
carcinogenicity is discussed in Section 3.4 of Appendix E and below.
4.4.6.2.7 Oxidative stress
Very limited data is available as to oxidative stress and related markers induced by the
oxidative metabolites of TCE. As discussed in Appendix E, above, there is limited data that do
not indicate significant oxidative stress and associated DNA damage associated with acute and
sub-acute TCE treatment. In regard to DC A and TCA, Larson and Bull (1992) exposed male
B6C3F1 mice or Fischer 344 rats to single doses TCA or DCA in distilled water by oral gavage
(n = 4). In the first experiment, TBARS was measured from liver homogenates and assumed to
be malondialdehyde. The authors stated that a preliminary experiment had shown that maximal
TBARS was increased 6 hours after a dose of DCA and 9 hours after a dose of TCA in mice
(data shown) and that by 24 hours TBARS concentrations had declined to control values (data
not shown). Time-course information in rats was not presented. A dose of 100 mg/kg DCA (rats
or mice) or TCA (mice) did not elevate TBARS concentrations over that of control liver with this
concentration of TCA not examined in rats. For TCA, there was a slight dose-related increase in
TBARS over control values starting at 300 mg/kg in mice with the increase in TBARS
increasing at a rate that was lower than the magnitude of increase in dose. Of note, is the report
that the induction of TBARS in mice is transient and has subsided within 24 hours of a single
dose of DCA or TCA, that the response in mice appeared to be slightly greater with DCA than
6/22/2009
538
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCA at similar doses, and that for DCA, there was similar TBARS induction between rats and
mice at similar dose levels.
Austin et al. (1996) appears to a follow-up publication of the preliminary experiment
cited in Larson and Bull (1992). Male B6C3F1 mice were treated with single doses of DCA or
TCA via gavage with liver examined for 8-OHdG. The authors stated that in order to conserve
animals, controls were not employed at each time point. There was a statistically significant
increase over controls in 8-OHdG for the 4- and 6-hour time points for DCA (~ 1.4-fold and 1.5-
fold of control, respectively) but not at 8 hours in mice. For TCA, there was a statistically
significant increase in 8-OHdG at 8 and 10 hours for TCA (~ 1.4 and 1.3-fold of control,
respectively).
Consistent results as to low, transient increases in markers of "oxidative stress" were also
reported by Parrish et al. (1996), who in addition to examining oxidative stress alone, attempted
to examine its possible relationship to PCO and liver weight in male B6C3F1 mice exposed to
TCA or DCA for 3 or 10 weeks (n = 6). The dose-related increase in PCO activity at 21 days for
TCA was reported to not be increased similarly for DCA. Only the 2.0 g/L dose of DCA was
reported to induce a statistically significant increase at 21-days of exposure of PCO activity over
control (~ 1.8-fold of control). After 71 days of treatment, TCA induced dose-related increases
in PCO activities that were ~ twice the magnitude as that reported at 21 days. Treatments with
DCA at the 0.1 and 0.5 g/L exposure levels produced statistically significant increase in PCO
activity of - 1.5-fold and 2.5-fold of control, respectively. The administration of 1.25 g/L
clofibric acid in drinking water, used as a positive control, gave ~ 6-7-fold of control PCO
activity at 21 and 71 days exposure. Parrish et al. (1996) reported that laurate hydroxylase
activity was reported to be elevated significantly only by TCA at 21 days and to approximately
the same extent (~ 1.4 to 1.6-fold of control) increased at all doses tested and at 71 days both the
0.5 and 2.0 g/L TCA exposures to a statistically significant increase in laurate hydroxylase
activity (i.e., 1.6-fold and 2.5-fold of control, respectively). No change was reported after DCA
exposure. Laurate hydroxylase activity within the control values varying 1.7-fold between 21
and 71 days experiments. Levels of 8-OHdG in isolated liver nuclei were reported to not be
altered from 0.1, 0.5, or 2.0 g/L TCA or DCA after 21 days of exposure and this negative result
was reported to remain even when treatments were extended to 71 days of treatment. The
authors noted that the level of 8-OHdG increased in control mice with age (i.e., ~ 2 fold increase
between 71-day and 21-day control mice). Thus, the increases in PCO activity noted for DCA
and TCA were not associated with 8-OHdG levels (which were unchanged) and also not with
changes laurate hydrolase activity observed after either DCA or TCA exposure. Of note, is that
the authors report taking steps to minimize artifactual responses for their 8-OHdG
6/22/2009
539
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
determinations. The authors concluded that their data suggests that peroxisome proliferative
properties of TCA were not linked to oxidative stress or carcinogenic response.
4.4.6.3 Comparisons of TCE-induced carcinogenic responses with TCA, DCA, and CH
studies
4.4.6.3.1 Studies in Rats
As discussed above, data on TCE carcinogenicity in rats, while not reporting statistically
significantly increased risks, are not entirely adequate due to low numbers of animals, increased
systemic toxicity, and/or increased treatment-related or accidental mortality. Notably, several
studies in rats noted a few very rare types of liver or biliary tumors (cystic cholangioma,
cholangiocarcinoma, or angiosarcomas) in treated animals. For TCA, DCA and CH, there are
even fewer studies in rats, so there is a very limited ability to assess the consistency or lack
thereof in rat carcinogenicity among these compounds.
For TCA, the only available study in rats (DeAngelo et al., 1997) has been frequently
cited in the literature to indicate a lack of response in this species for TCA-induced liver tumors.
However, this study does report an apparent dose-related increase in multiplicity of adenomas
and an increase in carcinomas over control at the highest dose. The use by DeAngelo et al.
(1997) of a relatively low number of animals per treatment group (n = 20-24) limits this study's
ability to determine a statistically significant increase in tumor response. Its ability to determine
an absence of treatment-related effect is similarly limited. In particular, a power calculation of
the study shows that for most endpoints (incidence and multiplicity of all tumors at all exposure
DCA concentrations), the type II error, which should be >50%, was less than 8%. The only
exception was for the incidence of adenomas and adenomas and carcinomas for the 0.5 g/L
treatment group (58%), at which, notably, there was a reported increase in reported adenomas or
adenomas and carcinomas combined over control (15% vs. 4%). Therefore, the likelihood of a
false null hypothesis was not negligible. Thus, while suggesting a lower response than for mice
for liver tumor induction, this study is inconclusive for determining of whether TCA induces a
carcinogenic response in the liver of rats.
For DCA, there are two reported long-term studies in rats (DeAngelo et al, 1996;
Richmond et al., 1995) that appear to have reported the majority of their results from the same
data set and which consequently were subject to similar design limitations and DCA-induced
neurotoxicity in this species. DeAngelo et al. (1996) reported increased hepatocellular adenomas
and carcinomas in male F344 rats exposed to DCA for 2 years. However, the data from
exposure concentrations at a 5 g/L dose had to be discarded and the 2.5 g/L DCA dose had to be
continuously lowered during the study due to neurotoxicity. There was a DCA-induced
increased in adenomas and carcinomas combined reported for the 0.5 g/L DCA (24.1 % vs. 4.4%
6/22/2009
540
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
adenomas and carcinomas combined in treated vs. controls) and an increase at a variable dose
started at 2.5 g/L DCA and continuously lowered (28.6% vs. 3.0% adenomas and carcinomas
combined in treated vs. controls). Only combined incidences of adenomas and carcinomas for
the 0.5 g/L DCA exposure group was reported to be statistically significant by the authors
although the incidence of adenomas was 17.2% vs. 4% in treated vs. control rats. Hepatocellular
tumor multiplicity was reported to be increased in the 0.5 g/L DCA group (0.31 adenomas and
carcinomas/animal in treated vs. 0.04 in control rats) but was reported by the authors to not be
statistically significant. At the starting dose of 2.5 g/L that was continuously lowered due to
neurotoxicity, the increased multiplicity of hepatocellular carcinomas was reported by the
authors to be to be statistically significant (0.25 carcinomas/animals vs. 0.03 in control) as well
as the multiplicity of combined adenomas and carcinomas (0.36 adenomas and
carcinomas/animals vs. 0.03 in control rats). Issues that affect the ability to determine the nature
of the dose-response for this study include (1) the use of a small number of animals (n = 23,
n = 21 and n = 23 at final sacrifice for the 2.0 g/L NaCl control, 0.05 g/L and 0.5 g/L treatment
groups) that limit the power of the study to both determine statistically significant responses and
to determine that there are not treatment-related effects (i.e. power) (2) apparent addition of
animals for tumor analysis not present at final sacrifice (i.e., 0.05 and 0.5 g/L treatment groups),
and (3) most of all, the lack of a consistent dose for the 2.5 g/L DCA exposed animals.
Similar issues are present for the study of Richmond et al. (1995) which was conducted
by the same authors as DeAngelo et al. (1996) and appeared to be the same data set. There was a
small difference in reports of the results between the two studies for the same data for the 0.5 g/L
DCA group in which Richmond et al. (1995) reported a 21% incidence of adenomas and
DeAngelo et al. (1996) reported a 17.2% incidence. The authors did not report any of the results
of DCA-induced increases of adenomas and carcinomas to be statistically significant. The same
issues discussed above for DeAngelo et al. (1996) apply to this study. Similar to the DeAngelo
et al. (1997) study of TCA in rats, the use in these DCA studies (DeAngelo et al., 1996;
Richmond et al., 1995) of relatively small numbers of rats limits the detection of treatment-
related effects and the ability to determine whether there was no treatment related effects (type II
error), especially at the low concentrations of DCA exposure.
For CH, George et al. (2000) exposed male F344/N rats to CH in drinking water for 2
years. Groups of animals were sacrificed at 13, 26, 52, and 78 weeks following the initiation of
dosing, with terminal sacrifices at week 104. Only a few animals received a complete
pathological examination. The number of animals surviving > 78 weeks and the number
examined for hepatocellular proliferative appeared to differ (42-44 animals examined but 32-35
surviving till the end of the experiment). Only the lowest treatment group had increased liver
tumors which were marginally significantly increased.
6/22/2009
541
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Leuschner and Beuscher (1998) examined the carcinogenic effects of CH in male and
female Sprague-Dawley rats (69-79 g, 25-29 days old at initiation of the experiment)
administered 0, 15, 45, and 135 mg/kg CH in unbuffered drinking water 7 days/week (n =
50/group) for 124 weeks in males and 128 weeks in females. Two control groups were noted in
the methods section without explanation as to why they were conducted as two groups. The
authors report no substance-related influence on organ weights and no macroscopic evidence of
tumors or lesions in male or female rats treated with CH for 124 or 128 weeks. However, no
data is presented on the incidence of tumors in either treatment or control groups. The authors
did report a statistically significant increase in the incidence of hepatocellular hypertrophy in
male rats at the 135 mg/kg dose (14/50 animals vs. 4/50 and 7/50 in controls I and II). For
female rats, the incidence of hepatocellular hypertrophy was reported to be 10/50 rats (control I)
and 16/50 (control II) rats with 18/50, 13/50 and 12/50 female rats having hepatocellular
hypertrophy after 15, 45, and 135 mg/kg CH, respectively. The lack of reporting in regard to
final body weights, histology, and especially background and treatment group data for tumor
incidences, limit the interpretation of this study. Whether this paradigm was sensitive for
induction of liver cancer cannot be determined.
Therefore, given the limitations in the available studies, a comparison of rat liver
carcinogenicity induced by TCE, TCA, DC A, and CH reveals no strong inconsistencies, but nor
does it provide much insight into the relative importance of different TCE metabolites in liver
tumor induction.
4.4.6.3.2 Studies in Mice
Similar to TCE, the bioassay data in mice for DCA, TCA, and CH is much more
extensive and have shown that all three compounds induce liver tumors in mice. Several two
year bioassays have been reported for CH (Daniel et al., 1992; George et al., 2000; Leakey et al.,
2003a). For many of the DCA and TCA studies, the focus was not carcinogenic dose-response
but rather investigation of the nature of the tumors and potential MOAs in relation to TCE. As a
result, studies often employed relatively high concentrations of DCA or TCA and/or were
conducted for a year or less. As shown previously in Section 4.4.4.2.1, the dose-response curves
for increased liver weight for TCE administration in male mice are more similar to those for
DCA administration and TCE oxidative metabolism than for direct TCA administration
(inadequate data was available for CH). An analogous comparison for DCA-, TCA-, and CH-
induced tumors would be informative, ideally using data from 2-year studies.
4.4.6.3.2.1 TCE carcinogenicity dose-response data
Unfortunately, the database for TCE, while consistently showing an induction of liver
tumors in mice, is very limited for making inferences regarding the shape of the dose-response
6/22/2009
542
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
curve. For many of these experiments multiplicity was not given only liver tumor incidence.
NTP (1990), Bull et al. (2002), Anna et al. (1994) conducted gavage experiments in which they
only tested one dose of - 1,000 mg/kg/day TCE. NCI (1976) tested two doses that were adjusted
during exposure to an average of 1,169 mg/kg/day and 2,339 mg/kg/day in male mice with only
2-fold dose spacing in only 2 doses tested. Maltoni et al. (1986) conducted inhalation
experiments in two sets of B6C3F1 mice and one set of Swiss mice at 3 exposure concentrations
that were 3-fold apart in magnitude between the low and mid-dose and 2-fold apart in magnitude
between the mid- and high-dose. However, for one experiment in male B6C3F1 mice (BT306),
the mice fought and suffered premature mortality and for two the experiments in B6C3F1 mice,
although using the same strain, the mice were obtained from differing sources with very different
background liver tumor levels. For the Maltoni et al. (1988) study a general descriptor of
"hepatoma" was used for liver neoplasia rather than describing hepatocellular adenomas and
carcinomas so that comparison of that data with those from other experiments is difficult. More
importantly, while the number of adenomas and carcinomas may be the same between treatments
or durations of exposure, the number of adenomas may decrease as the number of carcinomas
increase during the course of tumor progression. Such information is lost by using only a
hepatoma descriptor.
Given the limited database, it would be useful if different studies could be combined to
yield a more comprehensive dose-response curve, as was done for liver weight, above.
However, this is probably not appropriate for several reasons. First, only NTP (1990) was
performed with dosing duration and time of sacrifice both being the "standard" 104 weeks. NCI
(1976), Maltoni et al. (1986), Anna et al. (1994), and Bull et al. (2002) all had shorter dosing
periods and either longer (Maltoni et al., 1986) or shorter (the other three studies) observation
times. Therefore, because of potential dose-rate effects and differences in the degree of
expression of TCE-induced tumors, it is difficult to even come up with a comparable
administered dose metric across studies. Moreover, the background tumor incidences are
substantially different across experiments, even controlling for mouse strain and sex. For
example, across gavage studies in male B6C3F1 mice, the incidence of hepatocellular
carcinomas ranged from 1.2% to 16.7% (NCI, 1976; Anna et al., 1994; NTP, 1990) and the
incidence of adenomas ranged from 1.2% to 14.6% (Anna et al., 1994; NTP, 1990) in control
B6C3F1 mice. After ~ 1,000 mg/kg/day TCE treatment, the incidence of carcinomas ranged
from 19.4% to 62% (Bull et al., 2002; NCI, 1976; Anna et al., 1994; NTP, 1990), with three of
the studies (NCI, 1976; Anna et al., 1994; NTP, 1990) reporting a range of incidences between
42.8% to 62.0%). The incidence of adenomas ranged from 28% to 66.1% (Bull et al., 2002;
Anna et al., 1994; NTP, 1990). In the Maltoni et al. (1986) inhalation study as well, male
6/22/2009
543
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
B6C3F1 mice from two different sources had very different control incidences of hepatomas
(-2% versus about -20%).
Therefore, only data from the same experiment in which more than a single exposed dose
group was used provide reliable data on the dose-response relationship for TCE
hepatocarcinogenicity, and incidences from these experiments are shown in Figure 4.4.5 and
4.4.6. Except for one of the two Maltoni et al. (1986) inhalation experiments in male B6C3F1
mice, all of these datasets show relatively proportional increases with dose, albeit with somewhat
different slopes as may be expected across strains and sexes. Direct comparison is difficult,
since the "hepatomas" reported by Maltoni et al. (1986) are much more heterogeneous, including
neoplastic nodules, adenomas, and carcinomas, than the carcinomas reported by NCI (1976).
Nonetheless, although the data limitations preclude a conclusive statement, these data are
generally consistent with the linear relationship observed with TCE-induced liver weight
changes.
00%
mg/kg-d (oral gavage)
- NCI76 / B6C3F1 / F / oral
- NCI76 / B6C3F1 / M / oral
NTP90 / B6C3F1 / F / oral
NTP90 / B6C3F1 / M / oral
Bull02 / B6C3F1 / M / oral (aqueous)
a) '
o ra15
E -T10
500 1000 1500 2000
mg/kg-d (oral gavage)
2500
¦ NCI76 / B6C3F1 / F / oral
-NCI76 / B6C3F1 / M / oral
~ Anna94 / B6C3F1 / M / oral (corn oil controls)
Figure 4.4.5. Dose-response relationship, expressed as (A) % incidence and (B) fold-increase
over controls, for TCE hepatocarcinogenicity in NCI (1976). For comparison, incidences of
carcinomas for NTP (1990), Anna et al. (1994), and Bull et al. (2002) are included, but without
connecting lines since they are not appropriate for assessing the shape of the dose-response
relationship.
6/22/2009
544
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
-e- Maltoni86 / B6C3F1 / F / inhal I -©- Maltoni86 / B6C3F1 / F / inhal
-A- Maltoni86 / B6C3F1 / M / inhal [BT306] -A- Maltoni86 / B6C3F1 / M / inhal [BT306]
-A- Maltoni86 / B6C3F1 / M / inhal [BT306bis] -A- Maltoni86 / B6C3F1 / M / inhal [BT306bis]
-B- Maltoni86 / Swiss / M / inhal [ -B- Maltoni86 / Swiss / M / inhal
Figure 4.4.6. Dose-response relationship, expressed as (A) incidence and (B) fold-increase over
controls, for TCE hepatocarcinogenicity in Maltoni et al. (1986). Note that the BT306
experiment reported excessive mortality due to fighting, and so the paradigm was repeated in
experiment BT306bis using mice from a different source.
40%
35%
(D
c 30%
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
power of the experiment at this dose was only 16.9% to be able to determine that there was not a
treatment related effect. Indeed, Figure 4.4.7 replots the data from DeAngelo et al. (1999) with
an abscissa drawn to scale (unlike the figure in the original paper, which was not to scale),
suggests even a slightly greater than linear effect at the lowest dose (0.05 g/L, or 8 mg/kg-d) as
compared to the next lowest dose (0.5 g/L, or 84 mg/kg-d), though of course the power of such a
determination is limited. The authors did not report the incidence or multiplicity of adenomas
for the 0.05 g/L exposure group in the study or the incidence or multiplicity of adenomas and
carcinomas in combination. For the animals surviving from 79 to 100 weeks of exposure, the
incidence and multiplicity of adenomas peaked at 1 g/L while hepatocellular carcinomas
continued to increase at the higher doses. This would be expected where some portion of the
adenomas would either regress or progress to carcinomas at the higher doses.
100%
80%
a)
o
a)
u
60%
- 40%
o
20%
100 200 300 400
DCA mg/kg-d
500
100
200 300
DCA mg/kg-d
400
500
Figure 4.4.7. Dose-response data for hepatocellular carcinomas (HC) (A) incidence and (B)
multiplicity, induced by DCA from DeAngelo et al. (1999). Drinking water concentrations were
0, 0.05, 0.5, 1, 2, and 3.5 g/L, from which daily average doses were calculated using observed
water consumption in the study.
Associations of DCA carcinogenicity with various non-cancer, possibly precursor, effects
was also investigated. Importantly, the doses that induced tumors in DeAngelo et al. (1999)
were reported to not induce widespread cytotoxicity. An attempt was also made to relate
differing exposure levels to subchronic changes and peroxisomal enzyme induction.
Interestingly, DeAngelo et al. (1999) reported that peroxisome proliferation was significantly
increased at 3.5 g/L DCA only at 26 weeks, not correlated with tumor response, and to not be
increased at either 0.05 g/L or 0.5 g/L treatments. The authors concluded that DCA-induced
6/22/2009
546
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
carcinogenesis was not dependent on peroxisome proliferation or chemically sustained
proliferation, as measured by DNA synthesis. Slight hepatomegaly was present by 26 weeks in
the 0.5 g/L group and decreased with time. By contrast, increases in both % liver/body weight
and the multiplicity of hepatocellular carcinomas increased proportionally with DC A exposure
concentration after 79-100 weeks of exposure. DeAngelo et al. (1999) presented a figure
comparing the number of hepatocellular carcinomas/animal at 100 weeks compared with the %
liver/body weight at 26 weeks that showed a linear correlation (r2 = 0.9977) while peroxisome
proliferation and DNA synthesis did not correlate with tumor induction profiles. The
proportional increase in liver weight with DCA exposure was also reported for shorter durations
of exposure as noted previously. Therefore, for DCA, both tumor incidence and liver weight
appear to increase proportionally with dose.
4.4.6.3.2.3 TCA carcinogenicity dose-response data
With respect to TCA, Pereira (1996) reported that for 82 week exposure to TCA in
female B6C3F1 mice, TCA exposure concentrations of 0, 2, 6.67, and 20 mmol/L (0, 0.33, 1.1,
and 3.3 g/L) led to increasing incidences and multiplicity of adenomas and of carcinomas (Figure
4.4.8). DeAngelo et al. (2008) reported the results of three experiments exposing male B6C3F1
mice to neutralized TCA in drinking water (incidences also in Figure 4.4.8). Rather than using 5
exposure levels that were generally 2-fold apart, as was done in DeAngelo et al. (1999) for DCA,
DeAngelo et al. (2008) studied only 3 doses of TCA that were an order of magnitude apart which
limits the elucidation of the shape of the dose-response curve. In addition, the 104-week data,
DeAngelo et al. (2008) contained 2 studies, each conducted in a separate laboratories - the two
lower doses were studied in one study and the highest dose in another. The first 104-week study
was conducted using 2 g/L NaCl, or 0.05, 0.5, or 5 g/L TCA in drinking water for 60 weeks
(Study #1) while the other two were conducted for a period of 104 weeks (Study #2 with 2.5 g/L
neutralized acetic acid or 4.5 g/L TCA exposure groups and Study #3 with deionized water, 0.05
g/L TCA and 0.5 g/L TCA exposure groups). In addition, a relatively small number of animals
were used for the determination of a tumor response (n ~ 30 at final necropsy).
6/22/2009
547
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
100%
80%
60%
40%
20%
0%
0
2
4
6
TCA concentration (g/l)
-A- DeAngelo et al. (2008) 60 wk (Study #1, M)
-e-Pereira (1996) 82 wk (F)
DeAngelo et al. (2008) 104 wk (Study #2, M)
-A-DeAngelo et al. (2008) (Study #3, M)
o
Q.
| 0.6
O 0.4
X 0.2
0
2
4
6
TCA concentration (g/l)
-A- DeAngelo et al. (2008) 60 wk (Study #1, M)
-9-Pereira (1996) 82 wk (F)
-H- DeAngelo et al. (2008) 104 wk (Study #2, M)
-A- DeAngelo et al. (2008) (Study #3, M)
Figure 4.4.8. Reported incidences of hepatocellular carcinomas (HC) and adenomas plus
carcinomas (HA+HC) in various studies in B6C3F1 mice (Pereira, 1996; DeAngelo et al., 2008).
Combined HA+HC were not reported in (Pereira, 1996).
In Study #1, the incidence data for adenomas observed at 60 weeks at 0.05 g/L, 0.5 g/L
and 5.0 g/L TCA was 2.1-fold, 3.0-fold and 5.4-fold of control values, with similar fold increases
in multiplicity. As shown by Pereira (1996), 60 weeks does not allow for full tumor expression,
so whether the dose-response relationship is the same at 104 weeks is not certain. For instance,
Pereira (1996) examined the tumor induction in female B6C3 F1 mice and demonstrated that
foci, adenoma, and carcinoma development in mice are dependent on duration of exposure
(period of observation in controls). In control female mice a 360- vs. 576-day observation period
showed that at 360 days no foci or carcinomas and only 2.5% of animals had adenomas whereas
by 576 days of observation, 11% had foci, 2% adenomas, and 2% had carcinomas. For DCA and
TCA treatments, foci, adenomas, and carcinoma incidence and multiplicity did not reach full
expression until 82 weeks at the 3 doses employed. Although the numbers of animals were
relatively low and variable at the two highest doses (18-28 mice) there were 50-53 mice studied
at the lowest dose level and 90 animals studied in the control group.
Therefore, the 104-week DeAngelo et al. (2008) data from Studies #2 and #3 would
generally be preferred for elucidating the TCA dose-response relationship. However, Study #2
was only conducted at one dose, and although Study #3 used lower doses, it exhibited
extraordinarily high control incidences of liver tumors. In particular, while the incidence of
adenomas and carcinomas was 12% in Study #2, it was reported to be 64% in Study #3. The
mice in Study #3 were of very large size (weighing -50 g at 45 weeks) as compared to Study #1,
6/22/2009
548
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Study #2, or most other bioassays in general, and the large background rate of tumors reported is
consistent with the body-weight-dependence observed by Leakey et al. (2003b).
To put into context the 64% incidence data for carcinomas and adenomas reported in
DeAngelo et al. (2008) for the control group of Study #3, other studies cited in this review for
male B6C3F1 mice show a much lower incidence in liver tumors with: (1) NCI (1976) study of
TCE reporting a colony control level of 6.5% for vehicle and 7.1%> incidence of hepatocellular
carcinomas for untreated male B6C3F1 mice (n = 70-77) at 78 weeks, (2) Herren-Freund et al.
(1987) reporting a 9%> incidence of adenomas in control male B6C3F1 mice with a multiplicity
of 0.09 ± 0.06 and no carcinomas (n = 22) at 61 weeks, (3) NTP (1990) reporting an incidence of
14.6% adenomas and 16.6%> carcinomas in male B6C3F1 mice after 103 weeks (n = 48), and (4)
Maltoni et al. (1986) reporting that B6C3F1 male mice from the "NCI source" had a 1.1%
incidence of "hepatoma" (carcinomas and adenomas) and those from "Charles River Co." had a
18.9%) incidence of "hepatoma" during the entire lifetime of the mice (n = 90 per group). The
importance of examining an adequate number of control or treated animals before confidence
can be placed in those results in illustrated by Anna et al. (1994) in which at 76 weeks 3/10
control male B6C3F1 mice that were untreated and 2/10 control animals given corn oil were
reported to have adenomas but from 76 to 134 weeks, 4/32 mice were reported to have adenomas
(multiplicity of 0.13 ± 0.06) and 4/32 mice were reported to have carcinomas (multiplicity of
0.12 ± 0.06). Thus, the reported combined incidence of carcinomas and adenomas of 64%>
reported by DeAngelo et al. (2008) for the control mice of Study # 3, not only is inconsistent and
much higher than those reported in Studies #1 and #2, but also much higher than reported in a
number of other studies of TCE.
Therefore, this large background rate and the increased mortality for these mice limit
their use for determining the nature of the dose-response for TCA liver carcinogenicity. At the
two lowest doses of 0.05 g/L and 0.5 g/L TCA from Study #3, the differences in the incidences
and multiplicities for all tumors were 2-fold at 104 weeks. However, there was no difference in
any of the tumor results (i.e., adenoma, carcinoma, and combinations of adenoma and carcinoma
incidence and multiplicity) between the 4.5 g/L dose group in Study #2 and the 0.5 g/L dose
group in Study #3 at 104 weeks. By contrast, at 60 weeks of exposure, but within the same study
(Study #1), there was a 2-fold increase in multiplicity for adenomas, and for adenomas and
carcinomas combined between the 0.5 and 5.0 g/L TCA exposure groups. These results are
consistent with the two highest exposure levels reaching a plateau of response after a long
enough duration of exposure for full expression of the tumors (i.e., ~ 90% of animals having
liver tumors at the 0.5 g/L and 5 g/L exposures). However, whether such a plateau would have
been observed in mice with a more "normal" body weight, and hence a lower background tumor
burden cannot be determined.
6/22/2009
549
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Because of the limitations of different studies, it is difficult to discern whether the liver
tumor dose-response curves of TCA and DCA are different in a way analogous to that for liver
weight (see Figure 4.4.9). Certainly, it is clear that at the same concentration in drinking water
or estimated applied dose, DCA is more potent than TCA, as DCA induces nearly 100%
incidence of carcinomas at a lower dose than TCA. Therefore, like with liver weight gains, DCA
has a steeper dose-response function than TCA. However, the evidence for a "plateau" in tumor
response at high doses with TCA, as was observed for liver weight, is equivocal, as it is
confounded by the highly varying background tumor rates and the limitations of the available
study paradigms.
100%
90%
80%
g 70%
g 60%
^ 50%
.E 40%
O 30%
X 20%
10%
0%
0
200 400 600
mg/kg-d
DeAngelo et al. (2008) (TCA Study #2)
-A- DeAngelo et al. (2008) (TCA Study #3)
-O- DeAngelo et al. (1999) (DCA)
Figure 4.4.9. Reported incidence of hepatocellular carcinomas induced by DCA and TCA in 104
week studies (DeAngelo et al., 1999, 2008). Only carcinomas were reported in DeAngelo et al.
(1999), so combined adenomas and carcinomas could not be compared.
DeAngelo et al. (2008) attempt to identify a NOEL for tumorigenicity using tumor
multiplicity data and estimated TCA dose. However, it is not an appropriate descriptor for these
data, especially given that "statistical significance" of the tumor response is the determinant used
by the authors to support the conclusions regarding a dose in which there is no TC A-induced
effect. Due to issues related to the appropriateness of use of the concurrent control in Study #3,
only the 60-week experiment (i.e., Study # 1) is useful for the determination of tumor dose-
response. Not only is there not allowance for full expression of a tumor response at the 60-week
time point but a power calculation of the 60-week study shows that the type II error, which
should be > 50% and thus greater than the chances of "flipping a coin", was 41% and 71% for
6/22/2009
550
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
incidence and 7% and 15% for multiplicity of adenomas for the 0.05 g/L and 0.5 g/L TCA
exposure groups. For the combination of adenomas and carcinomas, the power calculation was
8% and 92% for incidence and 6% and 56% for multiplicity at 0.05 g/L and 0.5 g/L TCA
exposure. Therefore, the designed experiment could accept a false null hypothesis, especially in
terms of tumor multiplicity, at the lower exposure doses and erroneously conclude that there is
no response due to TCA treatment.
In terms of correlations with other non-cancer, possibly precursor effects, DeAngelo et al.
(2008) also reported that PCO activity, which varied 2.7-fold as baseline controls, was 1.3-fold,
2.4-fold, and 5.3-fold of control for the 0.05 g/L, 0.5 g/L and 5 g/L TCA exposure groups in
Study #1 at 4 weeks was for adenomas incidence 2.1-, 3.0-, and 5.4-fold of control and not
similar at the lowest dose level at 60 weeks. However, it is not clear whether the similarly
between PCO and carcinogenicity at 60 weeks would persist for tumor incidence at 104 weeks.
DeAngelo et al. (2008) report a regression analyses that compare "percent of hepatocellular
neoplasia," indicated by tumor multiplicity, with TCA dose, represented by estimations of the
TCA dose in mg/kg/day, and with PCO activity for the 60-week and 104-week data. Whether
adenomas and carcinomas combined or individual tumor type were used in these analysis was
not reported by the authors. However, it would be preferable to compare "precursor" levels of
PCO at earlier time points, rather than at a time when there was already a significant tumor
response. In addition, linear regression analyses of this data are difficult to interpret because of
the wide dose spacing of these experiments. In such a situation, for a linear regression, control
and 5 g/L exposure levels will basically determine the shape of the dose response curve since the
0.05 g/L and 0.5 g/L exposure levels are so close to the control (0) value. Thus, dose response
appears to be linear between control and the 5.0 g/L value with the two lowest doses not
affectively changing the slope of the line (i.e., "leveraging" the regression). Moreover, at the 5
g/L dose level, there is potential for effects due to palatability, as reported in one study in which
drinking water consumption declined at this concentration (DeAngelo et al., 2008). Thus, the
value of these analyses is limited by (1) use of data from Study # 3 in a tumor prone mouse that
is not comparable to those used in Studies #1 and #2, (2) the appropriateness of using PCO
values from later time points and the variability in PCO control values, (3) the uncertainty of the
effects of palatability on the 5 g/L TCA results which were reported in one study to reduce
drinking water consumption, and (4) the dose-spacing of the experiment.
4.4.6.3.2.4 CH carcinogenic dose-response
Although a much more limited database in rodents than for TCA or DC A, there is
evidence that Chloral hydrate is also a rodent liver hepatocarcinogen (see also Section 2.5 of
Appendix E and Caldwell and Keshava [2006]).
6/22/2009
551
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Daniel et al. (1992) exposed adult male B6C3F1 28-day old mice to 1 g/L CH in drinking
water for 30 and 60 weeks (n = 5 for interim sacrifice) and for 104 weeks (n = 40). The
concentration of CH was 1 g/L and estimated to provide a 166 mg/kg/day dose. It is not clear
from the report what control group better matched the CH group, as the mean initial body
weights of the groups as well as the number of animals varied considerably in each group (i.e., ~
40% difference in mean body weights at the beginning of the study). Liver tumors were
increased by CH treatment. The % incidence of liver carcinomas and adenomas in the surviving
animals was 15% in control and 71% in CH treated mice and the incidence of hepatocellular
carcinoma reported to be 46% in the CH treated group. The number of tumors/animals was also
significantly increased with CH treatment. However, because this was a single dose study, a
comparison with the dose-response relationship with TCE, TCA, or DCA is not feasible.
George et al. (2000) exposed male B6C3F1 mice to CH in drinking water for 2 years.
Groups of animals were sacrificed at 26, 52, and 78 weeks following the initiation of dosing,
with terminal sacrifices at week 104. Only a few animals received a complete pathological
examination. Preneoplastic foci and adenomas were reported to be increased in the livers of all
CH treatment groups at 104 weeks. The % incidence of hepatocellular adenomas was reported to
be 21.4%, 43.5%), 51.3%, and 50% in control, 13.5, 65.0 and 146.6 mg/kg/day CH treatment
groups, respectively. The % incidence of hepatocellular carcinomas was reported to be 54.8%,
54.3%), 59.0%) and 84.4% in these same groups. The resulting % incidence of hepatocellular
adenomas and carcinomas was reported to be 64.3%, 78.3%, 79.5% and 90.6%. Of concern is
the reporting of a 64% incidence of hepatocellular carcinomas and adenomas in the control group
of mice for this experiment, which is the same as that for another study published by this same
laboratory (DeAngelo et al., 2008). DeAngelo et al. (2008) did not identify them as being
contemporaneous studies or sharing controls, but a comparison of the control data published by
DeAngelo et al. (2008) for TCA and that published by George et al. (2000) for the CH studies
shows them to be the same data set. Therefore, as discussed above, this data set was derived
from B6C3F1 mice that were large (-50 g) and resultantly tumor prone, making determinations
of the dose-response of CH from this experiment difficult. Therefore, for the purposes of
comparison of dose-response relationships, this study has the same limitations as the DeAngelo
et al. (2008) study, discussed above.
Leakey et al. (2003a) studied the effects of CH exposure (0, 25, 50, and 100 mg/kg/day, 5
days/week, 104-105 weeks via gavage) in male B6C3F1 mice with dietary control used to
manipulate body growth (n = 48 for 2 year study and n = 12 for the 15-month interim study).
Dietary control was reported to decrease background liver tumor rates (decreased by 15-20%)
and was reported to be associated with decreased variation in liver-to-body weight ratios, thereby
potentially increasing assay sensitivity. In dietary-controlled groups and groups fed ad libitum,
6/22/2009
552
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
liver adenomas and carcinomas (combined) were reported to be increased with CH treatment.
With dietary restriction there was a more discernable CH tumor-response with overall tumor
incidence reduced, and time-to-tumor increased by dietary control in comparison to ad libitum
fed mice. Incidences of hepatocellular adenoma and carcinoma overall rates were reported to be
33%, 52%, 49%, and 46% for control, 25mg/kg, 50 mg/kg, and 100 mg/kg ad libitum-fed mice,
respectively. For dietary controlled mice the incidence rates were reported to be 22.9%, 22.9%,
29.2% and 37.5% for controls, 25mg/kg, 50 mg/kg, and 100 mg/kg CH, respectively. Body
weights were matched and carefully controlled in this study. These data are shown in Figure
4.4.10, relative to control incidences. It is evident from these data that dietary control
significantly changes the apparent shape of the dose response curve, presumably by reducing
variability between animals. While the ad libitum dose groups had an apparent "saturation" of
response, this was not evident with the dietary controlled group. Of note all the other bioassays
for TCE, TCA, DC A, and CH were in ad libitum fed mice. Therefore it is difficult to compare
the dose-response curves for CH-treated mice on dietary restriction to those fed ad libitum.
However, the rationale for dietary restriction in the B6C3F1 mouse is to prevent the types of
weight gain and corresponding high background tumor levels observed in DeAngelo et al. (2008)
and George et al. (2000). As stated previously, most other studies of TCA, DCA, and TCE had
background levels that, while varied, were lower than the ad libitum fed mice studied in Leakey
et al. (2003a).
0%
0 20 40 60 80 100
CH mg/kg-d
-©-ad libitum -b-dietary control
Figure 4.4.10. Effects of dietary control on the dose-response curves for changes in liver tumor
incidences induced by CH in diet (Leakey et al., 2003a).
60%
o
10%
6/22/2009
553
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Of note is that incidences of adenomas and carcinomas combined do not show
differences in tumor progression as carcinomas may increase and adenomas may regress. Liver
weight increases at 15-months did not correlate with 2-year tumor incidences in the ad libitum
group, but a consistent dose-response shape between these two measures is evident in the dietary
controlled group. However, of note is the reporting of liver weight at 15 months is for a time
period in which foci and liver tumors have been reported to have already occurred in other
studies, so hepatomegaly in the absence of these changes is hard to detect.
In terms of other non-cancer effects that may be associated with tumor induction, it is
notable that while dietary restriction reduced the overall level of CH-mediated tumor induction,
it led to greater CH-mediated induction of peroxisome proliferation-associated enzymes.
Moreover, between control groups, dietary restricted mice appeared to have higher levels of
lauric acid co-hydrolase activity than ad libitum-fed mice. Seng et al. (2003) report that lauric
acid P-hydroxylase and PCO were induced only at exposure levels >100 mg/kg CH, again with
dietary restricted groups showing the greatest induction. Such data argues against the role of
peroxisome proliferation in CH-liver tumor induction in mice.
Leakey et al. (2003a) gave no descriptions of liver pathology were given other than
incidence of mice with fatty liver changes. Hepatic malondialdehyde concentration in ad libitum
fed and dietary controlled mice did not change with CH exposure at 15 months but the dietary
controlled groups were all ~ half that of the ad libitum-fed mice. Thus, while overall increased
tumors observed in the ad libitum diet correlated with increased malondialdehyde concentration,
there was no association between CH dose and malondialdehyde induction for either diet.
Overall, from the CH studies in mice, there is an apparent increase in liver adenomas and
carcinomas induced by CH treatment by either drinking water or gavage with all available
studies performed in male B6C3F1 mice. However, the background levels of hepatocellular
adenomas and carcinomas in these mice in George et al. (2000) and body weight data from this
study are high, consistent with the association between large body weight and background tumor
susceptibility shown with dietary control (Leakey et al., 2003a). With dietary control, Leakey et
al. (2003a) report a dose-response relationship between exposure and tumor incidence that is
proportional to dose.
4.4.6.3.2.5 Degree of concordance among TCE, TCA, DCA, and CH dose-response relationships
Comparison of the dose-response for TCE hepatocarcinogenicity with that for TCA and
DCA is weakly suggestive a better concordance in dose-response shape between TCE and DCA
or TCE and CH than between TCE and TCA. However, differences across the databases of these
compounds, especially with respect to the comparability of study durations and control tumor
incidences, preclude a definitive conclusion from these data.
6/22/2009
554
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.4.6.3.3 Inferences from liver tumor phenotype and genotype
A number of studies have investigation tumor phenotypes, such as c-Jun staining,
tincture, and dysplacity, or genotypes, such as H-ras mutations, to inform both the identification
of the active agents of TCE liver tumor induction as well as what MOA(s) may be involved.
4.4.6.3.3.1 Tumor phenotype - staining and appearance
The descriptions of tumors in mice reported by the NCI, NTP, and Maltoni et al studies
are also consistent with phenotypic heterogeneity as well as spontaneous tumor morphology (see
Section 3.4.1.5 of Appendix E). As noted in Section 3.1 of Appendix E, hepatocellular
carcinomas observed in humans are also heterogeneous. For mice, Maltoni et al. (1986)
described malignant tumors of hepatic cells to be of different sub-histotypes, and of various
degrees of malignancy and were reported to be unique or multiple, and have different sizes
(usually detected grossly at necropsy) from TCE exposure. In regard to phenotype, tumors were
described as usual type observed in Swiss and B6C3F1 mice, as well as in other mouse strains,
either untreated or treated with hepatocarcinogens and to frequently have medullary (solid),
trabecular, and pleomorphic (usually anaplastic) patterns. For the NC I (1976) study, the mouse
liver tumors were described in detail and to be heterogeneous "as described in the literature" and
similar in appearance to tumors generated by carbon tetrachloride. The description of liver
tumors in this study and tendency to metastasize to the lung are similar to descriptions provided
by Maltoni et al. (1986) for TCE-induced liver tumors in mice via inhalation exposure. The NTP
(1990) study reported TCE exposure to be associated with increased incidence of hepatocellular
carcinoma (tumors with markedly abnormal cytology and architecture) in male and female mice.
Hepatocellular adenomas were described as circumscribed areas of distinctive hepatic
parenchymal cells with a perimeter of normal appearing parenchyma in which there were areas
that appeared to be undergoing compression from expansion of the tumor. Mitotic figures were
sparse or absent but the tumors lacked typical lobular organization. Hepatocellular carcinomas
were reported to have markedly abnormal cytology and architecture with abnormalities in
cytology cited as including increased cell size, decreased cell size, cytoplasmic eosinophilia,
cytoplasmic basophilia, cytoplasmic vacuolization, cytoplasmic hyaline bodies and variations in
nuclear appearance. Furthermore, in many instance several or all of the abnormalities were
reported to be present in different areas of the tumor and variations in architecture with some of
the hepatocellular carcinomas having areas of trabecular organization. Mitosis was variable in
amount and location. Therefore the phenotype of tumors reported from TCE exposure was
heterogeneous in appearance between and within tumors from all 3 of these studies.
Caldwell and Keshava (2006) report "that Bannasch (2001) and Bannasch et al. (2001)
describe the early phenotypes of preneoplastic foci induced by many oncogenic agents (DNA-
6/22/2009
555
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
reactive chemicals, radiation, viruses, transgenic oncogenes and local hyperinsulinism) as
insulinomimetic. These foci and tumors have been described by tincture as eosinophilic and
basophilic and to be heterogeneous. The tumors derived from them after TCE exposure are
consistent with the description for the main tumor lines of development described by Bannasch
etal. (2001) (see Section 3.4.1.5). Thus, the response of liver to DCA (glycogenesis with
emergence of glycogen poor tumors) is similar to the progression of preneoplastic foci to tumors
induced from a variety of agents and conditions associated with increased cancer risk."
Furthermore Caldwell and Keshava (2006) note that Bull et al. (2002) report expression of IR to
be elevated in tumors of control mice or mice treated with TCE, TCA and DCA but not in
nontumor areas suggesting that this effect is not specific to DCA.
There is a body of literature that has focused on the effects of TCE and its metabolites
after rats or mice have been exposed to "mutagenic" agents to "initiate" hepatocarcinogenesis
and this is discussed in Section 4.2 of Appendix E. TCE and its metabolites were reported to
affect tumor incidence, multiplicity, and phenotype when given to mice as a coexposure with a
variety of "initiating" agents and with other carcinogens. Pereira and Phelps (1996) reported that
MNU alone induced basophilic foci and adenomas. MNU and low concentrations of DCA or
TCA in female mice were reported to induce heterogeneous for foci and tumor with a higher
concentration of DCA inducing more eosinophilic and a higher concentration of TCA inducing
more tumors that were basophilic. Pereira et al. (2001) reported that not only dose, but gender
also affected phenotype in mice that had already been exposed to MNU and were then exposed
to DCA. As for other phenotypic markers, Lantendresse and Pereira (1997) reported that
exposure to MNU and TCA or DCA induced tumors that had some commonalities, were
heterogeneous, but for female mice were overall different between DCA and TCA as
coexposures with MNU.
With regard to the phenotype of TCA and DCA-induced tumors, Stauber and Bull (1997)
reported the for male B6C3F1 mice, DCA-induced "lesions" contained a number of smaller
lesions that were heterogeneous and more eosinophilic with larger "lesions" tending to less
numerous and more basophilic. For TCA results using this paradigm, the "lesions" were
reported to be less numerous, more basophilic, and larger than those induced by DCA. Carter et
al. (2003) used tissues from the DeAngelo et al. (1999) and examined the heterogeneity of the
DCA-induced lesions and the type and phenotype of preneoplastic and neoplastic lesions pooled
across all time points. Carter et al. (2003) examined the phenotype of liver tumors induced by
DCA in male B6C3 F1 mice and the shape of the dose-response curve for insight into its MOA.
They reported a dose-response of histopathologic changes (all classes of premalignant lesions
and carcinomas) occurring in the livers of mice from 0.05-3.5 g/L DCA for 26-100 weeks and
suggest foci and adenomas demonstrated neoplastic progression with time at lower doses than
6/22/2009
556
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
observed DCA genotoxicity. Preneoplastic lesions were identified as eosinophilic, basophilic
and/or clear cell (grouped with clear cell and mixed cell) and dysplastic. Altered foci were 50%
eosinophilic with about 30% basophilic. As foci became larger and evolved into carcinomas
they became increasingly basophilic. The pattern held true through out the exposure range.
There was also a dose and length of exposure related increase in atypical nuclei in "non-
involved" liver. Glycogen deposition was also reported to be dose-dependent with periportal
accumulation at the 0.5 g/L exposure level. Carter et al. (2003) suggested that size and evolution
into a more malignant state are associated with increasing basophilia, a conclusion consistent
with those of Bannasch (1996) and that there a greater periportal location of lesions suggestive as
the location from which they arose. Consistent with the results of DeAngelo et al. (1999), Carter
et al. (2003) reported that DCA (0.05-3.5 g/L) increased the number of lesions per animal
relative to animals receiving distilled water, shortened the time to development of all classes of
hepatic lesions, and that the phenotype of the lesions were similar to those spontaneously arising
in controls. Along with basophilic and eosinophilic lesions or foci, Carter et al. (2003)
concluded that DCA-induced tumors also arose from isolated, highly dysplastic hepatocytes in
male B6C3F1 mice chronically exposed to DCA suggesting another direct neoplastic conversion
pathway other than through eosinophilic or basophilic foci.
Rather than male B6C3F1 mice, Pereira (1996) studied the dose-response relationship for
the carcinogenic activity of DCA and TCA and characterized their lesions (foci, adenomas and
carcinomas) by tincture in females (the generally less sensitive gender). Like the studies of TCE
by Maltoni et al. (1986), female mice were also reported to have increased liver tumors after
TCA and DCA exposures. Pereira (1996) pool lesions were pooled for phenotype analysis so the
affect of duration of exposure could not be determined nor adenomas separated from carcinomas
for "tumors." However, as the concentration of DCA was decreased the number of foci was
reported by Pereira (1996) to be decreased but the phenotype of the foci to go from primarily
eosinophilic foci (i.e. ~ 95% eosinophilic at 2.58 g/L DCA) to basophilic foci (~ 57%
eosinophilic at 0.26 g/L). For TCA the number of foci was reported to ~ 40 basophilic and ~ 60
eosinophilic regardless of dose. Spontaneously occurring foci were more basophilic by a ratio of
7/3. Pereira (1996) described the foci of altered hepatocytes and tumors induced by DCA in
female B6C3F1 mice to be eosinophilic at higher exposure levels but at lower or intermittent
exposures to be half eosinophilic and half basophilic. Regardless of exposure level, half of the
TCA-induced foci were reported to be half eosinophilic and half basophilic with tumors 75%
basophilic. In control female mice, the limited numbers of lesions were mostly basophilic, with
most of the rest being eosinophilic with the exception of a few mixed tumors. The limitations of
descriptions tincture and especially for inferences regarding peroxisome proliferator from the
description of "basophilia" is discussed in Section 3.4.1.5 of Appendix E.
6/22/2009
557
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Thus, the results appear to differ between male and female B6C3F1 mice in regard to tincture for
DCA and TCA at differing doses. What is apparent is that the tincture of the lesions is
dependent on the stage of tumor progression, agent (DCA or TCA), gender, and dose. Also what
is apparent from these studies is the both DCA and TCA are heterogeneous in their tinctoral
characteristics
Overall, tumors induced by TCA, DCA, CH, and TCE are all heterogeneous in their
physical and tinctural characteristics in a manner this not markedly distinguishable from
spontaneous lesions or those induced by a wide variety of chemical carcinogens. For instance,
Daniel et al. (1992), which studies DCA and CH carcinogenicity (discussed above) noted that
morphologically, there did not appear to be any discernable differences in the visual appearance
of the DCA- and CH-induced tumors. Therefore, these data do not provide strong insights into
elucidating the active agent(s) for TCE hepatocarcinogenicity or their MOA(s).
4.4.6.3.3.2 C-Jun staining
Stauber and Bull (1997) reported that in male B6C3F1 mice, the oncoproteins c-jun and
c-Fos were expressed in liver tumors induced by DCA but not those induced by TCA. Although
Bull et al. (2004) have suggested that the negative expression of c-jun in TCA-induced tumors
may be consistent with a characteristic phenotype shown in general by peroxisome proliferators
as a class, as pointed out by Caldwell and Keshava (2006), there is no supporting evidence of
this. Nonetheless, the observation that TCA and DCA have different levels of oncogene
expression led to a number of follow-up studies by this group. No data on oncoprotein
immunostaining are available for CH.
Stauber et al. (1998) studied induction of "transformed" hepatocytes by DCA and TCE
treatment in vitro, including an examination of c-Jun staining. Stauber et al. (1998) isolated
primary hepatocytes from 5-8 week old male B6C3F1 mice (n = 3) and subsequently cultured
them in the presence of DCA or TCA. In a separate experiment 0.5 g/L DCA was given to mice
as pretreatment for 2 weeks prior to isolation. The authors assumed that the anchorage-
independent growth of these hepatocytes was an indication of an "initiated cell." After 10 days
in culture with DCA or TCA (0, 0.2, 0.5 and 2.0 mM), concentrations of 0.5 mM or more DCA
and TCA both induced an increase in the number of colonies that was statistically significant,
with DCA showing dose-dependence as well as slightly greater overall increases than TCA. In a
time course experiment the number of colonies from DCA treatment in vitro peaked by 10 days
and did not change through days 15-25 at the highest dose and, at lower concentrations of DCA,
increased time in culture induced similar peak levels of colony formation by days 20-25 as that
reached by 10 days at the higher dose. Therefore, the number of colonies formed was
independent of dose if the cells were treated long enough in vitro. However, not only did
6/22/2009
558
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
treatment with DCA or TCA induce anchorage independent growth but untreated hepatocytes
also formed larger numbers of colonies with time, although at a lower rate than those treated
with DCA. The level reached by untreated cells in tissue culture at 20 days was similar to the
level induced by 10 days of exposure to 0.5 mM DCA. The time course of TCA exposure was
not tested to see if it had a similar effect with time as did DCA. The colonies observed at 10
days were tested for c-Jun expression with the authors noting that "colonies promoted by DCA
were primarily c-Jun positive in contrast to TCA promoted colonies that were predominantly c-
Jun negative." Of the colonies that arose spontaneously from tissue culture conditions, 10/13
(76.9%) were reported to be c-Jun +, those treated with DCA 28/34 (82.3%) were c-Jun +, and
those treated with TCA 5/22 {22.1%) were c-Jun +. Thus, these data show heterogeneity in cell
in colonies but with more that were c-Jun + colonies occurring by tissue culture conditions alone
than in the presence of DCA, rather than in the presence of TCA.
Bull et al. (2002) administered TCE, TCA, DCA, and combinations of TCA and DCA to
male B6C3F1 mice by daily gavage (TCE) or drinking water (TCA, DCA, and TCA+DCA) for
52-79 weeks, in order to compare a number of tumor characteristics, including c-Jun expression,
across these different exposures. Bull et al. (2002) reported lesion reactivity to c-Jun antibody to
be dependent on the proportion of the DCA and TCA administered after 52 weeks of exposure.
Given alone, DCA was reported to produce lesions in mouse liver for which approximately half
displayed a diffuse immunoreactivity to a c-Jun antibody, half did not, and none exhibited a
mixture of the two. After TCA exposure alone, no lesions were reported to be stained with this
antibody. When given in various combinations, DCA and TCA co-exposure induced a few
lesions that were only c-Jun+, many that were only c-Jun-, and a number with a mixed phenotype
whose frequency increased with the dose of DCA. For TCE exposure of 79 weeks, TCE-induced
lesions were reported to also have a mixture of phenotypes (42% c-Jun+, 34% c-Jun-, and 24%
mixed) and to be most consistent with those resulting from DCA and TCA coexposure but not
either metabolite alone.
A number of the limitations of the experiment are discussed in Caldwell et al. (2008)
Specifically, for the DCA and TCA exposed animals, the experiment was limited by low
statistical power, a relatively short duration of exposure, and uncertainty in reports of lesion
prevalence and multiplicity due to inappropriate lesions grouping (i.e., grouping of hyperplastic
nodules, adenomas, and carcinomas together as "tumors"), and incomplete histopatholology
determinations (i.e., random selection of gross lesions for histopathology examination). For
determinations of immunoreactivity to c-Jun, Bull et al. (2002) combined hyperplastic nodules,
adenomas, and carcinomas in most of their treatment groups, so differences in c-Jun expression
across differing types of lesions were not discernable.
6/22/2009
559
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Nonetheless, these data collectively strongly suggest that TCA is not the sole agent of
TCE-induced mouse liver tumors. In particular, TCE-induced tumors that were, in order of
frequency, c-Jun+, c-Jun-, and of mixed phenotype, while c-Jun+ tumors have never been
observed with TCA treatment. Nor do these data support DCA as the sole contributor, since
mixed phenotypes were not observed with DCA treatment.
4.4.6.3.3.3 Tumor genotype: H-ras mutation frequency and spectrum
An approach to determine the potential MO As of DCA and TCA through examination of
the types of tumors each "induced" or "selected" was to examine H-ras activation (Ferreira-
Gonzalez et al., 1995; Anna et al., 1994; Bull et al., 2002; Nelson et al., 1990). No data of this
type were available for CH. This approach has also been used to try to establish an H-ras
activation pattern for "genotoxic" and "non-genotoxic" liver carcinogens compounds and to
make inferences concerning peroxisome proliferator-induced liver tumors. However, as noted
by Stanley et al. (1994), the genetic background of the mice used and the dose of carcinogen may
affect the number of activated H-ras containing tumors which develop. In addition, the stage of
progression of "lesions" (i.e., foci vs. adenomas vs. carcinomas) also has been linked the
observance of H-ras mutations. Fox et al. (1990) note that tumors induced by phenobarbital
(0.05% drinking H20, 1 yr), chloroform (200 mg/kg corn oil gavage, 2 times weekly for 1 year)
or ciprofibrate (0.0125% diet, 2 years) had a much lower frequency of H-ras gene activation than
those that arose spontaneously (2-year bioassays of control animals) or induced with the
"genotoxic" carcinogen benzidine-2 HC1 (120 ppm, drinking H20, 1 yr) in mice. In that study,
the term "tumor" was not specifically defined but a correlation between the incidence of H-ras
gene activation and development of either a hepatocellular adenoma or hepatocellular carcinoma
was reported to be made with no statistically significant difference between the frequency of H-
ras gene activation in the hepatocellular adenomas and carcinomas. Histopathological
examination of the spontaneous tumors, tumors induced with benzidine-2HCL, Phenobarbital,
and chloroform was not reported to reveal any significant changes in morphology or staining
characteristics. Spontaneous tumors were reported to have 64% point mutation in codon 61 (n =
50 tumors examined) with a similar response for Benzidine of 59% (n = 22 tumors examined),
whereas for Phenobarbital the mutation rate was 7% (n= 15 tumors examined), chloroform 21%
(n = 24 tumors examined) and ciprofibrate 21 % (n = 39 tumors examined). The ciprofibrate-
induced tumors were reported to be more eosinophilic as were the surrounding normal
hepatocytes.
Hegi et al. (1993) tested ciprofibrate-induced tumors in the NIH3T3 cotransfection-nude
mouse tumorigenicity assay, which the authors state is capable of detecting a variety of activated
protooncogenes. The tumors examined (ciprofibrate-induced or spontaneously arising) were
6/22/2009
560
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
taken from the Fox et al. study (1990), screened previously, and found to be negative for H-ras
activation. With the limited number of samples examined, Hegi et al concluded that ras
protooncogene activation or activation of other protooncogenes using the nude mouse assay were
not frequent events in ciprofibrate-induced tumors and that spontaneous tumors were not
promoted with it. Using the more sensitive methods, the H-ras activation rate was reported to be
raised from 21 to 31% for ciprofibrate-induced tumors and from 64 to 66% for spontaneous
tumors. Stanley et al. (1994) studied the effect of methylclofenapate (MCP) (25 mg/kg for up to
2 years), a peroxisome proliferator, in B6C3F1 (relatively sensitive) and C57BL/10J (relatively
resistant) mice for H-ras codon 61 point mutations in MCP-induced liver tumors (hepatocellular
adenomas and carcinomas). In the B6C3F1 mice the number of tumors with codon 61 mutations
was 11/46 and for C57BL/10J mice 4/31. Unlike the findings of Fox et al. (1990), Stanley et al.
(1994) reported an increase in the frequency of mutation in carcinomas, which was reported to be
twice that of adenomas in both strains of mice, indicating that stage of progression was related to
the number of mutations in those tumors, although most tumors induced by MCP did not have
this mutation.
Anna et al. (1994) reported that the H-ras codon 61 mutation frequency was not
statistically different in liver tumors from DCA and TCE-treated mice from a highly variable
number of tumors examined. From their concurrent controls they reported that H-ras codon 61
mutations in 17% (n = 6) of adenomas and 100% (n = 5) of carcinomas. For historical controls
(published and unpublished) they reported mutations in 73% (n = 33) of adenomas and mutations
in 70%) (n = 30) of carcinomas. For tumors from TCE treated animals they reported mutations in
35%) (n = 40) of adenomas and 69%> (n = 36) of carcinomas, while for DCA treated animals they
reported mutations in 54%> (n = 24) of adenomas and in 68%> (n = 40) of carcinomas. Anna et al.
(1994) reported more mutations in TCE-induced carcinomas than adenomas. In regard to
mutation spectra in H-ras oncogenes in control or spontaneous tumors, the patterns were slightly
different but those from TCE treatment were mostly similar to that of DCA-induced tumors
(0.5%o in drinking water).
The study of Ferreira-Gonzalez (1995) in male B6C3 F1 mice has the advantage of
comparison of tumor phenotype at the same stage of progression (hepatocellular carcinoma), for
allowance of the full expression of a tumor response (i.e. 104 weeks), and an adequate number of
spontaneous control lesions for comparison with DCA or TCA treatments. However, tumor
phenotype at an end stage of tumor progression may not be indicative of earlier stages of the
disease process. In spontaneous liver carcinomas, 58%> were reported to show mutations in H-61
as compared with 50%> of tumor from 3.5 g/L DCA-treated mice and 45%> of tumors from
4.5.g/L TCA-treated mice. A number of peroxisome proliferators have been reported to have a
much smaller mutation frequency that spontaneous tumors [e.g.,. 13-24%) H-ras codon 61
6/22/2009
561
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
mutations after methylclofenopate depending on mouser strain, Stanely et al. (1994): 21 to 31%
for ciprofibrate-induced tumors and from 64 to 66% for spontaneous tumors, Fox et al. (1990)
and Hegi et al (1993)]. Thus, there was a heterogeneous response for this phenotypic marker for
the spontaneous, DCA-, and TCA- treatment induced hepatocellular carcinomas had similar
patterns H-ras mutations that differed from the reduced H-ras mutation frequencies reported for a
number of peroxisome proliferators.
In his review, Bull (2000) suggested "the report by Anna et al. (1994) indicated that
TCE-induced tumors possessed a different mutation spectra in codon 61 of the H-ras oncogene
than those observed in spontaneous tumors of control mice." Bull (2000) stated that "results of
this type have been interpreted as suggesting that a chemical is acting by a mutagenic
mechanism" but went on to suggest that it is not possible to a priori rule out a role for selection
in this process and that differences in mutation frequency and spectra in this gene provide some
insight into the relative contribution of different metabolites to TCE-induced liver tumors. Bull
(2000) noted that data from Anna et al. (1994), Ferreira-Gonzalez et al. (1995), and Maronpot et
al. (1995) indicated that mutation frequency in DCA-induced tumors did not differ significantly
from that observed in spontaneous tumors. Bull (2000) also noted that the mutation spectra
found in DCA-induced tumors has a striking similarity to that observed in TCE-induced tumors,
and DCA-induced tumors were significantly different than that of TCA-induced liver tumors.
Bull et al. (2002) reported that mutation frequency spectra for the H-ras codon 61 in
mouse liver "tumors" induced by TCE (n = 37 tumors examined) were reported to be
significantly different than that for TCA (n = 41 tumors examined), with DCA-treated mice
tumors giving an intermediate result (n = 64 tumors examined). In this experiment, TCA-
induced "tumors" were reported to have more mutations in codon 61 (44%) than those from TCE
(21%) and DCA (33%). This frequency of mutation in the H-ras codon 61 for TCA is the
opposite pattern as that observed for a number of peroxisome proliferators in which the number
of mutations at H-ras 61 in tumors has been reported to be much lower than spontaneously
arising tumors (see above). Bull et al. (2002) noted that the mutation frequency for all TCE,
TCA or DCA tumors was lower in this experiment than for spontaneous tumors reported in other
studies (they had too few spontaneous tumors to analyze in this study), but that this study utilized
lower doses and was of shorter duration than that of Ferreira-Gonzalez (1995). Furthermore, the
disparities from previous studies may also be impacted by lesion grouping, mentioned above, in
which lower stages of progression are grouped with more advanced stages.
Overall, in terms of H-ras mutation, TCE-induced tumors appears to be more like DCA-
induced tumors (which are consistent with spontaneous tumors), or those resulting from a
coexposure to both DCA and TCA (Bull et al., 2002), than from those induced by TCA. As
noted above, Bull et al. (2002) reported the mutation frequency spectra for the H-ras codon 61 in
6/22/2009
562
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
mouse liver tumors induced by TCE to be significantly different than that for TCA, with DCA-
treated mice tumors giving an intermediate result and for TCA-induced tumors to have a H-ras
profile that is the opposite than those of a number of other peroxisome proliferators. More
importantly, however, these data, along with the measures discussed above, show that mouse
liver tumors induced by TCE are heterogeneous in phenotype and genotype in a manner similar
to that observed in spontaneous tumors.
4.4.6.3.4 "Stop" experiments
Several stop experiments, in which treatment is terminated early in some dose groups,
have attempted to ascertain the whether progression differences exist between TCA and DCA.
After 37 weeks of treatment and then a cessation of exposure for 15 weeks, Bull et al. (1990)
reported that after combined 52 week period, liver weight and % liver/body weight were reported
to still be statistically significantly elevated after DCA or TCA treatment. The authors partially
attribute the remaining increases in liver weight to the continued presence of hyperplastic
nodules in the liver. In terms of liver tumor induction, the authors stated that "statistical analysis
of tumor incidence employed a general linear model ANOVA with contrasts for linearity and
deviations from linearity to determine if results from groups in which treatments were
discontinued after 37 weeks were lower than would have been predicted by the total dose
consumed." The multiplicity of tumors (incidence was not used) observed in male mice exposed
to DCA or TCA at 37 weeks and then sacrificed at 52 weeks were compared with those exposed
for a full 52 weeks. The response in animals that received the shorter duration of DCA exposure
was very close to that which would be predicted from the total dose consumed by these animals.
By contrast, the response to TCA exposure for the shorter duration was reported by the authors
to deviate significantly (P = 0.022) from the linear model predicted by the total dose consumed.
However, in the prediction of "dose-response," foci, adenomas, and carcinomas were combined
into one measure. Therefore foci, a certain percentage of which have been commonly shown to
spontaneously regress with time, were included in the calculation of total "lesions." Moreover,
only a sample of lesions were selected for histological examination, and as is evident in the
sample, some lesions appeared "normal" upon microscopic examination (see below). Therefore,
while suggesting that cessation of exposure diminished the number of "lesions," methodological
limitations temper any conclusions regarding the identity and progression of lesion with
continuous vs. non-continuous DCA and TCA treatment.
Additionally, Bull et al. (1990) noted that after stopping treatment, DCA lesions appeared
to arrest their progression in contrast to TCA lesions, which appeared to progress. In particular,
among those in the stop treatment group (at 2 g/L) with 0/19 lesions examined histologically
were carcinomas, while in the continuous treatment groups, a significant fraction of lesions
examined were carcinomas at the higher exposure (6/23 at 2 g/L). By contrast, at terminal
6/22/2009
563
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
sacrifice, TCA lesions a larger fraction of the lesions examined were carcinomas in the stop
treatment group (3/5 at 2 g/L) than in the continuous treatment group (2/7 and 4/16 at 1 g/L and
2 g/L, respectively).
However, as mentioned above, these inferences are based on examination of only a
subset of lesions. Specifically, for TCA treatment the number of animals examined for
determination of which "lesions" were foci, adenomas, and carcinomas was 11 out of the 19
mice with "lesions" at 52 weeks while all 4 mice with lesions after 37 weeks of exposure and 15
weeks of cessation were examined. For DCA treatment the number of animals examined was
only 10 out of 23 mice with "lesions" at 52 weeks while all 7 mice with lesions after 37 weeks of
exposure and 15 weeks of cessation were examined. Most importantly, when lesions were
examined microscopically, some did not all turn out to be preneoplastic or neoplastic - for
example, two lesions appeared "to be histologically normal" and one necrotic.
While limited, the conclusions of Bull et al. (1990) are consistent with later experiments
performed by Pereira and Phelps (1996). They noted that in MNU-treated mice that were then
treated with DCA, the yield of altered hepatocytes decreases as the tumor yields increase
between 31 and 51 weeks of exposure suggesting progression of foci to adenomas, but that
adenomas did not appear to progress to carcinomas. For TCA, Pereira and Phelps (1996)
reported that "MNU-initiated" adenomas promoted with TCA continued to progress. However,
the use of MNU initiation complicates direct comparisons with treatment with TCA or DCA
alone.
No similar data comparing stop and continued treatment of TCE are available to assess
the consistency or lack-thereof with TCA or DCA. Moreover, the informative of such a
comparison would be limited by designs of the available TCA and DCA studies, which have
used higher concentrations in conjunction with the much lower durations of exposure. While
higher doses allow for responses to be more easily detected, it introduces uncertainty as to the
effects of the higher doses alone. In addition, because the overall duration of the experiments is
also generally much less than 104 weeks, it is not possible to discern whether the differences in
results between those animals in which treatment was suspended in comparison to those in which
had not had been conducted would persist with longer durations.
4.4.6.4 Conclusions regarding the role of TCA, DCA, and CH in TCE-induced effects in the
liver
In summary, it is likely that oxidative metabolism is necessary for TCE-induced effects in
the liver. However, the specific metabolite or metabolites responsible for both non-cancer and
cancer effects is less clear. TCE, TCA, and DCA exposures have all been associated with
induction of peroxisomal enzymes but are all weak PPARa agonists. The available data strongly
6/22/2009
564
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
support TCA not being the sole or predominant active moiety for TCE-induced liver effects.
With respect to hepatomegaly, TCE and TCA dose-response relationships are quantitatively
inconsistent, for TCE leads to greater increases in liver/body weight ratios that expected from
predicted rates of TCA production. In fact, above a certain dose of TCE, liver/body weight
ratios are greater than that observed under any conditions studied so far for TCA. Histological
changes and effects on DNA synthesis are generally consistent with contributions from either
TCA or DCA, with a degree of polyploidization, rather than cell proliferation, likely to be
significant for TCE, TCA, and DCA. With respect to liver tumor induction, TCE leads to a
heterogeneous population of tumors, not unlike those that occur spontaneously or that are
observed following TCA-, DCA-, or CH-treatment. Moreover, some liver phenotype
experiments, particularly those utilizing immunostaining for c-Jun, support a role for both DCA
and TCA in TCE-induced tumors, with strong evidence that TCA cannot solely account for the
characteristics of TCE-induced tumors. In addition, H-ras mutation frequency and spectrum of
TCE-induced tumors more closely resembles that of spontaneous tumors or of those induced by
DCA, and were less similar in comparison to that of TCA-induced tumors. The heterogeneity of
TCE-induced tumors is similar to that observed to be induced by a broad category of
carcinogens, and to that observed in human liver cancer. Overall, then, it is likely that multiple
TCE metabolites, and therefore multiple pathways, contribute to TCE-induced liver tumors.
4.4.7 MOA for TCE Liver Carcinogenicity
4.4.7.1 Mutagenicity
The hypothesis is that TCE acts by a mutagenic mode of action in TCE-induced
hepatocarcinogenesis. According to this hypothesis, the key events leading to TCE-induced liver
tumor formation constitute the following: TCE oxidative metabolite CH, after being produced in
the liver, cause direct alterations to DNA (e.g., mutation, DNA damage, and/or micronuclei
induction). Mutagenicity is a well established cause of carcinogenicity.
Experimental Support for the Hypothesized Mode of Action
The genotoxicity, as described by the ability of TCE, CH, TCA, and DCA to induce
mutations, was discussed previously in Section 4.1. The strongest data for mutagenic potential is
for CH, thought to be a relatively short-lived intermediate in the metabolism of TCE that is
rapidly converted to TCA and TCOH in the liver (see Section 3.3). CH causes a variety of
genotoxic effects in available in vitro and in vivo assays, with particularly strong data as to its
ability to induce aneuploidy. It has been argued that CH mutagenicity is unlikely to be the cause
of TCE carcinogenicity because the concentrations required to elicit these responses are
generally quite high, several orders of magnitude higher that achieved in vivo (Moore and
6/22/2009
565
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Harrington-Brock, 2000). For example, peak concentrations of CH in the liver of around 2-3
mg/kg have been reported after TCE administration at doses that are hepatocarcinogenic in
chronic bioassays (Abbas and Fisher, 1997; Greenberg et al., 1999). Assuming a liver density of
about 1 kg/L, these concentrations are orders of magnitude less than the minimum concentrations
reported to elicit genotoxic responses in the Ames test and various in vitro measures of
micronucleus, aneuploidy, and chromosome aberrations, which are in the 100-1,000 mg/L range.
However, it is not clear how much of a correspondence is to be expected from concentrations in
genotoxicity assays in vitro and concentrations in vivo, as reported in vivo CH concentrations are
in whole-liver homogenate while in vitro concentrations are in culture media. In addition, a few
in vitro studies have reported positive results at concentrations as low as 1 or 10 mg/L, including
Furnus et al. (1990) for aneuploidy in Chinese hamster CHED cells (10 mg/L), Eichenlaub-Ritter
et al. (1996) for bivalent chromosomes in meiosis I in MF1 mouse oocytes (10 mg/L), and
Gibson et al. (1995) for cell transformation in Syrian hamster embryo cells after 7 day treatment.
Moreover, some in vivo genotoxicity assays of CH reported positive results at doses similar to
those eliciting a carcinogenic response in chronic bioassays. For example, Nelson and Bull
(1988) reported increased DNA single strand breaks at 100 CH mg/kg (oral) in male B6C3F1
mice, although the result was not replicated by Chang et al. (1992). In another example, four of
six in vivo mouse genotoxicity studies reported that CH induced micronuclei in mouse bone-
marrow erythrocytes, with the lowest effective doses in positive studies ranging from 83 to 500
mg/kg (positive: Russo and Levis [1992], Russo et al. [1992], Marrazini et al. [1994], Beland et
al. [1999]; negative: Leuschner and Leuschner [1991], Leopardi et al. [1993]). However, the use
of ip administration in these and many other in vivo genotoxicity assays complicates the
comparison with carcinogenicity data. Also, it is difficult with the available data to assess the
contributions from the genotoxic effects of CH along with those from the genotoxic and non-
genotoxic effects of other oxidative metabolites (discussed below in Section 4.4.5.2 and 4.4.5.3).
Furthermore, altered DNA methylation, another heritable mechanism by which gene
expression may be altered, is discussed below in the in Section 4.4.1.3.2.6. As discussed
previously, the differential patterns of H-ras mutations observed in liver tumors induced by TCE,
TCA, and DCA may be more indicative of tumor selection and tumor progression resulting from
exposure to these agents rather than a particular mechanism of tumor induction. The state of the
science of cancer and the role of epigenetic changes, in addition to genetic changes, in the
initiation and progression of cancer and specifically liver cancer, are discussed in Section 3.1 of
Appendix E.
Therefore, while data are insufficient to conclude that a mutagenic MOA mediated by CH
is operant, a mutagenic MOA, mediated either by CH or by some other oxidative metabolite of
TCE, cannot be ruled out.
6/22/2009
566
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.4.7.IPPARa receptor activation
The hypothesis is that TCE acts by a PPARa agonism MOA in TCE-induced
hepatocarcinogenesis. According to this hypothesis, the key events leading to TCE-induced liver
tumor formation constitute the following: the TCE oxidative metabolite TCA, after being
produced in the liver, activates the PPARa receptor, which then causes alterations in cell
proliferation and apoptosis and clonal expansion of initiated cells. This MOA is assumed to
apply only to the liver.
Experimental Support for the Hypothesized Mode of Action
Proliferation of peroxisomes and increased activity of a number of related marker
enzymes has been observed in rodents treated with TCE, TCA, and DCA. The peroxisome-
related effects of TCE are most likely mediated primarily through TCA based on TCE
metabolism producing more TCA than DCA and the lower doses of TCA required to elicit a
response relative to DCA. However, Bull (2004) and Bull et al. (2004) have recently suggested
that peroxisome proliferation occurs at higher exposure levels than those that induce liver tumors
for TCE and its metabolites. They report that a direct comparison in the no-effect level or low-
effect level for induction of liver tumors in the mouse and several other endpoints shows that, for
TCA, liver tumors occur at lower concentrations than peroxisome proliferation in vivo but that
PPARa activation occurs at a lower dose than either tumor formation or peroxisome
proliferation. A similar comparison for DCA shows that liver tumor formation occurs at a much
lower exposure level than peroxisome proliferation or PPARa activation. In vitro transactivation
studies have shown that human and murine versions of PPARa are activated by TCA and DCA,
while TCE itself is relatively inactive in the in vitro system, at least with mouse PPARa
(Maloney and Waxman, 1999; Zhou and Waxman, 1998). In addition, Laughter et al. (2004)
reported that the responses of ACO, PCO, and CYP4A induction by TCE, TCA, and DCA were
substantially diminished in PPARa-null mice. Therefore, evidence suggests that TCE, through
its metabolites TCA and DCA, activate PPARa, and that at doses relevant to TCE-induced
hepatocarcinogenesis, the role of TCA in PPARa agonism is likely to predominate.
It has been suggested that PPARa receptor activation is both the MOA for TCA liver
tumor induction as well as the MOA for TCE liver tumor induction, as a result of the metabolism
of TCE to TCA (NRC, 2006). Section 3.4 in Appendix E addressed the status of the PPARa
MOA hypothesis for liver tumor induction and provides a more detailed discussion. However, as
discussed previously and in Section 2.1.10 of Appendix E, TCE-induced increases in liver
weight have been reported in male and female mice that do not have a functional PPARa
receptor (Nakajima et al., 2000). The dose-response for TCE-induced liver weight increases
differs from that of TCA (see Section 2.4.2 of Appendix E). The phenotype of the tumors
induced by TCE have been described to differ from those by TCA and to be more like those
6/22/2009
567
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
occurring spontaneously in mice, those induced by DCA, or those resulting from a combination
of exposures to both DCA and TCA (see Section 2.4.4 of Appendix E). As to whether TCA
induces tumors through activation of the PPARa receptor, the tumor phenotype of TC A-induced
mouse liver tumors has been reported to have a different pattern of H-ras mutation frequency
from other peroxisome proliferators (see Section 2.4.4.of Appendix E; Bull et al., 2002; Stanely
et al., 1994; Fox et al., 1990; Hegi et al., 1993). While TCE, DCA, and TCA are weak
peroxisome proliferators, liver weight induction from exposure to these agents has not correlated
with increases in peroxisomal enzyme activity (e.g. PCO activity) or changes in peroxisomal
number or volume. By contrast, as discussed above, liver weight induction from subchronic
exposures appears to be a more accurate predictor of carcinogenic response for DCA, TCA and
TCE in mice (see also Section 2.4.4 of Appendix E). The database for cancer induction in rats is
much more limited than that of mice for determination of a carcinogenic response to these
chemicals in the liver and the nature of such a response.
While many compounds known to cause rodent liver tumors with long-term treatment
also activate the nuclear receptor peroxisome proliferator activated receptor alpha (PPARa), the
mechanisms by which PPARa activation contributes to tumorigenesis are not completely known
(Klaunig et al., 2003; NRC, 2006; Yang et al., 2007). As reviewed by Keshava and Caldwell
(2006), PPARa activation leads to a highly pleiotropic response and may play a role in toxicity
in multiple organs as well as in multiple chronic conditions besides cancer (obesity,
atherosclerosis, diabetes, inflammation). Klaunig et al. (2003) and NRC (NRC, 2006) proposed
that the key causal events for PPARa agonist-induced liver carcinogenesis, after PPARa
activation, are perturbation of cell proliferation and/or apoptosis, mediated by gene expression
changes, and selective clonal expansion. It has also been proposed that sufficient evidence for
this MOA consists of evidence of PPARa agonism (i.e., in a receptor assay) in combination with
either light- or electron-microscopic evidence for peroxisome proliferation or both increased
liver weight and one more of the in vivo markers of peroxisome proliferation (Klaunig et al.,
2003). However, it should be noted that peroxisome proliferation and in vivo markers such as
PCO are not considered causal events (Klaunig et al., 2003; NRC, 2006), and that their
correlation with carcinogenic potency is poor (Marsman et al., 1988). Therefore, for the
purposes of this discussion, peroxisome proliferation and its markers are considered indicators of
PPARa activation, as it is well established that these highly specific effects are mediated through
PPARa (Klaunig et al., 2003; Peters et al., 1997).
As recently reviewed by Guyton et al. (2009), recent data suggest that PPARa activation
along with these hypothesized causal events may not be sufficient for carcinogenesis. In
particular, Yang et al. (2007) reported comparisons between mice treated with Wy-14643 and
transgenic mice in which PPARa was constitutively activated in hepatocytes without the
6/22/2009
568
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
presence of ligand. Yang et al. (2007) reported that, in contrast to Wy-14643-treatment, the
transgene did not induce liver tumors at 11 months, despite inducing PPARa-mediated effects of
a similar type and magnitude seen in response to tumorigenic doses of Wy-14643 in wild type
mice (decreased serum fatty acids, induction of PPARa target genes, altered expression of cell-
cycle control genes, and a sustained increase in cellular proliferation). Nonetheless, it is
important to discuss the extent to which PPARa activation mediates the effects proposed by
Klaunig et al. (2003) and NRC (2006), even if they may not be themselves sufficient for
carcinogenesis, and investigation continues into additional events that may also contribute, such
as non-parenchymal cell activation and micro-RNA-based regulation of protooncogenes (Yang et
al., 2007; Shah et al., 2007). Specifically addressed below are gene expression changes,
proliferation, clonal expansion, and mutation frequency or spectrum.
With respect to gene expression changes due to TCE, Laughter et al. (2004) evaluated
transcript profiles induced by TCE in wild-type and PPARa-null mice. As noted in Appendix E
sections 3.4.1.3. and 3.1.2., there are limitations to the interpretation of such studies, some of
which are discussed below. Also noted in Appendix E are discussions of how studies of
peroxisome proliferators, indicate of the need for phenotypic anchoring, especially since gene
expression is highly variable between studies and within studies using the same experimental
paradigm. Section 3.4 in Appendix E also provides detailed discussions of the status of the
PPARa hypothesis. Of note, all null mice at the highest TCE dose (1,500 mg/kg-d) were
moribund prior to the end of the planned 3 week experiment(Laughter et al., 2004), and it was
proposed that this may reflect a greater sensitivity in PPARa-null mice to hepatotoxins due to
defects in tissue repair abilities. Laughter et al. (2004) also noted that four genes known to be
regulated by other peroxisome proliferators also had altered expression with TCE treatment in
wild-type, but not null mice. However, in a comparative analysis, Bartosiewicz et al. (2001)
concluded that TCE induced a different pattern of transcription than two other peroxisome
proliferators, DEHP and clofibrate. In addition, Keshava and Caldwell (2006) compared gene
expression data from Wy-14643, DBP, GEM, and DEHP, and noted a lack of consistent results
across PPARa agonists. Thus, available data are insufficient to conclude that TCE gene
expression changes are similar to other PPAR agonists, or even that there are consistent changes
(beyond the in vivo markers of peroxisome proliferation, such as ACO, PCO, CYP4A, etc.)
among different agonists. It should also be noted that Laughter et al. (2004) did not compare
baseline (i.e., control levels of) gene expression between null and wild-type control mice,
hindering interpretation of these results (Keshava and Caldwell, 2006). The possible relationship
between PPARa activation and hypomethylation are discussed below in section 4.4.7.1.9.
In terms of proliferation, mitosis itself has not been examined in PPARa-null mice, but
BrdU incorporation, a measure of DNA synthesis that may reflect cell division, polyploidization,
6/22/2009
569
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
or DNA repair, was observed to be diminished in null mice as compared to wild-type mice at 500
and 1,000 mg/kg-d TCE (Laughter et al., 2004). However, BrdU incorporation in null mice was
still about 3-fold higher than controls, although it was not statistically significantly different due
to the small number of animals, high variability, and the 2 to 3-fold higher baseline levels of
BrdU incorporation in control null mice as compared to control wild-type mice. Therefore,
while PPARa appears to contribute to the short-term increase in DNA synthesis observed with
TCE treatment, these results cannot rule out other contributing mechanisms. However, since it is
likely that both cellular proliferation and increased ploidy contribute to the observed TCE-
induced increases in DNA synthesis, it is not clear to whether the observed decrease in BrdU
incorporation is due to reduced proliferation, reduced polyploidization, or both.
With respect to clonal expansion, it has been suggested that tumor characteristics such as
tincture (i.e., the staining characteristics light microscopy sections of tumor using H&E stains)
and oncogene mutation status can be used to associate chemical carcinogens with a particular
MOA such as PPARa agonism (Klaunig et al., 2003; NRC, 2006). This approach is problematic
primarily because of the lack of specificity of these measures. For example, with respect to
tincture, it has been suggested that TCA-induced foci and tumors resemble those of other
peroxisome proliferators in basophilia and lack of expression of GGT and GST-pi. However, as
discussed in Caldwell and Keshava (2006), the term "basophilic" in describing foci and tumors
can be misleading, because, for example, multiple lineages of foci and tumors exhibit basophilia,
including those not associated with peroxisome proliferators (Bannasch, 1996; Bannasch et al.,
2001; Carter et al., 2003). Moreover, a number of studies indicate that foci and tumors induced
by other "classic" peroxisome proliferators may have different phenotypic characteristics from
that attributed to the class through studies of WY-14643, including DEHP (Voss et al., 2005) and
clofibric acid (Michel et al., 2007). Furthermore, even the combination of GGT and GSTpi
negative, basophilic foci are non-specific to peroxisome proliferators, as they have been
observed in rats treated with AfBl and AfBl plus PB, none of which are peroxisome
proliferators (Kraupp-Grasl et al., 1998; Grasl-Kraupp et al., 1993). Finally, while Bull et al.
(2004) suggested that negative expression of c-jun in TCA-induced tumors may be consistent
with a characteristic phenotype of peroxisome proliferators, no data could be located to support
this statement. Therefore, of phenotypic information does not appear to be reliable for
associating a chemical with a PPARa agonism MOA.
Mutation frequency or spectrum in oncogenes has also been suggested to be an indicator
of a PPARa agonism MOA being active (NRC, 2006), with the idea being that specific
genotypes are being promoted by PPARa agonists. Although not a highly specific marker, H-ras
codon 61 mutation frequency and spectra data do not support a similarity between mutations in
TCE-induced, TCA-, or DCA- tumors and those due to other peroxisome proliferators. For
6/22/2009
570
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
example, while ciprofibrate and methylclofenopate had lower mutation frequencies than
historical controls (Hegi et al., 1993; Stanley et al., 1994), TCA-induced tumors had mutation
frequencies similar to or higher than historical controls (Ferreira-Gonzalez et al., 1995; Bull et
al., 2002). Anna et al. (1994) and Ferreira-Gonzalez et al. (1995) also reported TCE and DCA-
induced tumors to have mutation frequencies similar to historical controls, although Bull et al.
(2002) reported lower frequencies for these chemicals. However, the data reported by Bull et al.
(2002) consist of mixed lesions at different stages of progression, and such differing stages, in
addition to differences in genetic background and dose, can influence the frequency of H-ras
mutations (Stanley et al., 1994). In addition, a greater frequency of mutations was reported in
carcinomas than adenomas, and Bull et al. (2002) stated that this suggested that H-ras mutations
were a late event. Moreover, Fox et al. (1990) noted that tumors induced by phenobarbital,
chloroform, and ciprofibrate all had a much lower frequency of H-ras gene activation than those
that arose spontaneously, so this marker does not have good specificity. Mutation spectrum is
similarly of low utility for supporting a PPARa agonism MOA. First, because many peroxisome
proliferators been reported to have low frequency of mutations, the comparison of mutation
spectrum would be limited to a small fraction tumors. In addition to the low power due to small
numbers, the mutation spectrum is relatively non-specific, as Fox et al. (1990) reported that of
the tumors with mutations, the spectra of the peroxisome proliferator ciprofibrate, historical
controls, and the genotoxic carcinogen benzidine-2 HC1 were similar.
In summary, TCE clearly activates PPARa, and some of the effects contributing to
tumorigenesis that Klaunig et al. (2003) and NRC (2006) propose to be the result of PPARa
agonism are observed with TCE, TCA, or DCA treatment. While this consistency is supportive a
role for PPARa, all of the proposed key causal effects with the exception of PPARa agonism
itself are non-specific, and may be caused by multiple mechanisms. There is more direct
evidence that several of these effects, including alterations in gene expression and changes in
DNA synthesis, are mediated by multiple mechanisms in the case of TCE, and a causal linkage
to PPARa specifically is lacking. Therefore, because, as discussed further in the MOA
discussion below, there are multiple lines of evidence supporting the role of multiple pathways
of TCE-induced tumorigenesis, the hypothesis that PPARa agonism and the key causal events
proposed by Klaunig et al. (2003) and NRC (2006) constitute the sole or predominant MOA for
TCE-induced carcinogenesis is considered unlikely.
Furthermore, as reviewed by Guyton et al. (2009), recent data strongly suggest that
PPARa and key events hypothesized by Klaunig et al. (2003) are not sufficient for
carcinogenesis induced by the purported prototypical agonist Wy-14643. Therefore, the
proposed PPARa MOA is likely "incomplete" in the sense that the sequence of key events
necessary for cancer induction has not been identified. A recent two-year bioassay of the
6/22/2009
571
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
peroxisome proliferator DEHP showed that it can induce a liver tumor response in mice lacking
PPARa similar to that in wild-type mice (Ito et al., 2007). Klaunig et al. (2003) previously
concluded that PPARa agonism was the sole MOA for DEHP-induced liver tumorigenesis based
on the lack of tumors in PPARa-null mice after 11 months treatment with Wy-14643 (Peters et
al., 1997). They also assumed that due to the lack of markers of PPARa agonism in PPARa-null
mice after short-term treatment with DEHP (Ward et al., 1998), a long-term study of DEHP in
PPARa-null mice would yield the same results as for Wy-14643. However, due the finding by
Ito et al. (2007) that PPARa-null mice exposed to DEHP do develop liver tumors, they
concluded that DEHP can induce liver tumors by multiple mechanisms (Ito et al., 2007;
Takashima et al., 2008). Hence, since there is no 2-year bioassay in PPARa-null mice exposed
to TCE or its metabolites, it is not justifiable to use a similar argument based on Peters et al.
(1997) and short-term experiments to suggest that the PPARa MOA is operative. Therefore, the
conclusion is supported that the hypothesized PPARa MOA is inadequately specified because
the data do not adequately show the proposed key events individually being required for
hepatocarcinogenesis, nor do they show the sequence of key events collectively to be sufficient
for hepatocarcinogenesis.
4 A.1.3 Additional Proposed Hypotheses and Key Events with Limited Evidence or
Inadequate Experimental Support
Several effects that been hypothesized to be associated with liver cancer induction are
discussed in more detail below, including increased liver weight, DNA hypomethylation, and
pathways involved in glycogen accumulation such as insulin signaling proteins. As discussed
above, TCE and its metabolites reportedly increase nuclear size and ploidy in hepatocytes, and
these effects likely account for much of the increases in labeling index and DNA synthesis
caused by TCE. Importantly, these changes appear to persist with cessation of treatment, with
liver weights, but not nuclear sizes, returning to control levels(Kjellstrand et al., 1983a). In
addition, glycogen deposition, DNA synthesis, increases in mitosis, or peroxisomal enzyme
activity do not appear correlated with TCE-induced liver weight changes.
4.4.7.3.1 Increased liver weight
Increased liver weight or liver/body weight ratios (hepatomegaly) is associated with
increased risk of liver tumors in rodents, but it is relatively non-specific (Allen et al., 2004). The
evidence presented above for TCE and its metabolites suggest a similarity in dose-response
between liver weight increases at short-term durations of exposure and liver tumor induction
observed from chronic exposure. Liver weight increases may results from several concurrent
processes that have been associated with increase cancer risk (e.g., hyperplasia, increased ploidy,
6/22/2009
572
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
and glycogen accumulation) and when observed after chronic exposure may result from the
increased presence of foci and tumors themselves. Therefore, there are inadequate data to
adequately define a MOA hypothesis for hepatocarcinogenesis based on liver weight increases.
4.4.7.3.2 "Negative selection"
As discussed above, TCE, TCA, and DCA all cause transient increases in DNA synthesis.
This DNA synthesis has been assumed to result from proliferation of hepatocytes. However, the
dose-related TCA- and DCA-induced increases in liver weight not correlate with patterns of
DNA synthesis; moreover, there have been reports that DNA synthesis in individual hepatocytes
does not correlate with whole liver DNA synthesis measures (Sanchez and Bull, 1990; Carter et
al., 1995). With continued treatment, decreases in DNA synthesis have been reported for DCA
(Carter et al., 1995). More importantly, several studies show that transient DNA synthesis is
confined to a very small population of cells in the liver in mice exposed to TCE for 10 days or to
DCA or TCA for up to 14 days of exposure. Therefore, generalized mitogenic stimulation is not
likely to play a role in TCE-induced liver carcinogenesis.
Bull has proposed that the TCE metabolites TCA and DCA may contribute to liver tumor
induction through so-called "negative selection" by way of several possible processes (Bull,
2000). First, it is hypothesized that the mitogenic stimulation by continued TCA and DCA
exposure is down-regulated in normal hepatocytes, conferring a growth advantage to initiated
cells that either do not exhibit the down-regulation of response or are resistant to the down-
regulating signals. This is implausible as both the normal rates of cell division in the liver and
the TCE-stimulated increases are very low. Polyploidization has been reported to decrease the
normal rates of cell division even further. That the transient and relatively low level of DNA
synthesis reported for TCE, DCA, and TCA is reflective of proliferation rather than
polyploidization is not supported by data on mitosis. A mechanism for such "down-regulation"
has not been identified experimentally.
A second proposed contributor to "negative-selection" is direct enhancement by TCA and
DCA in the growth of certain populations of initiated cells. While differences in phenotype of
end stage tumors have been reported between DCA and TCA, the role of selection and
emergence of potentially different foci has not been elucidated. Neither have pathway
perturbations been identified that are common to liver cancer in human and rodent for TCE,
DCA, and TCA. The selective growth of clones of hepatocytes that may progress fully to cancer
is a general feature of cancer and not specific to at TCE, TCA, or DCA MOA.
A third proposed mechanism by which TCE may enhance liver carcinogenesis within this
"negative selection" paradigm is through changing apoptosis. However as stated above, TCE
has been reported to either not change apoptosis or to cause a slight increase at high doses.
Rather than increases in apoptosis, peroxisome proliferators have been suggested to inhibit
6/22/2009
573
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
apoptosis as part of their carcinogenic MOA. However, the age and species studied appear to
greatly affect background rates of apoptosis (Snyder et al., 1995) with the rat having a greater
rate of apoptosis than the mouse. DCA has been reported to induce decreases in apoptosis in the
mouse (Carter et al., 1995; Snyder et al., 1995). However, the significance of the DCA-induced
reduction in apoptosis, from a level that is already inherently low in the mouse, for the MOA for
induction of DCA-induce liver cancer is difficult to discern.
Therefore, for a MOA for hepatocarcinogenesis based on "negative selection," there are
inadequate data to adequately define the MOA hypothesis, or the available data do not support
such a MOA being operative.
4.4.7.3.3 Polyploidization
Polyploidization may be an important key event in tumor induction. For example, in
addition to TCE, partial hepatectomy, nafenopin, methylclofenopate, DEHP, DEN, N-
nitrosomorpholine, and various other exposures that contribute to liver tumor induction also shift
the hepatocyte ploidy distribution to be increasingly diploid or polypoid (Hasmal and Roberts,
2000; Styles et al., 1988; Melchiorri et al., 1993; Miller et al., 1996; Vickers et al., 1996). As
discussed by Gupta (2000), "[wjorking models indicate that extensive polyploidy could lead to
organ failure, as well as to oncogenesis with activation of precancerous cell clones." However,
the mechanism(s) by which increased polypoidy enhances carcinogenesis is not currently
understood. Due to increased DNA content, polypoid cells will generally have increased gene
expression. However, polyploid cells are considered more highly differentiated and generally
divide more slowly and are more likely to undergo apoptosis, perhaps thereby indirectly
conferring a growth advantage to initiated cells (See Section 1 of Appendix E). Of note is that
changes in ploidy have been observed in transgenic mouse models that are also prone to develop
liver cancer (See Section 3.3.1 of Appendix E). It is likely that polyploidization occurs with
TCE exposure and it is biologically plausible that polyploidization can contribute to liver
carcinogenesis, although the mechanism(s) is (are) not known. However, whether
polyploidization is necessary for TCE-induced carcinogenesis is not known, as no experiment in
which polyploidization specifically is blocked or diminished has been performed and the extent
of polyploidization has not been quantified. Therefore, there are inadequate data to adequately
define a MOA hypothesis for hepatocarcinogenesis based on polyploidization.
4.4.7.3.4 Glycogen storage
As discussed above, several studies have reported that DCA causes accumulation of
glycogen in mouse hepatocytes. Such glycogen accumulation has been suggested to be
pathogenic, as it is resistant to mobilization by fasting (Kato-Weinstein et al., 1998). In humans,
glycogenesis due to glycogen storage disease or poorly controlled diabetes has been associated
6/22/2009
574
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
with increased risk of liver cancer (LaVecchia et al., 1994; Adami et al., 1996; Wideroff et al.,
1997; Rake et al., 2002). Glycogen accumulation has also been reported to occur in rats exposed
to DC A.
For TCE exposure in mice or rats, glycogen content of hepatocytes has been reported to
be somewhat less than or the same as controls, or not remarked upon in the studies. TCA
exposure has been reported to decrease glycogen content in rodent hepatocytes while DCA has
been reported to increase it (Kato-Weinstein et al., 2001). There is also evidence that DCA-
induced increases in glycogen accumulation are not proportional to liver weight increases and
only account for a relatively small portion of increases in liver mass. DCA-induced increases in
liver weight are not a function of cellular proliferation but probably include hypertrophy
associated with polyploidization, increased glycogen deposition and other factors.
While not accounting for increases in liver weight, excess glycogen can still be not only
be pathogenic but a predisposing condition for hepatocarcinogenesis. Some hypotheses
regarding the possible relationship between glycogenesis and carcinogenesis have been posed
that lend them biological plausibility. Evert et al. (2003), using an animal model of hepatocyte
exposure to a local hyperinsulinemia from transplanted islets of Langerhans with remaining
tissue is hypoinsulinemic, reported that insulin induces alterations resembling preneoplastic foci
of altered hepatocytes (FAH) that develop into hepatocellular tumors in later stages of
carcinogenesis. Lingohr et al. (2001) suggest that normal hepatocytes down-regulate insulin-
signaling proteins in response to the accumulation of liver glycogen caused by DCA and that the
initiated cell population, which does not accumulate glycogen and is promoted by DCA
treatment, responds differently from normal hepatocytes to the insulin-like effects of DCA. Bull
et al. (Bull et al., 2002) reported increased insulin receptor protein expression in tumor tissues
regardless of whether they were induced by TCE, TCA, or DCA. Given the greater activity of
DCA relative to TCA on carbohydrate metabolism, it is unclear whether changes in these
pathways are causes or simply reflect the effects of tumor progression. Therefore, it is
biologically plausible that changes in glycogen status may occur from the opposing actions of
TCE metabolites, but changes in glycogen content due to TCE exposure has not been
quantitatively studied. The possible contribution of these effects to TCE-induced
hepatocarcinogenesis is unclear. Therefore, there are inadequate data to adequately define a
MOA hypothesis for TCE-induced hepatocarcinogenesis based on changes in glycogen storage
or even data to support increased glycogen storage to result from TCE exposure.
4.4.7.3.5 Inactivation of GST-Zeta
DCA has been shown to inhibit its own metabolism in that pre-treatment in rodents prior
to a subsequent challenge dose leads to a longer biological half-life (Schultz et al., 2002). This
self-inhibition is hypothesized to occur through inactivation of GST-zeta (Schultz et al., 2002).
6/22/2009
575
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
In addition, TCE has been shown to cause the same prolongation of DC A half-life in rodents,
suggesting that TCE inhibits GST-zeta, probably through the formation of DCA (Schultz et al.,
2002). DCA-induced inhibition of GST-zeta has also been reported in humans, with GST-zeta
polymorphisms reported to influence the degree of inactivation (Blackburn et al., 2000;
Blackburn et al., 2001; Tzeng et al., 2000). Board et al. (2001) report one variant to have
significantly higher activity with DCA as a substrate than other GST zeta isoforms, which could
affect DCA susceptibility.
GST-zeta, which is identical to maleylacetoacetate isomerase, is part of the tyrosine
catabolism pathway which is disrupted in type 1 hereditary tyrosinemia, a disease associated
with the development of hepatocellular carcinoma at a young age (Tanguay et al., 1996). In
particular, GST-zeta metabolizes maleylacetoacetate (MAA) to fumarylacetoacetate (FAA) and
maleylacetone (MA) to fumarylacetone(Cornett et al., 1999; Tanguay et al., 1996). It has been
suggested that the increased cancer risk with this disease, as well as through DCA exposure,
results from accumulation of MAA and MA, both alkylating agents, or FAA, which displays
apoptogenic, mutagenic, aneugenic, and mitogenic activities (Bergeron et al., 2003; Cornett et
al., 1999; Jorquera and Tanguay, 2001; Kim et al., 2000; Tanguay et al., 1996). However, the
possible effects of DCA through this pathway will depend on whether MAA, MA, or FAA is the
greater risk factor, since inhibition of GST-zeta will lead to greater concentrations of MAA and
MA and lower concentrations of FAA. Therefore, if MAA is the more active agent, DCA may
increase carcinogenic risk, while if FAA is the more active, DCA may decrease carcinogenic
risk. Tzeng et al. (2000) propose the later based on the greater genotoxicity of FAA, and in fact
suggest that DCA may "merit consideration for trial in the clinical management of hereditary
tyrosinemia type 1."
Therefore, TCE-induced inactivation GST-zeta, probably through formation of DCA,
may play a role in TCE-induced hepatocarcinogenesis. However, this mode of action is not
sufficiently delineated at this point for further evaluation, as even the question of whether its
actions through this pathway may increase or decrease cancer risk has yet to be experimentally
tested.
4.4.7.3.6 Oxidative stress
Several studies have attempted to study the possible effects of "oxidative stress" and
DNA damage resulting from TCE exposures. The effects of induction of metabolism by TCE, as
well as through co-exposure to ethanol, have been hypothesized to in itself increase levels of
"oxidative stress" as a common effect for both exposures (see Section 4.2.4. of Appendix E). In
terms of contributing to a carcinogenic MO A, the term "oxidative stress" is a somewhat non-
specific term, as it is implicated as part of the pathophysiologic events in a multitude of disease
processes and is part of the normal physiologic function of the cell and cell signaling.
6/22/2009
576
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Commonly, it appears to refer to the formation of reactive oxygen species leading to cellular or
DNA damage. As discussed above, however, measures of oxidative stress induced by TCE,
TCA, and DCA appear to be either not apparent, or at the very most transient and non-persistent
with continued treatment (Larson and Bull, 1992; Channel et al., 1998; Toraason et al., 1999;
Parrish et al., 1996). Therefore, while the available data are limited, there is insufficient
evidence to support a role for such effects in TCE-induced liver carcinogenesis.
Oxidative stress has been hypothesized to be part of the MOA for peroxisome
proliferators, but has been found to neither be correlated with cell proliferation nor carcinogenic
potency of peroxisome proliferators (see Section 3.4.1.1 of Appendix E). For instance, Parrish et
al. (1996) reported that increases in PCO activity noted for DCA and TCA were not associated
with 8-OHdG levels (which were unchanged) and also not with changes laurate hydrolase
activity observed after either DCA or TCA exposure. The authors concluded that their data does
not support an increase in steady state oxidative damage to be associated with TCA initiation of
cancer and that extension of treatment to time periods sufficient to insure peroxisome
proliferation failed to elevate 8-OHdG in hepatic DNA. The authors thus suggested that
peroxisome proliferative properties of TCA were not linked to oxidative stress or carcinogenic
response.
4.4.7.3.7 Changes in gene expression (e.g., hypomethylation)
Studies of gene expression as well as considerations for interpretation of studies of using
the emerging technologies of DNA, siRNA, and miRNA microarrays for MOA analyses are
included in Sections 3.1.2. and 3.4.2.2. of Appendix E. Caldwell and Keshava (2006) and
Keshava and Caldwell (2006) report on both genetic expression studies and studies of changes in
methylation status induced by TCE and its metabolites as well as differences and difficulties in
the patterns of gene expression between differing PPARa agonists. In particular are concerns for
the interpretation of studies which employ pooling of data as well as interpretation of "snapshots
in time of multiple gene changes." For instance, in the Laughter et al. (2004) study, it is not
clear whether transcription arrays were performed on pooled data as well as the issue of
phenotypic anchoring as data on % liver/body weight indicates significant variability within TCE
treatment groups, especially in PPARa-null mice. For studies of gene expression using
microarrays Bartosiewicz et al. (2001) used a screening analysis of 148 genes for xenobiotic-
metabolizing enzymes, DNA repair enzymes, heat shock proteins, cytokines, and housekeeping
gene expression patterns in the liver in response TCE. The TCE-induced gene induction was
reported to be highly selective; only Hsp 25 and 86 and Cyp2a were up-regulated at the highest
dose tested. Collier et al. (2003) reported differentially expressed mRNA transcripts in
embryonic hearts from Sprague-Dawley rats exposed to TCE with sequences down-regulated
with TCE exposure appearing to be those associated with cellular housekeeping, cell adhesion,
6/22/2009
577
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
and developmental processes. TCE was reported to induce up-regulated expression of numerous
stress-response and homeostatic genes.
For the Laughter et al. (2004) study, transcription profiles using macroarrays containing
approximately 1,200 genes were reported in response to TCE exposure with 43 genes reported to
be significantly altered in the TCE-treated wild-type mice and 67 genes significantly altered in
the TCE-treated PPARa knockout mice. However, the interpretation of this information is
difficult because in general, PPARa knockout mice have been reported to be more sensitive to a
number of hepatotoxins partly because of defects in the ability to effectively repair tissue damage
in the liver (Shankar et al., 2003; Mehendale, 2000) and because a comparison of gene
expression profiles between controls (wild-type and PPARa knockout) were not reported. As
reported by Voss et al. (2006), dose-, time course-, species-, and strain-related differences should
be considered in interpreting gene array data. The comparison of differing PPARa agonists
presented in Keshava and Caldwell (2006) illustrate the pleiotropic and varying liver responses
of the PPARa receptor to various agonists, but did not imply that these responses were
responsible for carcinogenesis.
As discussed above in Section 3.3.5 of Appendix E, Aberrant DNA methylation is a
common hallmark of all types of cancers, with hypermethylation of the promoter region of
specific tumor suppressor genes and DNA repair genes leading to their silencing (an effect
similar to their mutation) and genome-wide hypomethylation (Ballestar and Esteller, 2002;
Berger and Daxenbichler, 2002; Herman et al., 1998; Pereira et al., 2004; Rhee et al., 2002).
Whether DNA methylation is a consequence or cause of cancer is a long-standing issue
(Ballestar and Esteller, 2002). Fraga et al. (2004, 2005) reported global loss of monoacetylation
and trimethylation of histone H4 as a common hallmark of human tumor cells; they suggested,
however, that genomewide loss of 5-methylcytosine (associated with the acquisition of a
transformed phenotype) exists not as a static predefined value throughout the process of
carcinogenesis but rather as a dynamic parameter (i.e., decreases are seen early and become more
marked in later stages).
DNA methylation is a naturally occurring epigenetic mechanism for modulating gene
expression, and disruption of this mechanism is known to be relevant to human carcinogenesis.
As reviewed by Calvisi et al. (2007), "[ajberrant DNA methylation occurs commonly in human
cancers in the forms of genome-wide hypomethylation and regional hypermethylation. Global
DNA hypomethylation (also known as demethylation) is associated with activation of
protooncogenes, such as c-Jun, c-Myc, and c-HA-Ras, and generation of genomic instability.
Hypermethylation on CpG islands located in the promoter regions of tumor suppressor genes
results in transcriptional silencing and genomic instability." While clearly associated with
cancer, it has not been conclusively established whether these epigenetic changes play a
6/22/2009
578
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
causative role or are merely a consequence of transformation (Tryndyak et al., 2006). However,
as Calvisi et al. (2007) note, "Current evidence suggests that hypomethylation might promote
malignant transformation via multiple mechanisms, including chromosome instability, activation
of protooncogenes, reactivation of transposable elements, and loss of imprinting."
Although little is known about how it occurs, a hypothesis has also been proposed that
that the toxicity of TCE and its metabolites may arise from its effects on DNA methylation
status. In regard to methylation studies, many are coexposure studies as they have been
conducted in initiated animals with some studies being very limited in their reporting and
conduct. Caldwell and Keshava (2006) review the body of work regarding TCE, DCA and TCA.
Methionine status has been noted to affect the emergence of liver tumors (Counts et al., 1996).
Tao et al. (2000) and Pereira et al. (2004) have studied the effects of excess methionine in the
diet to see if it has the opposite effects as a deficiency (i.e. and reduction in a carcinogenic
response rather than enhancement). However, Tao et al. (2000) report that the administration of
excess methionine in the diet is not without effect and can result in % liver/body weight ratios.
Pereira et al. (2004) report that methionine treatment alone at the 8 g/kg level was reported to
increase liver weight, decrease lauryl-CoA activity and to increase DNA methylation.
Pereira et al. (2004) reported that very high level of methionine supplementation to an
AIN-760A diet, affected the number of foci and adenomas after 44 weeks of co-exposure to
3.2.g/L DCA. However, while the highest concentration of methionine (8.0 g/kg) was reported
to decrease both the number of DCA-induce foci and adenomas, the lower level of methionine
co-exposure (4.0 g./kg) increased the incidence of foci. Co-exposure of methionine (4.0 or 8.0
g/kg) with 3.2 g/L DCA was reported to decrease by ~ 25% DCA-induced glycogen
accumulation, increase mortality, but not to have much of an effect on peroxisome enzyme
activity (which was not elevated by more than 33% over control for DCA exposure alone). The
authors suggested that their data indicate that methioninine treatment slowed the progression of
foci to tumors. Given that increasing hypomethylation is associated with tumor progression,
decreased hypomethylation from large doses of methionine are consistent with a slowing of
progression. Whether, these results would be similar for lower concentrations of DCA and lower
concentrations of methionine that were administered to mice for longer durations of exposure,
cannot be ascertained from this data. It is possible that in a longer-term study, the number of
tumors would be similar. Finally, a decrease in tumor progression by methionine
supplementation is not shown to be a specific event for the MOA for DCA-induced liver
carcinogenicity.
Tao et al. (2000) reported that 7 days of gavage dosing of TCE (1,000 mg/kg in corn oil),
TCA (500 mg/kg, neutralized aqueous solution), and DCA (500 mg/kg, neutralized aqueous
solution) in 8-week old female B6C3F1 mice resulted in not only increased liver weight but also
6/22/2009
579
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
increased hypomethylation of the promoter regions of c-Jun and c-Myc genes in whole liver
DNA. However, data were shown for 1-2 mice per treatment. Treatment with methionine was
reported to abrogate this response only at a 300 mg/kg i.p dose with 0-100 mg/kg doses of
methionine having no effect. Ge et al. (2001) reported DC A- and TCA-induced DNA
hypomethylation and cell proliferation in the liver of female mice at 500 mg/kg and decreased
methylation of the c-myc promoter region in liver, kidney and urinary bladder. However,
increased cell proliferation preceded hypomethylation. Ge et al. (2002) also reported
hypomethylation of the c-myc gene in the liver after exposure to the peroxisome proliferators
2,4-dichlorophenoxyacetic acid (2,4-D)(l,680 ppm), dibutyl phthalate (DBP) (20,000 ppm),
gemfibrozil (8,000 ppm), and Wy-14,643 (50-500 ppm, with no effect at 5 or 10 ppm) after six
days in the diet. Caldwell and Keshava (2006) concluded that hypomethylation did not appear to
be a chemical-specific effect at these concentrations. As noted Section 3.3.5 of Appendix E,
chemical exposure to a number of differing carcinogens have been reported to lead to
progressive loss of DNA methylation..
After initiation by N-methyl-N-nitrosourea (25 mg/kg) and exposure to 20 mmol/L DCA
or TCA (46 weeks), Tao et al. (2004) report similar hypomethylation of total mouse liver DNA
by DCA and TCA with tumor DNA showing greater hypomethylation. A similar effect was
noted for region-2 (DMR-2) of the insulin-like growth factor-II (IGF-II) gene. The authors
suggest that hypomethylation of total liver DNA and the IGF-II gene found in non-tumorous
liver tissue would appear to be the result of a more prolonged activity and not cell proliferation,
while hypomethylation of tumors could be an intrinsic property of the tumors. As pointed out by
Caldwell and Keshava (2006) over expression of IGF-II gene in liver tumors and preneoplastic
foci has been shown in both animal models of hepatocarcinogenesis and humans, and may
enhance tumor growth, acting via the over-expressed IGF-I receptor (Scharf et al., 2001; Werner
and Le Roith, 2000).
Diminished hypomethylation was observed in Wy-14643-treated PPARa-null mice as
compared to wild type mice, suggestive of involvement of PPARa in mediating hypomethylation
(Pogribny et al., 2007), but it is unclear how relevant these results are to TCE and its metabolites.
First, the doses of Wy-14643 administered are associated with substantial liver necrosis and
mortality with long-term treatment (Woods et al., 2007), adding confounding factors the
interpretation of their results. Hypomethylation by Wy-14643 progressively increased with time
up to 5 months (Pogribny et al., 2007), consistent with the sustained DNA synthesis caused by
Wy-14643 and a role for proliferation in causing hypomethylation. Regardless, as discussed
above, it is unlikely that PPARa is the mediator of the observed transient increase in DNA
synthesis by DCA, so even if it is important for hypomethylation by TCA, there may be more
than one pathway for this effect.
6/22/2009
580
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
To summarize, aberrant DNA methylation status, including hypomethylation, is clearly
associated with both human and rodent carcinogenesis. Hypomethylation itself appears to be
sufficient for carcinogenesis, as diets deficient in choline and methionine that induce
hypomethylation have been shown to cause liver tumors in both rats and mice (Ghoshal and
Farber, 1984; Mikol et al., 1983; Henning and Swendseid, 1996; Wainfan and Poirier, 1992).
However, it is not known to what extent hypomethylation is necessary for TCE-induced
carcinogenesis. However, as noted by Bull (2004) and Bull et al. (2004), the doses of TCA and
DCA that have been tested for induction of hypomethylation are quite high compared to doses at
which tumor induction occurs - at least 500 mg/kg/day. Whether these effects are still manifest
at lower doses relevant to TCE carcinogenicity, particularly with respect to DCA, has not been
investigated. Finally, the role of PPARa in modulating hypomethylation, possibly through
increased DNA synthesis as suggested by experiments with Wy-14643, are unknown for TCE
and its metabolites.
4.4.7.3.8 Cytotoxicity
Cytotoxicity and subsequent induction of reparative hyperplasia have been proposed as
key events for a number of chlorinated solvents, such as chloroform and carbon tetrachloride..
However, as discussed above and discussed by Bull (2004) and Bull et al. (2004), TCE treatment
at doses relevant to liver carcinogenicity results in relatively low cytotoxicity. While a number
of histological changes with TCE exposure are observed, in most cases necrosis is minimal or
mild, associated with vehicle effects, and with relatively low prevalence. This is consistent with
the low prevalence of necrosis observed with TCA and DCA treatment at doses relevant to TCE
exposure. Therefore, it is unlikely that cytotoxicity and reparative hyperplasia play a significant
role in TCE carcinogenicity
4A.1AMOA Conclusions
Overall, although a role for many of the proposed key events discussed above cannot be
ruled out, there are inadequate data to support the conclusion that any of the particular MOA
hypotheses reviewed above are operant. Thus, the MOA of liver tumors induced by TCE is
considered unknown at this time, and the answer to the first key question "2. Is the hypothesized
mode of action sufficiently supported in the test animals?" is "no" at this time. Consequently,
the other key questions of "2. Is the hypothesized mode of action relevant to humans?" and "3.
Which populations or lifestages can be particularly susceptible to the hypothesized mode of
action?" will not be discussed in a MOA-specific manner. Rather, they are discussed below in
more general terms, first qualitatively and then quantitatively, using available relevant data.
6/22/2009
581
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.4.7.4.1 Qualitative human relevance and susceptibility
No data exist that suggests that TCE-induced liver tumorigenesis is caused by processes
that irrelevant in humans. In addition, as discussed above, several of the other effects such as
polyploidization, changes in glycogen storage, and inhibition of GST-zeta - are either clearly
related to human carcinogenesis or areas of active research as to their potential roles. For
example, the effects of DCA on glycogen storage parallel the observation that individuals with
conditions that lead to glycogenesis appear to be at an increased risk of liver cancer (LaVecchia
et al., 1994; Adami et al., 1996; Wideroff et al., 1997; Rake et al., 2002). In addition, there may
be some relationship between the effects of DCA and the mechanism of increased liver tumor
risk in childhood in those with type 1 hereditary tyrosinemia, though the hypotheses needs to be
tested experimentally. Similarly, with respect to PPARa activation and downstream events
hypothesized to be causally related to liver carcinogenesis, it is generally acknowledged that "a
point in the rat/mouse key events cascade where the pathway is biologically precluded in humans
cannot be identified, in principle" (Klaunig et al, 2003; NRC, 2006).
In terms of human relevance and susceptibility, it is also useful to briefly review what is
known about human HCC. A number of risk factors have been identified for human
hepatocellular carcinoma, including ethanol consumption, hepatitis B and C virus infection,
aflatoxin B1 exposure, and, more recently, diabetes and perhaps obesity (El-Serag and Rudolph,
2007). However, it is also estimated that a substantial minority of HCC patients, perhaps 15% to
50%, have no established risk factors (El-Serag and Rudolph, 2007). In addition, cirrhosis is
present in a large proportion of HCC patients, but the prevalence of HCC without underlying
cirrhosis, while not precisely known, is still significant, with estimates based on relatively small
samples ranging from 7% to 54% (Fattovisch, 2004).
However, despite the identification of numerous factors that appear to play a role in the
human risk of HCC, the mechanisms are still largely unclear (Yeh et al., 2007). Interestingly,
the observation by Leakey et al. (2003a, b) that body weight significantly and strongly impacts
background liver tumor rates in B6C3F1 mice parallels the observed epidemiologic associations
between liver cancer and obesity (review in El-Serag and Rudolph [2007]). This concordance
suggests that similar pathways may be involved in spontaneous liver tumor induction between
mice and humans. The extent to which TCE exposure may interact with known risk factors for
HCC cannot be determined at this point, but several hypotheses can be posed based on existing
data. If TCE affects some of the same pathways involved in human HCC, as suggested in the
discussion of several TCE-induced effects above, then TCE exposure may lead a risk that is
additive to background.
As discussed above, there are several parallels between the possible key events in TCE-
induced liver tumors in mice and what is known about mechanisms of human HCC, though none
6/22/2009
582
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
have been experimentally tested. Altered ploidy distribution and DNA hypomethylation are
commonly observed in human HCC (Zeppa et al., 1998; Lin et al., 2003; Calvisi et al., 2007).
Interestingly, El-Serag and Rudolph (2007) have been suggested that the risk of HCC increases
with cirrhosis in part because the liver parenchymal cells have decreased proliferative capacity,
resulting in an altered milieu that promotes tumor cell proliferation. This description suggests a
similarity in mode of action, though via different mechanisms, with the "negative selection"
hypothesis proposed by Bull (2000) for TCE and its metabolites although for TCE changes in
apoptosis and cell proliferation have not been noted or examined to such an extent to provide
evidence of a similar environment. Increased ploidy decreases proliferative capacity, so that
may be another mechanism through which the effects of TCE mimic the conditions thought to
facilitate the induction of human HCC.
In sum, from the perspective of hazard characterization, the available data support the
conclusion that the mode of action for TCE-induced mouse liver tumors is relevant to humans.
No data suggest that any of the key events are biologically precluded in humans, and a number of
qualitative parallels exist between hypotheses for the mode of action in mice and what is known
about the etiology and induction of human HCC. A number of risk factors have been identified
that appear to modulate the risk of human HCC, and these may also modulate the susceptibility
to the effects from TCE exposure. As noted in Section 4. of Appendix E, TCE exposure in the
human population is accompanied not only by external exposures to its metabolites, but
brominated analogues of those metabolites that are also rodent carcinogens, a number of
chlorinate solvents that are hepatocarcinogenic and alcohol consumption. The types of tumors
and the heterogeneity of tumors induced by TCE in rodents parallel those observed in humans
(see Section 3.1.8 of Appendix E). The pathways identified for induction of cancer in humans
for cancer are similar to those for the induction of liver cancer (see Section 3.2.1. of Appendix
E). However, while risk factors have been identified for human liver cancer that have
similarities to TCE-induced effects and those of its metabolites, both the mechanism for human
liver cancer induction and that for TCE-induced liver carcinogenesis in rodents are not known.
4.4.7.4.2 Quantitative species differences
As a precursor to the discussion of quantitative differences between humans and rodents
and among humans, it should be noted that an adequate explanation for the difference in
response for TCE-liver cancer induction between rats and mice has yet to be established or for
that difference to be adequately described given the limitations in the rat database. For TCA,
there is only one available long-term study in rats that, while suggestive that TCA is less potent
in rats than mice, is insufficient to determine if there was a TCA-induced effect or what its
magnitude may be. While some have proposed that the lower rate of TCA formation in rats
relative to mice would explain the species difference, PBPK modeling suggests that the
6/22/2009
583
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
differences (3-5 fold) may be inadequate to fully explain the differences in carcinogenic
potency. Moreover, inferences from comparing the effects of TCE and TCA on liver weight,
using PBPK model-based estimates of TCA produced from TCE, indicate that TCA is not likely
to play a predominant role in hepatomegaly. Combined with the qualitative correlation between
rodent hepatomegaly and hepatocarcinogenesis observed across many chemicals, this suggests
that TCA similarly is not a predominant factor in TCE-induced hepatocarcinogenesis. Indeed,
there are multiple lines of evidence that TCA is insufficient to account for TCE-induced tumors,
including data on tumor phenotype (e.g., c-Jun immunostaining) and genotype (e.g., H-ras
mutation frequency and spectrum). For DCA, only a single experiment in rats is available
(reported in two publications), and although it suggests lower hepatocarcinogenic potency in rats
relative to mice, its relatively low power limits the inferences that can be made as to species
differences.
As TCA induces peroxisome proliferation in the mouse and the rat, some have suggested
that difference in peroxisomal enzyme induction is responsible for the difference in susceptibility
to TCA liver carcinogenesis. The study of DeAngelo et al. (1989) has been cited in the literature
as providing evidence of differences between rats and mice for peroxisomal response to TCA.
However, data from the most resistant strain of rat (Sprague-Dawley) has been cited in
comparisons of peroxisomal enzyme effects but the Osborne-Mendel and F344 rat were not
refractory and showed increased PCO activity so it is not correct to state that the rat is refractory
to TCA-induction of peroxisome activity (see Section 2.3.1.5 of Appendix E). In addition, as
discussed above, inferences based on PCO activity are limited by its high variability, even in
control animals, as well as its not necessarily being predictive of the peroxisome number or
cytoplasmic volume.
The same assumption of lower species sensitivity by measuring peroxisome proliferation
has been applied to humans, as peroxisome proliferation caused by therapeutic PPARa agonists
such as fibrates in humans is generally lower (<2-fold induction) than that observed in rodents
(20- to 50-fold induction). However, as mentioned above, it is known that peroxisome
proliferation is not a good predictor of potency (Marsman et al., 1988).
Limited data exist on the relative sensitivity of the occurrence of key events for liver
tumor induction between mice and humans and among humans. Pharmacokinetic differences are
addressed with PBPK modeling to the extent that data allow, so the discussion here will
concentrate on pharmacodynamic differences. Most striking is the difference in "background"
rates of liver tumors. Data from NTP indicates that control B6C3F1 mice in 2-year bioassays
have a background incidence of hepatocellular carcinomas of 26% in males and 10% in females,
with higher incidences for combined hepatocellular adenomas and carcinomas (Maronpot, 2007).
However, as discussed above, Leakey et al. (2003a, b) report that the background incidence rates
6/22/2009
584
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
are very dependent on the weight of the mice. By contrast, the estimated lifetime risk of liver
and biliary tract cancer in the United States (about 75% of which are hepatocellular carcinomas)
is 0.97% for men and 0.43% for women (Ries et al., 2008). However, regions of the world
where additional risk factors (hepatitis infection, alflatoxin exposure) have high prevalence have
liver cancer incidences up to more than 6-fold greater than the United States (Ferlay et al., 2004).
Therefore, one possible quantitative difference that can be flagged for use in dose-response
assessment is the background rate of liver tumors between species. Biologically-based dose-
response modeling by Chen (2000) suggested that the data were consistent with a purely
promotional model in which potency would be proportional to background tumor incidence.
However, it is notable that male Swiss mice, which have lower background liver tumor rates than
the B6C3F1 strain, were also positive in one long-term bioassay (Maltoni et al., 1986).
Similarly, in terms of intra-species susceptibility, to the extent that TCE may
independently promote pre-existing initiated cells, it can be hypothesized that those with greater
risk for developing HCC due to one more of the known risk factors would have a proportional
increase in the any contributions from TCE exposure. In addition, in both humans and mice,
males appear to be at increased risk of liver cancer, possibly due to sexually dimorphism in
inflammatory responses (Lawrence et al., 2007; Naugler et al., 2007; Rakoff-Nahoun and
Medzhitov, 2007), suggesting that men may also be more susceptible to TCE-induced liver
tumorigenesis than women. It has been observed that human HCC is highly heterogeneous
histologically, but within patients and between patients, studies are only beginning to distinguish
the different pathways that may be responsible for this heterogeneity (Feitelson et al., 2002;
Chen et al., 2002; Yeh et al., 2007).
Appropriate quantitative data is generally lacking on inter-species differences in the
occurrence of most other proposed key events, although many have argued that there are
significant quantitative differences between rodents and humans related to PPARa activation
(Klaunig et al., 2003; NRC, 2006). For instance, it has been suggested that lower levels of
PPARa receptor in human hepatocytes relative to rodent hepatocytes contributes to lower human
sensitivity (Tugwood et al., 1996; Palmer et al., 1998; Klaunig et al., 2003). However, out of a
small sample of human livers (n = 6) show similar protein levels to mice (Walgren et al., 2000a).
Another proposed species difference has been ligand affinity, but while transactivation assays
showed greater affinity of Wy-14643 and PFOA for rodent relative to human PPARa, they
showed TCA and DCA had a similar affinities between species (Maloney and Waxman, 1999).
Furthermore, it is not clear that receptor-ligand kinetics (capacity and affinity) are rate-limiting
for eliciting hepatocarcinogenic effects, as it is known that maximal receptor occupation is not
necessary for a maximal receptor mediated response (Stephenson, 1956, see also review by
Danhof et al., 2007).
6/22/2009
585
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
There is also limited in vivo and in vitro data suggesting that increases in cell
proliferation mediated by PPARa agonists are diminished in humans and other primates relative
to rodents (Klaunig et al., 2003; NRC, 2006; Hoivik et al., 2004). However, Walgren et al.
(2000b) reported that TCA and DCA were not mitogenic in either human or rodent hepatocytes
in vitro. Furthermore, TCE, TCA, and DCA all induce only transient increases in cell
proliferation, so the relevance to TCE of inter-species differences from PPARa agonists that to
produce sustained proliferation, such as Wy-14643, is not clear. In addition, comparisons
between primate and rodent models should take into account the differences in the ability to
respond to any mitogenic stimulation, (see Section 3.2 of Appendix E). Primate and human
liver respond differently (and much more slowly) to a stimulus such as partial hepatectomy.
Recent studies in "humanized" mice (PPARa-null mice in which a human PPARa gene
was subsequently inserted and expressed in the liver) reported that treatment with a PPARa
agonist lead to greatly lower incidence of liver tumors as compared to wild-type mice (Morimura
et al., 2006). However, these experiments were performed with WY-14643 at a dose causing
systemic toxicity (reduced growth and survival), had a duration of less than one year, and
involved a limited number of animals. In addition, because liver tumors in mice at less than one
year are extremely rare, the finding a one adenoma in WY-14643-treated humanized mice
suggests carcinogenic potential that could be further realized with continued treatment (Keshava
and Caldwell, 2006). In addition, Yang et al. (2007) recently noted that let-7C, a microRNA
involved in cell growth and thought to be a regulatory target of PPARa (Shah, 2008), was
inhibited by Wy-14643 in wild-type mice, but not in "humanized mice" in which had human
PPARa was expressed throughout the body on a PPARa-null background. However, these
humanized mice had about a 20-fold higher baseline expression of let-7C, as reported in control
mice, potentially masking any treatment effects. More generally, it is not known to what extent
PPARa-related events are rate limiting in TCE-induced liver tumorigenesis, for which multiple
pathways appear to be operative. So even if quantitative differences mediated by PPARa were
well estimated, they would not be directly usable for dose-response assessment in the absence of
way to integrate the contributions from the different pathways.
In sum, the only quantitative data and inter- and intra-species susceptibility suitable for
consideration in dose-response assessment are differences background liver tumor risk. These
may modulate the effects of TCE if relative risk, rather than additional risk, is the appropriate
common inter- and intra-species metric. However, the extent to which relative risk would
provide a more accurate estimate of human risk is unknown.
6/22/2009
586
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
References
Abbas R, Fisher JW. (1997). physiologically based pharmacokinetic model for trichloroethylene
and its metabolites, chloral hydrate, trichloroacetate, dichloroacetate, trichloroethanol,
and trichloroethanol glucuronide inB6C3Fl mice. Toxicol ApplPharmacol. 147:15-30.
Adami HO, Chow WH, Nyrene O, Berne C, Linet MS, Ekbom A, et al. 1996. Excess risk of
primary liver cancer in patients with diabetes mellitus. J Natl Cancer Inst 88:1472-1477.
Allen DG, Pearse G, Haseman JK, Maronpot RR. 2004. Prediction of rodent carcinogenesis: an
evaluation of prechronic liver lesions as forecasters of liver tumors in NTP
carcinogenicity studies. Toxicol Pathol 32: 393-401.
Alexander DD, Kelsh MA, Mink PJ, Mandel JH, Basu R, Weingart M. 2007. A meta-analysis
of occupational trichloroethylene exposure and liver cancer. Int Arch Occup Environ
Health 81(2): 127-43.
Anna CH, Maronpot RR, Pereira MA, Foley JF, Malarkey DE, Anderson MW. 1994. ras proto-
oncogene activation in dichloroacetic-, trichloroethylene- and tetrachloroethylene-
induced liver tumors in B6C3F1 mice. Carcinogenesis 15:2255-2261.
Anttila A, Pukkala E, Sallmen M, Hernberg S, Hemminki K. 1995. Cancer incidence among
Finnish workers exposed to halogenated hydrocarbons. J Occup Environ Med
37:797-806.
ATSDR(Agency for Toxic Substances and Disease Registry). 2003. Health Consultation.
View-Master Factory Supply Well (a.k.a. Mattel Portland Operations) Beaverton,
Washington County, Oregon. Prepared by Oregon Department of Human Services under
Cooperative Agreement with the Agency for Toxic Substances and Disease Registry.
October 20, 2003.
ATSDR. (2004) Feasibility Investigation of Worker Exposure to Trichloroethylene at the View-
Master Factory in Beaverton, Oregon. Prepared by the Oregon Department of Human
Services. Report No. PB2005-102173.
ATSDR. (2006) Health Consultation: Cancer and Birth Outcome Analysis, Endicott Area, Town
of Union, Broome County, New York. U.S. Department of Health and Human Services,
May 2006, Atlanta, GA.
6/22/2009
587
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Austin H, Delzell E, Grufferman S, Levine R, Morrison AS, Stolley PD, Cole P. 1987. Case-
control study of hepatocellular carcinoma, occupation, and chemical exposures. J Occup
Med 29:665-669.
Austin EW, Parrish JM, Kinder DH, Bull RJ. 1996. Lipid peroxidation and formation of 8-
hydroxydeoxygaunosine from acute doses of halogenated acetic acids. Fundamental and
Applied Toxicology 31: 77-82.
Axelson O, Selden A, Andersson K, Hogstedt C. 1994. Updated and expanded Swedish cohort
study on trichloroethylene and cancer risk. J Occup Environ 36:556-562.
Bai CL, CanfieldPJ, StaceyNH. 1992a. Effects of hexachloro-1,3-butadiene and 1,1,2,2-
tetrachloroethylene on individual serum bile acids. Toxicol Ind Health 8: 191-203.
Bai CL, Canfield PJ, Stacey NH. 1992b. Individual serum bile acids as early indicators of
carbon tetrachloride- and chloroform-induced liver injury. Toxicology 75:221-234.
Ballestar E, Esteller M. 2002. The impact of chromatin in human cancer: linking DNA
methylation to gene silencing. Carcinogenesis 23:1103-1109
Bannasch P. 2001. Hormonal and hormone-like effects eliciting hepatocarcinogenesis. Folia
Histochem Cytobio 139(Suppl 2):28~29.
Bannasch P, Nehrbass D, Kopp-Schneider A. 2001 Significance of hepatic preneoplasia for
cancer chemoprevention. IARC Sci Publ 154:223-240.
Barton HA, Bull R, Schultz I, Andersen ME. (1999). Dichloroacetate (DCA) dosimetry:
interpreting DCA-induced liver cancer dose response and the potential for DCA to
contribute to trichloroethylene-induced liver cancer. Toxicol Lett. 106:9-21.
Bartosiewicz M, Penn S, Buckpitt A. 2001. Applications of gene arrays in environmental
toxicology: fingerprints of gene regulation associated with cadmium chloride,
benzo(a)pyrene, and trichloroethylene. Environ Health Perspect 109:71-74.
Beland, F.A. (1999) NTP technical report on the toxicity and metabolism studies of chloral
hydrate (CAS No. 302-17-0). Administered by gavage to F344/N rats and B6C3F1 mice.
Toxic. Rep. Ser., 59, 1-66, A1-E7
Berger J, Daxenbichler G. 2002. DNA methylation of nuclear receptor genes—possible role in
malignancy. J Steroid Biochem Mol Bio 180:1-11.
Bergeron, A; Jorquera, R; Tanguay, RM. (2003) [Hereditary tyrosinemia: an endoplasmic
reticulum stress disorder?]. Med Sci (Paris) 19:976-980.
6/22/2009
588
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Berman, E; Schlicht, M; Moser, VC; et al. (1995) A Multidisciplinary Approach to
Toxicological Screening: I. Systemic Toxicity. Journal of Toxicology and Environmental
Health 45:127-143.
Blackburn, AC; Tzeng, HF; Anders, MW; et al. (2000) Discovery of a functional polymorphism
in human glutathione transferase zeta by expressed sequence tag database analysis.
Pharmacogenetics 10:49-57.
Blackburn, AC; Coggan, M; Tzeng, HF; et al. (2001) GSTZld: a new allele of glutathione
transferase zeta and maleylacetoacetate isomerase. Pharmacogenetics 11:671-678.
Blair A, Hartge P, Stewart PA, McAdams M, Lubin J. 1998. Mortality and cancer incidence of
aircraft maintenance workers exposed to trichloroethylene and other organic solvents and
chemicals: extended follow up. Occup Environ Med 55:161-171.
Blake JE, Compton KV, Schmidt W, Orrego H. 1988. Accuracy of death certificates in the
diagnosis of alcoholic liver cirrhosis. Alcoholism: Clinical and Experimental Research
12:168-172.
Blechzcz BRA, Gores GJ. 2008. Cholangiocarcinoma. Clin Liver Disease 12:131-150.
Board, PG; Chelvanayagam, G; Jermiin, LS; et al. (2001) Identification of novel glutathione
transferases and polymorphic variants by expressed sequence tag database analysis. Drug
Metab Dispos 29:544-547.
Boice JD, Marano DE, Fryzek JP, Sadler CJ, McLaughlin JK. 1999. Mortality among aircraft
manufacturing workers. Occup Environ Med 56:581-597.
Boice JD, Marano DE, Cohen SS, Mumma MR, Blott WJ, Brill AB, Fryzek JP, Henderson BE,
McLaughlin JK. 2006. Mortality among Rocketdyne worker who tested rocket engines,
1948-1999. J Occup Environ Med 48:1070-1092.
Brodsky WY, Uryvaeva IV. 1977. Cell polyploidy: its relation to tissue growth and function. Int
Rev Cytol 50: 275-332.
Buben, JA; O'Flaherty, EJ. (1985) Delineation of the role of metabolism in the hepatotoxicity of
trichloroethylene and perchloroethylene: a dose-effect study. Toxicol Appl Pharmacol
78:105-122.
Bull, RJ. (2000) Mode of action of liver tumor induction by trichloroethylene and its metabolites,
trichloroacetate and dichloroacetate. Environ Health Perspect 108(Suppl 2):241-259.
6/22/2009
589
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Bull RJ. 2004. Trichloroethylene and liver tumors in mice. In: Proceedings of New Scientific
Research Related to the Health Effects of Trichloroethylene, 26-27 February 2004,
Washington DC. Washington, DC:U.S. Environmental Protection Agency. Available:
http://cfpub.epa.gov/ncea/cfm/recordisply.cfm?deid=75934 [accessed 3 August 2006],
Bull RJ, Sanchez IM, Nelson MA, Larson JL, Lansing AJ. 1990. Liver tumor induction in
B6C3F1 mice by dichloroacetate and trichloroacetate. Toxicology 63:341-359.
Bull, RJ; Orner, GA; Cheng, RS; et al. (2002) Contribution of dichloroacetate and
trichloroacetate to liver tumor induction in mice by trichloroethylene. Toxicol Appl
Pharmacol 182:55-65.
Caldwell JC, KeshavaN. 2006. Key issues in the modes of action and effects of
trichloroethylene metabolites for liver and kidney tumorigenesis. Environ Health Perspect
114:1457-1463.
Caldwell JC, Keshava N, Evans MV. 2008(b). Difficulty of mode of action determination for
trichloroethylene: an example of complex interactions of metabolites and other chemical
exposures. Environ Mol Mutagen 49(2): 142-54
Calvisi DF, Ladu S, Gorden A, Farina M, Lee JS, Conner EA, Schroeder I, Factor VM,
Thorgeirsson SS. 2007. Mechanistic and prognostic significance of aberrant methylation
in the molecular pathogenesis of human hepatocellular carcinoma. J Clin Invest 117(9):
2713-2711.
Carter JH, Carter HW, DeAngelo AB. 1995. Biochemical pathologic and morphometric
alterations induced in male B6C3F1 mouse liver by short-term exposure to dichloroacetic
acid. Toxicology Letters 81: 55-71.
Carter JH, Carter HW, Deddens JA, Hurst BM, George MH, DeAngelo AB. 2003. A 2-year
dose-response study of lesion sequences during hepatocellular carcinogenesis in the male
B6C3F1 mouse given the drinking water chemical dichloroacetic acid. Environ Health
Perspect 111(1): 53-64.
Chang LW, Daniel FB, DeAngelo AB. 1992. Analysis of DNA strand breaks induced in rodent
liver in vivo, hepatocytes in primary culture, and a human cell line by chlorinated acetic
acids and chlorinated acetaldehydes. Environ Mol Mutagen 20L 277-288.
Channel, SR; Latendresse, JR; Kidney, JK; et al. (1998) A subchronic exposure to
trichloroethylene causes lipid peroxidation and hepatocellular proliferation in male
B6C3F1 mouse liver. Toxicol Sci 43:145-154.
6/22/2009
590
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Charbonneau Mm Couture J, Plaa GL. 1991. Inhalation versus oral administration of acetone:
effect of the vehicle on the potentiation of CCL4- induced liver injury. Toxicology
Letters 57: 47-54.
Chen CW. 2000. Biologically based dose-response model for liver tumors induced by
trichloroethylene. Environ Health Perspect. 108 Suppl 2:335-42.
Chen C, Cheung ST, So S, Fan ST, Barry S, Higgins J, Lai KM, Ji J, Dudoit S, Ng IOI, van de
Rijn M, Botstein D, Brown PO. 2002. Gene expression patters in human liver cancers.
Molecular Biology of the Cell. 13: 1929-1939.
Chiesa R, Donato F, Tagger A, Favret, Ribero Nardi et al. 2000. Etiology of hepatocellular
carcinoma in Italian patients with and without cirrhosis. Cancer Epidemiol Biomarkers
Prev 9:213-216.
Cicmanec, JL; Condie, LW; Olson, GR; et al. (1991) 90-Day toxicity study of dichloroacetate in
dogs. Fundam Appl Toxicol 17:376-389.
Clapp RW, Hoffman K. 2008. Cancer mortality in IBM Endicott plant workers, 1969-2001: an
update on a NY production plant. Environ Health 7:13.
Collier JM, Selmin O, Johnson PD, Runyan RB. 2003. Trichloroethylene effects on gene
expression during cardiac development. Birth Defects Res A Clin Mol Teratol
67:488-495.
Cornett, R; James, MO; Henderson, GN; et al. (1999) Inhibition of glutathione S-transferase zeta
and tyrosine metabolism by dichloroacetate: a potential unifying mechanism for its
altered biotransformation and toxicity. Biochem Biophys Res Commun 262:752-756.
Counts JL, Sarmiento JI, Harbions ML, Downing JC, McClain RM, Goodman JI. 1996.
Carcinogenesis. 17(6): 1251-1257.
Daniel FB, DeAngelo AB, Stober JA, Olson GR, Page NP. 1992. Hepatocarcinogenicity of
chloral hydrate, 2-chloroacetaldehyde, and dichloroacetic acid in the male B6C3F1
mouse. Fundamental and Applied Toxicology 19: 159-168.
Davis SI, Pallos L, Wu JQ, Sapp JH, Cusack C. 2006. ATSDR's trichloroethylene subregistry
methods and results: 1989-2000. Arch Environ Occup Health 60:130-139.
DeAngelo, AB; Daniel, FB; McMillan, L; et al. (1989) Species and strain sensitivity to the
induction of peroxisome proliferation by chloroacetic acids. Toxicol Appl Pharmacol
101:285-298.
6/22/2009
591
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
DeAngelo, AB; Daniel, FB; Most, BM; et al. (1996) The carcinogenicity of dichloroacetic acid
in the male Fischer 344 rat. Toxicology 114:207-221.
DeAngelo, AB; Daniel, FB; Most, BM; et al. (1997) Failure of monochloroacetic acid and
trichloroacetic acid administered in the drinking water to produce liver cancer in male
F344/N rats. J Toxicol Environ Health 52:425-445.
DeAngelo AB, George MH, House DE. 1999. Hepatocarcinogenicity in the male B6C3F1
mouse following a lifetime exposure to dichloroacetic acid in the drinking water: dose-
response determination and modes of action. J Toxicol Environ Health 58:485-507.
DeAngelo AB, Daniel FB, Wong DM, George MH. 2008. The induction of hepatocellular
neoplasia by trichloroacetic acid administered in the drinking water of the male B6C3F1
mouse. J Toxicol and Environmental Health Part A 71: 1056-1068.
Dees, C; Travis, C. (1993) The mitogenic potential of trichloroethylene in B6C3F1 mice.
Toxicol Lett 69:129-137.
Dossing M, Petersen KT, Vyberg M, Olsen JH. 1997. Liver cancer among employees in
Denmark. Am J Ind Med 32:248-254.
Driscoll TR, Hamdan HH, Wang G, Wright PFA, Stacey NH. 1992. Concentrations of
individual serum or plasma bile acids in workers exposed to chlorinated aliphatic
hydrocarbons. Br J Ind Med 49:700-705.
Eichenlaub-Ritter, U., Baart, E., Yin, H. & Betzendahl, I. (1996) Mechanisms of spontaneous
and chemically-induced aneuploidy in mammalian oogenesis: Basis of sex-specific
differences in response to aneugens and the necessity for further tests. Mutat. Res., 372,
279-294
Elcombe, CR; Rose, MS; Pratt, IS. (1985) Biochemical, histological, and ultrastructural changes
in rat and mouse liver following the administration of trichloroethylene: possible
relevance to species differences in hepatocarcinogenicity. Toxicology and Applied
Pharmacology 79:365-376.
El-SeragHB. 2007. Epidemiology of hepatocellular carcinoma in USA. Hepatology Res
37(Suppl 2):S88-S94.
El-Serag HB, Rudolph KL. 2007. Hepatocellular carcinoma: epidemiology and molecular
carcinogenesis. Gastroenterology 132: 2557-2576.
6/22/2009
592
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Environmental Health Strategies. 1997. Final Report. Correspondence from Paul A. Cammer,
Ph.D., Trichloroethylene Issues Group, to Cheryl Siegel Scott, US Environmental
Protection Agency. December 22, 1997.
Evans, MV., Chiu, WA., Okino, MS., and Caldwell, JC. Development of an updated PBPK
model for trichloroethylene and metabolites in mice, and its application to discern the
role of oxidative metabolism in TCE-induced hepatomegaly, Toxicol. Appl. Pharmacol.,
236: 329-340.
Evert M, Schneider-Stock R, Dombrowski F. 2003. Apolipoprotein A-IV mRNA overexpression
in early preneoplastic hepatic foci induced by low-number pancreatic islet transplants in
streptozotocin-diabetic rats. Pathol Res Pract 199(6):373-379.
Fattovich G, Stroffolini T, Zagni I, Donato F. 2004. Hepatocellular carcinoma in cirrhosis:
incidence and risk factors. Gastroenterology, 127: S35-S50.
Feitelson MA, Sun B, Satiroglu- Tufan NL, Liu J, Pan J, Lian Z. 2002. Genetic mechanisms of
hepatocarcinogenesis. Oncogene 21: 2593-2604.
Ferlay, J; Bray, F; Pisani, P; et al. (2004) GLOBOCAN 2002: cancer incidence, mortality and
prevalence worldwide. IARC CancerBase No. 5, Version 2.0. Lyon, France: IARC Press.
Ferreira-Gonzalez DeAngelo AB, Nasim S, Garret CT. 1995. Ras oncogene activation during
hepatocarcinogenesis in B6 C3F1 male mice by dichloroacetic and trichloroacetic acids.
Carcinogenesis 16(3): 495-500.
Fox TR, Schumann AM, Watanabe PG, Yano BL, Maher BM. 1990. Mutational analysis of H-
ras oncogene in spontaneous C57B176 X C3H/He mouse liver tumors and tumors induced
with genotoxic and nongenotoxic hepatocarcinogens. Cancer Research 50: 4014-4019.
Fraga MF, Herranz M, Espada J, Ballestar E, Paz MF, Ropero S, et al. 2004. A mouse skin
multistage carcinogenesis model reflects the aberrant DNA methylation patterns of
human tumors. Cancer Res 64:5527-5534.
Fraga MF, Ballestar E, Villar-Garea A, Boix-Chornet M, Espada J, Schotta G, et al. 2005. Loss
of acetylation at Lysl6 and trimethylation at Lys20 of histone H4 is a common hallmark
of human cancer. Nat Genet 37:391-400
Fukuda K, Takemoto K, Tsuruta H. 1983. Inhalation carcinogenicity of trichloroethylene in
mice and rats. Industrial Health 21: 243-254.
Furnus CC, Ulrich MA, Terreros MC, Dulout FN. (1990) The induction of aneuploidy in
cultured Chinese hamster cells by propionaldehyde and chloral hydrate. Mutagenesis.
5(4):323-6.
6/22/2009
593
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Garabrant DH, Held J, Langholz B, Bernstein L. 1988. Mortality of Aircraft Manufacturing
Workers in Southern California. Am J Ind Med 13:683-693.
Ge R, Yang S, Kramer PM, Tao L, Pereira MA. 2001. The effect of dichloroacetic acid and
trichloroacetic acid on DNA methylation and cell proliferation in B6C3F1 mice. J
Biochem Mol Toxicol 15:100-106.
Ge R, Tao L, Kramer PM, Cunningham ML, Pereira MA. 2002. Effect of peroxisome
proliferators on the methylation and protein level of c-myc protooncogene in B6C3F1
mice liver. J Biochem Mol Toxicol 16(1): 41-47.
George MH, Moore T, Kilburn S, Olson GR, DeAngelo AB. 2000. Carcinogenicity of Chloral
hydrate administered in drinking water to the Male F344/N rat and Male B6C3F1 mouse.
Toxicologic Pathology 28(4): 610-618.
Ghoshal AK, Farber E. The induction of liver cancer by dietary deficiency of choline and
methionine without added carcinogens. Carcinogenesis. 1984 Oct;5(10): 1367-70.
Gibson, D.P., Aardema, M.J., Kerckaert, G.A., Carr, G.J., Brauninger, R.M. & LeBoeuf, R.A.
(1995) Detection of aneuploidy-inducing carcinogens in the Syrian hamster embryo
(SHE) cell transformation assay. Mutat. Res., 343, 7-24
Goel, SK; Rao, GS; Pandya, KP; et al. (1992) Trichloroethylene toxicity in mice: a biochemical,
hematological and pathological assessment. Indian J Exp Biol 30:402-406.
Goldsworthy, TL; Popp, JA. (1987) Chlorinated hydrocarbon-induced peroxisomal enzyme
activity in relation to species and organ carcinogenicity. Toxicol Appl Pharmacol
88:225-233.
Grasl-Kraupp B, Waldho' r T, Huber W, Schulte-Hermann R. 1993. Glutathione S-transferase
isoenzyme patters in different subtypes of enzyme-altered rat liver foci treated with the
peroxisome proliferator nafenopin or with Phenobarbital. Carcinogenesis 14:2407-2412.
Greenberg, M. S., Burton, G. A., and Fisher, J. W. (1999). Physiologically based
pharmacokinetic modeling of inhaled trichloroethylene and its oxidative metabolites in
B6C3F1 mice. Toxicol Appl Pharmacol 154, 264-278.
Greenland S, Salvan A, Wegman DH, HallockMF, Smith TJ. 1994. A case-control study of
cancer mortality at a transformer-assembly facility. Int Arch Occup Environ Health.
66(l):49-54.
Gupta S. 2000. Hepatic polyploidy and liver growth control. Seminars in Cancer Biology. 10:
161-171.
6/22/2009
594
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Guyton KZ, Chiu WA, Bateson TF, Jinot J, Scott CS, Brown RC, Caldwell JC. 2009. A
Reexamination of the PPAR-a Activation Mode of Action as a Basis for Assessing
Human Cancer Risks of Environmental Contaminants. Environ Health Perspect.
doi: 10.1289/ehp.0900758 available via http://dx.doi.org/ [Online 15 May 2009]
Hamdan H, Stacey NH. 1993. Mechanisms of trichloroethylene-induced elevation of individual
serum bile acids: I. Correlation of trichloroethylene concentrations to bile acids in rat
serum. Toxicol and Appl Pharmacol 121: 291-295.
Hansen J, Raaschou-Nielsen O, Christensen JM, Johansen I, McLaughlin JK, Lipworth L, Blot
WJ, Olsen JH. 2001. Cancer incidence among Danish workers exposed to
trichloroethylene. J Occup Environ Med 43:133-139.
Hardell L, Bengtsson NO, Johsson U, Eriksson S, Larsson LG. 1984. Aetiological aspects o
primary liver cancer with special regard to alcohol, organic solvents and acute
intermittent porphyria - an epidemiological investigation. Br J Cancer 50:389-297.
Hasmall SC, Roberts RA. 2000. The nongenotoxic hepatocarcinogen diethylhexylphthalate and
methylclofenebrate induce DNA synthesis preferentially in octoploid rat hepatocytes.
Toxicologic pathology 28(4): 503-509.
Hegi ME, Fox TR, Belinsky SA, Devereux TR, Anderson MW. 1993. Analysis of activated
protooncogenes in B6C3F1 mouse liver tumors induced by ciprofibrate, a potent
peroxisome proliferator. Carcinogenesis 14(1): 145-149.
Heinemann K, Willich SN, Heinemann AJ, DoMinh T, Mohner M, Heuchert GE, and the
Collaborative MILTS Project Team. 2000. Occupational exposure and liver cancer in
women: results of the Multicentre International Liver Tumour Study (MILTS). Occup
Med 50:422-429.
Henning SM, Swendseid ME. The role of folate, choline, and methionine in carcinogenesis
induced by methyl-deficient diets. Adv Exp Med Biol. 1996;399:143-55.
Henschler D, Romen W, Elsasser HM, Reichaert D, Eder E, Radwan Z. 1980. Carcinogenicity
study of trichloroethylene by longterm inhalation in three animal species. Arch Toxicol
43: 237-248 (1980).
Henschler D, Elsasser H, Romen W, Eder E. 1984. Carcinogenicity study of trichloroethylene,
with and without epoxide stabilizers, in mice. J Cancer Res Clin Oncol.
1984; 107(3): 149-56.
6/22/2009
595
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Hernberg S, Kauppinen T, Riala R, Korkala M-L, Aslkainen U. 1988. Increased risk for
primary liver cancer among women exposed to solvents. Schan J work Environ Health
14:356-365.
Hernberg S, Korkala M-L, Asikainen U, Riala R. 1984. Primary liver cancer and exposure to
solvents. Int Arch Occup Environ Health 54:147-153.
Herman JG, Umar A, Polyak K, Graff JR, Ahuja N, Issa JP, et al. 1998. Incidence and functional
consequences of hMLHl promoter hypermethylation in colorectal carcinoma. Proc Natl
Acad Sci USA 95(12):6870-6785.
Herren-Freund SL, Pereiira MA, Khoury MD, Olson G. 1987. The carcinogenicity of
trichloroethylene and its metabolites trichloroacetic acid and dichloroacetic acid, in
mouse liver. Toxicol Appl Pharmacol. 90: 183-189.
Hoivik DJ, Quails CW, Mirabile RC, Cariello NF, Kimbrough CL, Colton HM, Anderson SP,
Santostefano MJ, Morgan RJO, Dahl RR, Brown AR, Zhao Z, Mudd PN, Oliver WB,
Brown HR, and Miller RT. 2004. Fibrates induce hepatic peroxisome and mitochondrial
proliferation without evidence of cellular proliferation and oxidative stress in cynomolgus
monkeys. Carcinogenesis 25(9): 1757-1769.
Huang HL, Li LY, Chen B, Huang JX, Huang SR. 2002. New problems caused by occupational
trichloroethylene exposure. Int J Immunopath Pharm 15:30-32.
Ito Y, Yamanoshita O, Asaeda N, Tagawa Y, Lee CH, Aoyama T, Ishihara G, Furuhashi K,
Kamijima M, Gonzalez FJ, Nakajima T. 2007. Di(2-ethylhexyl)phthalate induced hepatic
tumorigenesis through a peroxisome proliferator-activated receptor a-independent
pathway. J Occup Health 49:172-182.
Jemal A, Siegel R, Ward E, Hao Y, Xu J, Murray T, Thun MJ. 2008. Cancer Statistics, 2008.
CA Cancer J Coin 58:71-96.
Ji J, Hemminki K. 2005. Variation in the risk for liver and gallbladder cancers in
socioeconomic and occupational groups in Sweden with etiological implications. Int
Arch Occup Environ health 78:641-649.
Jorquera, R; Tanguay, RM. (2001) Fumarylacetoacetate, the metabolite accumulating in
hereditary tyrosinemia, activates the ERK pathway and induces mitotic abnormalities and
genomic instability. Hum Mol Genet 10:1741-1752.
6/22/2009
596
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Juran BD, Lazaridis KN. 2006. Genomics and complex liver disease: Challenges and
Opportunities. Hepatology 44:1380-1390.
Kamijima M, Hisanaga N, Wang H, Nakajima T. 2007. Occupational trichloroethylene exposure
as a cause of idiosyncratic generalized skin disorders and accompanying hepatitis similar
to drug hypersensitivities. Int Arch Occup Environ Health 80:357-370.
Kaneko, T; Saegusa, M; Tasaka, K; et al. (2000) Immunotoxicity of trichloroethylene: a study
with MRL-lpr/lpr mice. J Appl Toxicol 20:471-475.
Kato-Weinstein, J; Lingohr, MK; Orner, GA; et al. (1998) Effects of dichloroacetate on glycogen
metabolism inB6C3Fl mice. Toxicology 130:141-154.
Kato-Weinstein J, Stauber AJ, Orner GA., Thrall BD, Bull RJ. 2001. Differential effects of
dihalogenated and trihalogenated acetates in the liver of B6C3 F1 mice. Journal of
Applied Toxicology 21: 81-89.
Kawamoto T, Hobara T, Nakamura K, Imamura A, Ogino K, Kobayashi H, Iwamoto S, Sakai T.
1988. Induction of cytochrome P450, cytochrome b-5, NADPH cytochrome c reductase
and change in cytochrome P-450 isozymes with long-term trichloroethylene treatment.
Toxicology 53: 239-249.
Kelsh MA, Weihgart M, Mandel J, Mink P, Alexander D, Basu R, Kalmes R, Goodman M.
2005. A meta-analysis of epidemiology studies of occupational TCE exposure and
selected cancers. Presented to National Research Committee on Health Risks of
Trichloroethylene. June 9, 2005.
Keshava, N; Caldwell, JC. (2006) Key issues in the role of peroxisome proliferator-activated
receptor agonism and cell signaling in trichloroethylene toxicity. Environ Health Perspect
114:1464-1470.
Kim HJ, Odend'hal S, Bruckner JV. 1990. Effect of oral dosing vehicles on the acute
hepatotoxicity of carbon tetrachloride in rats. Toxicol Appl Pharmacol 102: 34-49.
Kim, SZ; Kupke, KG; Ierardi-Curto, L; et al. (2000) Hepatocellular carcinoma despite long-term
survival in chronic tyrosinaemia I. J Inherit Metab Dis 23:791-804.
Kjellstrand P, Kanje M, Mansson L, Bjerkemo M. Mortensen I, Lanke J, Holmquist B. 1981.
Toxicology 21: 105-115.
6/22/2009
597
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Kjellstrand, P; Holmquist, B; Aim, P; et al. (1983a) Trichloroethylene: further studies of the
effects on body and organ weights and plasma butyrylcholinesterase activity in mice.
Acta Pharmacol Toxicol (Copenh) 53:375-384.
Kjellstrand, P; Holmquist, B; Mandahl, N; et al. (1983b) Effects of continuous trichloroethylene
inhalation on different strains of mice. Acta Pharmacol Toxicol (Copenh) 53:369-374.
Klaunig, JE; Babich, MA; Baetcke, KP; et al. (2003) PPARalpha agonist-induced rodent tumors:
modes of action and human relevance. Crit Rev Toxicol 33:655-780.
Kraupp-Grasl B, Huber W, Putz Bm Gerbracht U, Schulte-Herman R. 1990. Tumor promotion
by peroxisome proliferator nafenopin involving specific subtype of altered foci in rat
liver. Cancer Research 50:37-1-3706.
Kumar, P; Prasad, AK; Maji, BK; et al. (2001) Hepatotoxic alterations induced by inhalation of
trichloroethylene (TCE) in rats. Biomed Environ Sci 14:325-332.
Kumar M, Kumar R, Hissar SS, SaraswatMK, SharmaBC, SakhujaP, Sarin SK. 2007. Risk
factors analysis for hepatocellular carcinoma in patients with and without cirrhosis: A
case-control study of 213 hepatocellular carcinoma patients from India. J Gastroent
Hepatol 22:1104-1111.
Kung H-C, Hoyert DL, Xu J, Murphy SL. 2008. Deaths: Final Data for 2005. National Vital
Statistics Reports, U.S. Department of Health and Human Services, Centers for Disease
Control and Prevention. Vol 56, Issue 10. Accessed May 29, 2008;
http://www.cdc.gov/nchs/data/nvsr/nvsr56/nvsr56 10.pdf
Kvam BM, Romundstad PR, Boffestta P, Andersen A. 2005. Cancer in the Norwegian printing
industry. Scand J Work Environ Health 31:36-43.
Larson JL, Bull RJ. 1992. Metabolism of lipoperoxidative activity of trichloroacetate and
dichloroacetate in rats and mice. Toxicology and Applied Pharmacology 115L 268-277.
Latendresse JR, PereiraMA. 1997. Dissimilar characteristics of N-methyl-N-nitrosourea-
initiated foci and tumors promoted by dichloroacetic acid or trichloroacetic acid in the
liver of female B6C3F1 mice. Toxicol Pathol. 25(5):433-40.
Laughter, AR; Dunn, CS; Swanson, CL; et al. (2004) Role of the peroxisome proliferator-
activated receptor alpha (PPARalpha) in responses to trichloroethylene and metabolites,
trichloroacetate and dichloroacetate in mouse liver. Toxicology 203:83-98.
LaVecchia C, Negri E, Franceschi S, D'Avanzo B, Boyle P. 1994. A case-control study of
diabetes mellitus and cancer risk. Br J Cancer 70:950-953.
6/22/2009
598
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Lawrence T, Hagemann T, Balkwill F. 2007. Sex, cytokines, and cancer. Science 317: 51-52.
Leakey JEA, Seng JE, Lantendresse JR, Hussain N, Allen LJ, Allaben WT. 2003a. Dietary
controlled carcinogenicity study of chloral hydrate in male B6C3F1 mice. Toxicology
and Applied Pharmacology 193: 266-280.
Leakey JEA, Seng JE, Allaben WT. 2003b. Body weight considerations in the B6C3F1 mouse
and the use of dietary control to standardize background tumor incidence in chronic
bioassays. Toxicology and Applied Pharmacology 193: 237-265.
Lee JH, Chung CW, Ma YC, Want GS, Chen PC, Hwang YH, et al. 2003. Increased mortality
odds ratio of male liver cancer in a community contaminated by chlorinated
hydrocarbons in groundwater. Occup Environ Med 60:364-369.
Leigh JP, Jiang W. 1993. liver cirrhosis deaths within occupations and industries in the
California occupational mortality study. Addiction 88:767-779.
Leopardi, P., Zijno, A., Bassani, B. & Pacchierotti, F. (1993) In vivo studies on chemically
induced aneuploidy in mouse somatic and germinal cells. Mutat. Res., 287, 119-130
Leuschner J, Beuscher N. 1998. Studies on the mutagenic and carcinogenic potential of chloral
hydrate. Arzneim-Forsch./Drug Res. 48(11): 961-968.
Lindbohm ML, Sallmen M, Kyyronen P, kauppinen T, Pukkala E. 2009. Risk of liver cancer
and exposure to organic solvents and gasoline vapors among Finnish workers. Int J
Cancer EPUB DOI 10.1002/ijc.24309
Lingohr, MK; Thrall, BD; Bull, RJ. (2001) Effects of dichloroacetate (DCA) on serum insulin
levels and insulin-controlled signaling proteins in livers of male B6C3F1 mice. Toxicol
Sci 59:178-184.
Maloney, EK; Waxman, DJ. (1999) trans-Activation of PPARalpha and PPARgamma by
structurally diverse environmental chemicals. Toxicol Appl Pharmacol 161:209-218.
Maltoni, C; Lefemine, G; Cotti, G. (1986) Experimental research on trichloroethylene
carcinogenesis. In: Maltoni, C; Mehlman MA., eds. Vol. 5. Archives of research on
industrial carcinogenesis. Princeton, NJ: Princeton Scientific Publishing;
Maltoni C, Lefemine G, Cotti G, Perino G. 1988. Long-term carcinogenicity bioassays on
trichloroethylene administered by inhalation to Sprague-Dawley rats and Swiss and
B6C3F1 mice. Annals of the New York Academy of Sciences 534: 316-343.
6/22/2009
599
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Mandel JH, Kelsh MA, Mink PJ, Alexander DD, Kalmes RM, Weingart M, Yost L, Goodman
M. 2006. Occupational trichloroethylene exposure and non-Hodgkin's lymphoma: a
meta-analysis and review. Occup Environ Med 2006 63:597-607.
Maronpot RR, Fox T„ Malarkey DE, Goldsworthy TL. 1995. Mutations in the ras proto-
oncogene clues to etiology and molecular pathogenesis of mouse liver tumors.
Toxicology 101: 125-156.
Marrazzini, A., Betti, C., Bernacchi, F., Barrai, I. & Barale, R. (1994) Micronucleus test and
metaphase analyses in mice exposed to known and suspected spindle poisons.
Mutagenesis, 9, 505-515
Marsman, DS; Cattley, RC; Conway, JG; et al. (1988) Relationship of hepatic peroxisome
proliferation and replicative DNA synthesis to the hepatocarcinogenicity of the
peroxisome proliferators di(2-ethylhexyl)phthalate and [4-chloro-6-(2,3-xylidino)-2-
pyrimidinylthiojacetic acid (Wy-14,643) in rats. Cancer Res 48:6739-6744.
Mather, GG; Exon, JH; Koller, LD. (1990) Subchronic 90 day toxicity of dichloroacetic and
trichloroacetic acid in rats. Toxicology 64:71-80.
Melchiorri C, Chieco P, Zedda AI, Coni P, Ledda-Columbano GM, Columbano A. 1993. Ploidy
and nuclearity of rat hepatocytes after compensatory regeneration or mitogen-induced
liver growth. Carcinogenesis 14(9): 1825-1830.
Melnick, RL; Jameson, CW; Goehl, TJ; et al. (1987) Application of microencapsulation for
toxicology studies. II. Toxicity of microencapsulated trichloroethylene in Fischer 344
rats. Fundam Appl Toxicol 8:432-442.
Merrick BA, Robinson M, Condi LW. 1989. Differing hepatotoxicity and lethality after
subacute trichloroethylene exposure in aqueous or corn oil gavage vehicles in B6C3F1
mice. Journal of Applied Toxicology. 9(1): 15—21.
Michel C, Desdouets C, Slaoui M, Isaacs KR, Roberts RA, Boitier E. 2007. Diethylnitrosamine
initiation does not alter clofibric acid-induced hepatocarcinogenesis in the rat.
Toxicological Sciences. 99(1): 58-69.
Mikol YB, Hoover KL, Creasia D, Poirier LA. 1983. Hepatocarcinogenesis in rats fed methyl-
deficient, amino acid-defined diets. Carcinogenesis. 4(12): 1619-29.
Miller RT, Shah RS, Cattley RC, Popp JA. 1996. The peroxisome proliferator WY-14,643 and
methylclofenapate induce hepatocellular ploidy alterations and ploidy-specific DNA
synthesis in F344 rats. Toxicology and Applied Pharmacology 138: 317-323.
6/22/2009
600
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Mirsalis JC, Tyson CK, Steinmetz KL, Loh EK, Hamilton CM, Bakke JP, Spalding JW. 1989.
Measurement of unscheduled DNA synthesis and S-phase synthesis in rodent hepatocytes
following in vivo treatment: testing of 24 compounds. Environmental and Molecular
Mutagenesis 14: 155-164.
Moore MM, Harrington-Brock K. (2000) Mutagenicity of trichloroethylene and its metabolites:
implications for the risk assessment of trichloroethylene. Environ Health Perspect. 108
Suppl 2:215-23.
Morgan RW, Kelsh MA, Zhao K, Heringer S. 1998. Mortality of aerospace workers exposed to
trichloroethylene. Epidemiology 9:424-431.
Morgan RW, Kelsh MA, Zhao K, Heringer S. 2000. Mortality of aerospace workers exposed to
trichloroethylene [erratum]. Epidemiology 11:360.
Morgan JW, Cassady RE. 2002. Community cancer assessment in response to long-time
exposure to perchlorate and trichloroethylene in drinking water. J Occup Environ Med
44:616-621.
Morimura K, Cheung C, Ward JM, Reddy JK, Gonzalez FJ. 2006. Differential susceptibility of
mice humanized for peroxisome proliferator-activated receptor {alpha} to Wy-14,643
induced liver turnorigenesis. Carcinogenesis. 27:1074-1080.
Nagaya T, Ishikawa N, Hata H, Otobe T. 1993. Subclinical and reversible hepatic effects of
occupational exposure to trichloroethylene Int Arch Occup Environ Health 64:561-563.
Nakajima, T; Okino, T; Okuyama, S; et al. (1988) Ethanol-induced enhancement of
trichloroethylene metabolism and hepatotoxicity: difference from the effect of
phenobarbital. Toxicol Appl Pharmacol 94:227-237.
Nakajima, T; Kamijo, Y; Usuda, N; et al. (2000) Sex-dependent regulation of hepatic
peroxisome proliferation in mice by trichloroethylene via peroxisome proliferator-
activated receptor alpha (PPARalpha). Carcinogenesis 21:677-682.
Naugler WE, Sakurai T, Kim S, Maeda S, Kim K, Elsharkawy AM, Karin M. 2007. Gender
disparity in liver cancer due to sex differences in MyD88-dependent IL-6 production.
Science 317: 121-124.
Neghab M, Stacey NH. 1997. Toluene-induced elevation of serum bile acids: relationship to
bile acid transport. J Toxicol and Environ Health. 52: 249-268.
6/22/2009
601
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Neghab M, Qu Sm, Bai CL, Caples J, Stacey NH. 1997. Raised concentration of serum bile
acids following occupational exposure to halogenated solvents, 1,1,2-trichloro-1,2,2-
trifluoroethane and trichloroethylene. Int Arch Occup Environ Health 70: 187-194.
Nelson MA, Bull RJ. 1988. Induction of strand breaks in DNA by trichloroethylene and
metabolites in rat and mouse liver in vivo. Toxicology and Applied Pharmacol 94:
45-54.
Nelson MA, Lansing AJ, Sanches AM, Bull RJ. 1989. Dichloroacetic acid and trichloroacetic
acid-induced DNA strand breaks are independent of peroxisome proliferation.
Toxicology 58:239-248.
Nelson MA, Sanchez, Bull RJ, Sylvester SR. 1990. Increased expression of c-myc and c-H-ras
in dichloroacetate and trichloroacetate-induced liver tumors in B6C3F1 mice.
Toxicology 64: 47-57.
NRC (National Research Council). (2006) Assessing the human health risks of trichloroethylene:
key scientific issues. Washington, DC: National Academies Press.
National Toxicology Program (NTP) Technical Report No. 273. Toxicology and Carcinogenesis
studies of trichloroethylene (Cas No. 79-01-6) in four strains of rats (ACI, August,
Mashal, Osborne-Mendel) (gavage studies). April 1988.
NTP (National Toxicology Program). (1990) Carcinogenesis studies of trichloroethylene
(without epichlorohydrin) in F344/N rats and B6C3F1 mice (gavage studies). Public
Health Service, U.S. Department of Health and Human Services; NTP TR-243; NIH
Publication No. 90-1779. Available from the National Institute of Environmental Health
Sciences, Research Triangle Park, NC.
http://ntp.niehs.nih.gov/ntp/htdocs/LT_rpts/tr243.pdf.
Nunes J, Ehrich M, Robertson J. 2001. Toxicosis associated with dual oral exposure of rats to
lead and trichloroethylene. Toxicologic Pathology. 29(4): 451-457.
Okino, T; Nakajima, T; Nakano, M. (1991) Morphological and biochemical analyses of
trichloroethylene hepatotoxicity: differences in ethanol- and phenobarbital-pretreated
rats. Toxicol Appl Pharmacol 108:379-389.
Palmer CN, Hsu MH, Griffin KJ, Raucy JL, Johnson EF. 1998. Peroxisome proliferator activated
receptor-alpha expression in human liver. Mol Pharmacol. 53(1): 14-22.
Parrish, JM; Austin, EW; Stevens, DK; et al. (1996) Haloacetate-induced oxidative damage to
DNA in the liver ofmaleB6C3Fl mice. Toxicology 110:103-111.
6/22/2009
602
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Pereira MA, Phelps BJ. 1996. Promotion by dichloroacetic acid and trichloroacetic acid of N-
methyl-N-nitrosurea-initiated cancer in the liver of female B6C3F1 mice. Cancer Letters
102: 133-141.
Pereira MA, Kramer PM, Conran PB, Tao L. 2001. Effect of chloroform on dichloroacetic acid
and trichloroacetic acid-induced hypomethylation and expression of the c-myc gene and
on their promotion of liver and kidney tumors in mice. Carcinogenesis 22(9):
1511-1519.
Pereira MA, Wang W, Kramer PM, Tao L. 2004. Prevention by methionine of dichloroacetic
acid-induced liver cancer and DNA hypomethylation in mice. Toxicological Sciences
77: 243-248.
Pereira MA, Phelps BJ. 1996. Promotion by dichloroacetic acid and trichloroacetic acid of N-
methyl-N-nitrosurea-initiated cancer in the liver of female B6C3F1 mice. Cancer Letters
102: 133-141.
Peters JM, Cattley RC, Gonzalez FJ. 1997. Role of PPAR alpha in the mechanism of action of
the nongenotoxic carcinogen and peroxisome proliferator Wy-14,643. Carcinogenesis.
18:2029-33.
Pogribny IP, Tryndyak VP, Woods CG, Witt SE, Rusyn I. 2007. Epigenetic effects of the
continuous exposure to peroxisome proliferator WY-14,643 in mouse liver are dependent
on peroxisome proliferator activated receptor a. Mutation Research 625:62-71.
Porru S, Placidi D, Carta A, Gelati U, Ribero ML, Tagger A, Boffetta P, Donato F. 2001.
Primary liver cancer and occupation in men: A case-control study in a high-incidence
area in Northern Italy. Int J Cancer 94:878-883.
Raaschou-Nielsen O, Hansen J, McLaughlin JK, Kolstad H, Christensen JM, Tarone RE, Olsen
JH. 2003. Cancer risk among workers at Danish companies using trichloroethylene: a
cohort study. Am JEpidemiol 158:1182-1192.
Radican L, Blair A, Stewart P, Wartenberg D. 2008. Mortality follow-up in hydrocarbon-
exposed workers. J Occup Environ Med 50:1306-1319.
Rake JP, Visser G, Labrune P, Leonard JV, Ullrich K, Smit GP. 2002. Glycogen storage disease
type I: diagnosis, management, clinical course and outcome. Results of the European
Study on Glycogen Storage Disease Type I (ESGSD I). Eur J Pediatr 161(Suppl
1):S20-S34.
6/22/2009
603
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Rakoff-Nahoum S, Medzhitov R. 2007. Regulation of spontaneous intestinal tumorigenesis
through the adaptor protein MyD88. Science 317: 124-127.
Ramdhan DH, Kamijima M, Yamada N, Ito Y, Yanagiba Y, Nakamura D, Okamura A, Ichihara
G, Aoyama T, Gonzalez FJ, Nakajima T. 2008. Molecular mechanisms of
trichloroethylene-induced hepatotoxicity mediated by CYP2E1. Toxicol and Appl
Pharmacol 231: 300-307.
Rasmussen K, Brogren C-H, Sabroe S. 1993. Subclinical affection of liver and kidney function
and solvent exposure. Int Arch Occup Environ Health 64:445-448.
Ries LAG, Melbert D, Krapcho M, Stinchcomb DG, Howlader N, Horner MJ, Mariotto A,
Miller BA, Feuer EJ, Altekruse SF, Lewis DR, Clegg L, Eisner MP, Reichman M,
Edwards BK (eds). SEER Cancer Statistics Review, 1975-2005, National Cancer
Institute. Bethesda, MD, http://seer.cancer.gov/csr/1975_2005/, accessed June 9, 2008.
Rhee I, Bachman KE, Park BH, Jair KW, Yen RW, et al. 2002. DNMT1 and DNMT3b
cooperate to silence genes in human cancer cells. Nature 416(6880):552-556.
Richmond, RE; Carter, JH; Carter, HW; et al. (1995) Immunohistochemical analysis of
dichloroacetic acid (DCA)-induced hepatocarcinogenesis in male Fischer (F344) rats.
Cancer Lett 92:67-76.
Ritz B. 1999. Cancer mortality among workers exposed to chemicals during uranium processing.
J Occup Environ Med 41:556-566.
Russo, A., Levis, A.G. (1992) Further evidence for the aneuploidogenic properties of chelating
agents: Induction of micronuclei in mouse male germ cells by EDTA. Environ, mol.
Mutag., 19, 125-131
Russo A, Stocco A, Majone F (1992) Identification of kinetochore-containing (CREST+)
micronuclei in mouse bone marrow erythrocytes. Mutagenesis 7(3): 195-197.
Rusyn I, Bradham CA, Cohn L, Schoonhoven R, Swenberg JA, Brenner DA, Thurman RG.
1999. Corn oil rapidly activates nuclear factor - kB in hepatic Kupffer cells by oxidant-
dependent mechanisms. Carcinogenesis 20 (11): 2095-2100.
Rusyn I, Peters JM, Cunningham ML. 2006. Modes of action and species-specific effects of di-
(2-ethylhexl)phthalate in the liver. Critical Reviews in Toxicology. 36:459-479.
Sanchez IM, Bull RJ. 1990. Early induction of reparative hyperplasia in the liver of B6C3F1
mice treated with dichloroacetate and trichloroacetate. Toxicology 64: 33-46.
6/22/2009
604
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Sanders, V.M., Kauffmann, B.M., White, K.L., Jr, Douglas, K.A., Barnes, D.W., Sain, L.E.,
Bradshaw, T.J., Borzelleca, J.F. & Munson, A.E. (1982) Toxicology of chloral hydrate in
the mouse. Environ. Health Perspect., 44, 137-146
Scharf JG, Dombrowski F, Ramadori G. 2001. The IGF axis and hepatocarcinogenesis. Mol
Pathol 54:138-144.
Schultz, IR; Merdink, JL; Gonzalez-Leon, A; et al. (2002) Dichloroacetate toxicokinetics and
disruption of tyrosine catabolism in B6C3F1 mice: dose-response relationships and age
as a modifying factor. Toxicology 173:229-247.
Seng JE, Agrawal N, Horsley ETM, Leakey TI, Scherer EM, Xia S, Allaben WT, Leakey JEA.
2003. Toxicokinetics of chloral hydrate in ad libitum-fed, dietary controlled and
calorically restricted male B6C3F1 mice following short-term exposure. Toxicol and
Appl Pharmacol 193(2): 281-292.
Shah YM, Morimura K, Yang Q, Tanabe T, Takagi M, Gonzalez FJ. 2007. Peroxisome
proliferator-activated receptor a regulates a microRNA-mediated signaling cascade
responsible for Hepatocellular proliferation. Molecular and cellular biology
27(12):423 8-4247.
Shaib Y, El-Serag HB. 2004. The epidemiology of cholangiocarcinoma. Semin Liver Disease.
24:115-125.
Shankar K, Vaidya VS, Corton JJ, Bucci TJ, Liu J, Waalkes MP, Mehendale HM. 2003.
Activation of PPAR-a in streptozotocin-induced diabetes is essential for resistance
against acetaminophen toxicity. FASEB J. 17(12): 1748-50
Snyder RD, Pullman J, Carter JH, Carter HW, DeAngelo AB. 1995. In vivo administration of
dichloroacetic acid suppresses spontaneous apoptosis in murine hepatocytes. Cancer
Research 55: 3702-3705.
Stanley LA, Blackburn DR, Devereaux S, Foley J, Lord PG, Maronpot RR, Orton TC, Anderson
MW. 1994. Ras mutations in methylclofenapate-induced B6C3F1 and C57BL/10J
mouse liver tumors. Carcinogenesis 15(6): 1125-1131.
Stauber AJ, Bull RJ. 1997. Differences in phenotype and cell replicative behavior of hepatic
tumors induced by dichloroacetate (DCA) and trichloroacetate (TCA). Toxicology and
Applied pharmacology 144: 235-246.
6/22/2009
605
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Stauber AJ, Bull RJ, Thrall BD. 1998. Dichloroacetate and trichloroacetate promote clonal
expansion of anchorage-independent hepatocytes in vivo and in vitro. Toxicology and
Applied Pharmacology 150: 287-294.
Stemhagen A, Slade J, Altman R, Bill J. 1983. Occupational risk factors and liver cancer.
Amer J Epidemiol 117:443-454.
Stephenson RP. 1956. A modification of receptor theory. Br. J. Pharmacol. 11:379-93
Styles JA, Kelly M, Pritchard NR, Elcombe CR. 1988. A species comparison of acute
hyperplasia induced by peroxisome proliferatory methylcofenapate: involvement of the
binucleated hepatocyte. Carcinogenesis 9: 1647-1655.
Styles JA, Wyatt I, Coutts C. 1991. Trichloroacetic acid: studies on uptake and effects on
hepatic DNA and liver growth in mouse. Carcinogenesis. 12(9): 1715-1719.
Takashima K, Ito Y, Gonzalez FJ, Nakajima T. 2008. Different mechanisms of DEHP-induced
hepatocellular adenoma tumorigenesis in wild-type and PPAR alpha-null mice. J Occup
Health. 2008 Mar;50(2): 169-80.
Tanguay, RM; Jorquera, R; Poudrier, J; et al. (1996) Tyrosine and its catabolites: from disease to
cancer. Acta Biochim Pol 43:209-216.
Tao L, Yan S, Xie M, Kramer PM, Pereira MA. 2000. Effect of trichloroethylene and its
metabolites, dichloroacetic acid and trichloroacetic acid, on the methylation and
expression of c-Jun and c-Myc protooncogenes in mouse liver: prevention by methionine.
Toxicological Sciences 54: 399-407.
Tao L, Li Y, Kramer PM, Wang W, Pereira MA. 2004. Hypomethylation of DNA and the
insulin-like growth factor-II gene in dichloroacetic and trichloroacetic acid-promoted
mouse liver tumors. Toxicology 196: 127-136.
Thiele DL, Eigenbrodt EH, Ware AJ. 1982. Cirrhosis after reported trichloroethylene and 1,1,
1-trichl or ethane exposure. Gastroenterol 83:926-929.
Toraason M, Clark J, Dankovic D, Mathias P, Skaggs S, Walker C, Werren D. 1999. Oxidative
stress and DNA damage in Fischer rats following acute exposure to trichloroethylene or
perchloroethylene. Toxicology 138: 43-53.
Tryndyak VP, Muskhelishvili L, Kovalchuk O, Rodriguez-Juarez R, Montgomery B, Churchwell
MI, Ross SA, Beland FA, Pogribny IP. 2006. Effect of long-term tamoxifen exposure on
genotoxic and epigenetic changes in rat liver; implications for tamoxifen-induced
hepatocarcinogenesis. Carcinogenesis 27(8) 1713-1720.
6/22/2009
606
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Tucker AN, Sanders VM, Barnes DW, Bradshaw TJ, White KL, Sain LE, Borzelleca JF,
Munson AE. 1982. Toxicology of trichloroethylene in the mouse. Toxicology and
Applied Pharmacology 62: 351-357.
Tugwood JD, Aldridge TC, Lambe KG, Macdonald N, Woodyatt NJ. 1996. Peroxisome
proliferator-activated receptors: structures and function. Ann N Y Acad Sci 804:
252-265.
Tzeng, HF; Blackburn, AC; Board, PG; et al. (2000) Polymorphism- and species-dependent
inactivation of glutathione transferase zeta by dichloroacetate. Chem Res Toxicol
13:231-236.
US Environmental Protection Agency (EPA). 2008. Integrated Risk Information System. 1,1,
2-Trichloro-l, 2, 2-trifluoroethane. Accessed June 2, 2008.
http://www.epa. gov/IRIS/sub st/0123. htm.
Van Duuren BL, Goldschmidt BM, Loewengart G, Smith AC, Meichionne S, Seidman I, Roth
D. 1979. Carcinogenicity of halogenated olefinic and aliphatic hydrocarbons in mice.
JNCI 63(6): 1433-1439.
Vartiainen T, Pukkala E, Rienoja T, Strandman T, Kaksonen K. 1993. Population exposure to
tri- and tetrachloroethene and cancer risk: two cases of drinking water pollution.
Chemosphere 27:1171-1181.
Vickers AEM, Lucier GW. 1996. Estrogen receptor levels and occupancy in hepatic sinusoidal
endothelial and Kupffer cells are enhanced by initiation with diethylnitroamine and
promotion with 17 a-ethinylestradiol in rats. Carcinogenesis 17(6): 1235-1242.
Voss C, Zerban H, Bannasch P, Berger MR. 2005. Lifelong exposure to di-(2-ethylhexyl)-
phthalated induced tumors in liver and testes of Sprague-Dawley rats. Toxicology 2006:
359-371.
Voss KA, Riley R, Dunn C, Corton JC. 2006. The role of tumor necrosis factor alpha and the
peroxisome proliferator-activated receptor alpha in modulating the effects of fumonisin in
mouse liver. Toxicology. 222:165-74.
Wainfan E, Poirier LA. 1992. Methyl groups in carcinogenesis: effects on DNA methylation and
gene expression. Cancer Res.;52(7 Suppl):2071s-2077s.
Walgren, JE; Kurtz, DT; McMillan, JM. (2000a) Expression of PPAR(alpha) in human
hepatocytes and activation by trichloroacetate and dichloroacetate. Res Commun Mol
Pathol Pharmacol 108:116-132.
6/22/2009
607
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Walgren, JE; Kurtz, DT; McMillan, JM. (2000b) The effect of the trichloroethylene metabolites
trichloroacetate and dichloroacetate on peroxisome proliferation and DNA synthesis in
cultured human hepatocytes. Cell Biol Toxicol 16:257-273.
Wang G, Stacey NH. 1990. Elevation of individual serum bile acids on exposure to
trichloroethylene or a- naphthylisothioncyanate. Toxicol and Appl Pharmacol 105:
209-215.
Ward JM, Peters JM, Perella CM, Gonzalez FJ. 1998. Receptor and nonreceptor-mediated
organ-specific toxicity of di(2-ethylhexyl)phthalate (DEHP) in peroxisome proliferator-
activated receptor alpha-null mice. Toxicol Pathol. 26(2):240-6.
Wartenberg D, Reyner D, Scott CS. 2000. Trichloroethylene and Cancer: Epidemiologic
Evidence. Environ Health Perspect 108(suppl 2): 161-176.
Watanabe, S; Fukui, T. (2000) Suppressive effect of curcumin on trichloroethylene-induced
oxidative stress. JNutr Sci Vitaminol (Tokyo) 46:230-234.
Weiderpass E, Vainion H, Kauppinen T, Vasama-Neuvonen K, Partanen T, Pukkala E. 2003.
Occupational exposure and gastrointestinal cancers among Finnish women. J Occup
Environ Med 45:305-315.
Welzel TM, McGlynn KA, Hsing AW, O'Brien TR, Pfeiffer RM. 2006. Impact of classification
of Hilar Cholangiocarcinomas (Klatskin Tumros) on the incidence of intra- and
extrahepatic cholangiocarcinoma in the United States. J Natl Cancer Inst 98:873-875.
Werner H, Le Roith D. 2000. New concepts in regulation and function of the insulin-like growth
factors: implications for understanding normal growth and neoplasia. Cell Mol Life Sci
57:932-942.
Wideroff L, Gridley G, Mellemkjaer L, Chow WH, Linet M, Keehn S, et al. 1997. Cancer
incidence in a population-based cohort of patients hospitalized with diabetes mellitus in
Denmark. J Natl Cancer Inst 89:1360-1365.
Woods CG, Burns AM, Bradford BU, Ross PK, Kosyk O, Swenburg JA, Cunningham ML,
Rusyn I. 2007. Wy-14,643-induced cell proliferation and oxidative stress in mouse liver
are independent of NADPH oxidase. Toxicological Sciences 98(2): 366-374.
Woolhiser MR, Krieger SM, Thomas J, Hotchkiss JA. 2006. Report from the Dow Chemical
Company to ATSDR and EPA, Study ID # 031020.Trichloroethylene (TCE):
immunotoxicity potential in CD rats following a 4-week vapor inhalation exposure
6/22/2009
608
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Xu X, Yang R, Wu N, Zhong P, Ke Y, Zhou L, Yuan J, Li G, Huang H, Wu B. 2009. Severe
hypersensitivity dermatitis and liver dysfunction induced by occupational exposure to
trichloroethylene. Ind Health 47:107-112.
Yang Q, Ito S, Gonzalez FJ. 2007. Hepatocyte-restricted constitutive activation of PPAR alpha
induces hepatoproliferation but not hepatocarcinogenesis. Carcinogenesis. 28(6):
1171-1177.
Yeh MM, Larson AM, Campbell JS, Fausto N, Rulyak SJ, Swanson PE. 2007. The expression
of transforming growth factor -a in cirrhosis, dysplastic nodules, and hepatocellular
carcinoma; an immunohistochemical study in 70 cases. Am J Surg Pathol 31(5):
681-689.
Zeppa P, Benincasa G, Troncone G, Lucariello A, Zabatta A, Cochand-Priollet B, Fulciniti F,
Vetrani A, De Rosa G, Palombini L. 1998. Retrospective evaluation of DNA ploidy of
hepatocarcinoma on cytological samples. Diagnostic Cytopathology 19(5): 323-329.
Zhao Y, Krishnadasan A, Kennedy N, Morgenstern H, Ritz B. 2005. Estimated effects of
solvents and mineral oils on cancer incidence and Mortality in a cohort of aerospace
workers. Am J Ind Med 48:249-258.
Zhou, YC; Waxman, DJ. (1998) Activation of peroxisome proliferator-activated receptors by
chlorinated hydrocarbons and endogenous steroids. Environ Health Perspect 106(Suppl
4):983-988.
6/22/2009
609
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.5 Immunotoxicity and Cancers of the Immune System
Chemical exposures may result in a variety of adverse immune-related effects, including
immunosuppression (decreased host resistance), autoimmunity, and allergy-hypersensitivity, and
may result in specific diseases such as infections, systemic or organ-specific autoimmune
diseases, or asthma. Measures of immune function (e.g., T-cell counts, immunoglobulin (Ig) E
levels, specific autoantibodies, cytokine levels) may provide evidence of altered an immune
response that precedes the development of clinically expressed diseases. The first section of this
chapter discusses effects relating to immunotoxicity, including risk of autoimmune diseases,
allergy and hypersensitivity, measures of altered immune response, and lymphoid cancers.
Studies pertaining to effects in humans are presented first, followed by a section discussing
relevant studies in animals. The second section of this chapter discusses evidence pertaining to
trichloroethylene in relation to lymphoid tissue cancers, including childhood leukemia.
4.5.1 Human Studies
4.5.1.1 Noncattcer Immune-Related Effects
4.5.1.1.1 Immunosuppression, Asthma, and Allergies
In 1982, Lagakos et al. conducted a telephone survey of residents of Woburn,
Massachusetts, collecting information on residential history and history of 14 types of medically
diagnosed conditions (Lagakos, 1986). The survey included 4,978 children born since 1960 who
lived in Woburn before age 19. Completed surveys were obtained from approximately 57% of
the town residences with listed phone numbers. Two of the wells providing the town's water
supply from 1964 to 1979 had been found to be contaminated with a number of solvents,
including tetrachloroethylene (21 ppb) and trichloroethylene (267 ppb) [as cited in (Lagakos,
1986)]. Lagakos et al. used information from a study by the Massachusetts Department of
Environmental Quality and Engineering to estimate the contribution of water from the two
contaminated wells to the residence of each participant, based on zones within the town
receiving different mixtures of water from various wells, for the period in which the
contaminated wells were operating. This exposure information was used to estimate a
cumulative exposure based on each child's length of residence in Woburn. A higher cumulative
exposure measure was associated with conditions indicative of immunosuppression (e.g.,
bacterial or viral infections) or hypersensitivity (e.g., asthma). In contrast, a recent study using
the National Health and Nutrition Examination Survey (NHANES) data collected from
6/22/2009
610
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1999-2000 in a representative sample of the U.S. population (n = 550) did not find an
association between a TCE exposure and self-report of a history of physician-diagnosed asthma
(OR 0.94, 95% CI 0.77, 1.14) (Arif and Shah, 2007). TCE exposure, as well as exposure to 9
other volatile organic compounds, was determined through a passive monitor covering a period
of 48-72 hours. No clear trend was seen with self-reported wheeze episodes (OR 1.29, 95% CI
(0.98, 1.68) for 1-2 episodes; OR 0.21, 95% CI 0.04, 10.05) for 3 or more episodes in the past
12 months).
Allergy and hypersensitivity, as assessed with measures of immune system parameters or
immune function tests (e.g., atopy) in humans, have not been extensively studied with respect to
the effects of trichloroethylene (Table 4.5-1). Lehmann et al. reported data pertaining to IgE
levels and response to specific antigens in relation to indoor levels of volatile organic compounds
among children (age 36 months) selected from a birth cohort study in Leipzig, Germany
(Lehmann et al., 2001). Enrollment into the birth cohort occurred between 1995 and 1996. The
children in this allergy study represent a higher-risk group for development of allergic disease,
with eligibility criteria that were based on low birth weight (between 1,500 and 2,500 g), or cord
blood IgE greater than 0.9 kU/L with double positive family history of atopy. These eligibility
criteria were met by 429 children; 200 of these children participated in the allergy study
described below, but complete data (IgE and volatile organic compound measurements) were
available for only 121 of the study participants. Lehmann et al. measured 26 volatile organic
compounds via passive indoor sampling in the child's bedroom for a period of 4 weeks around
the age of 36 months. The median exposure of trichloroethylene was 0.42 (J,g/m3 (0.17 |ig/m3
and 0.87 (j,g/m3 for the 25th and 75th percentiles, respectively). Blood samples were taken at the
36-month-study examination and were used to measure the total IgE and specific IgE antibodies
directed to egg white, milk, indoor allergens (house dust mites, cat, molds), and outdoor
allergens (timothy-perennial grass, birch- tree). There was no association between
trichloroethylene exposure and any of the allergens tested in this study, although some of the
other volatile organic compounds (e.g., toluene, 4-ethyltoluene) were associated with elevated
total IgE levels and with sensitization to milk or eggs.
4.5.1.1.2 Generalized hypersensitivity skin diseases, with or without hepatitis
Occupational exposure to trichloroethylene has been associated with a severe,
generalized skin disorder that is distinct from contact dermatitis in the clinical presentation of the
skin disease (which often involves mucosal lesions), and in the accompanying systemic effects
that can include lymphadenopathy, hepatitis, and other organ involvement. Kamijima et al.
recently reviewed case reports describing 260 patients with trichloroethylene-related generalized
skin disorders (Kamijima et al., 2007). Six of the patients were from the United States or
6/22/2009
611
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Europe, with the remainder occurring in China, Singapore, Philippines, and other Asian
countries. One study in Guangdong province, in southeastern China, included more than 100 of
these cases in a single year (Huang et al., 2002). Kamijima et al. categorized the case
descriptions as indicative of hypersensitivity syndrome (n = 124) or a variation of erythema
multiforme, Stevens-Johnson syndrome, and toxic epiderma necrolysis (n = 115), with 21 other
cases unclassified in either category. The fatality rate, approximately 10%, was similar in the
two groups, but the prevalence of fever and lymphadenopathy was higher in the hypersensitivity
syndrome patients. Hepatitis was seen in 92%-94% of the multiforme, Stevens-Johnson
syndrome, and toxic epiderma necrolysis patients, but the estimates within the hypersensitivity
syndrome group were more variable (46%-94%) (Kamijima et al., 2007).
Some of the case reports reviewed by Kamijima et al. provided information on the total
number of exposed workers, working conditions, and measures of exposure levels. From the
available data, generalized skin disease within a worksite occurred in 0.25% to 13% of workers
in the same location, doing the same type of work (Kamijima et al., 2007). The measured
concentration of trichloroethylene ranged from < 50 mg/m3 to more than 4,000 mg/m3, and
exposure scenarios included inhalation only and inhalation with dermal exposures. Disease
manifestation generally occurred within 2-5 weeks of initial exposure, with some intervals up to
3 months. Most of the reports were published since 1995, and the geographical distribution of
cases reflects the newly industrializing areas within Asia.
Kamijima and colleagues recently conducted an analysis of urinary measures of
trichloroethylene metabolites (trichloroacetic acid and trichloroethanol) in 25 workers
hospitalized for hypersensitivity skin disease in 2002 (Kamijima et al., 2008). Samples taken
within 15 days of the last exposure to trichloroethylene exposure were available for 19 of the 25
patients, with a mean time of 8.4 days. Samples from the other patients were not used in the
analysis because the half life of urinary trichloroacetic acid is 50-100 hours. In addition, 3-6
healthy workers doing the same type of work in the factories of the affected worker, and 2
control workers in other factories not exposed to trichloroethylene were recruited in 2002-2003
for a study of breathing zone concentration of volatile organochlorines and urinary measures of
trichloroethylene metabolites. Worksite measures of trichloroethylene concentration were also
obtained. Adjusting for time between exposure and sample collection, mean urinary
concentration at the time of last exposure among the 19 patients was 206 mg/mL for
trichloroacetic acid. Estimates for trichloroethanol were not presented because of the shorter
half-life for this compound. Urinary trichloroacetic acid levels in the healthy exposed workers
varied among the 4 factories, with means (± SD) of 41.6 (± 18.0), 131 (± 90.2), 180 (± 92), and
395 (± 684). The lower values were found in a factory in which the degreasing machine had
been partitioned from the workers after the illnesses had occurred. Trichloroethylene
6/22/2009
612
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
concentrations (personal time weighed averages) at the factories of the affected workers ranged
from 164-2,330 mg/m3 (30-431 ppm). At the two factories with no affected workers in the past
3 years, the mean personal time weighted average trichloroethylene concentrations were 44.9
mg/m3 (14 ppm) and 1,803 mg/m3 (334) ppm). There was no commonality of additives or
impurities detected among the affected factories that could explain the occurrence of the
hypersensitivity disorder.
To examine genetic influences on disease risk, Dai et al. conducted a case-control study
of 111 patients with trichloroethylene-related severe generalized dermatitis and 152
trichloroethylene-exposed workers who did not develop this disease (Dai et al., 2004). Patients
were recruited from May 1999 to November 2003 in Guangdong Province, and were employed
in approximately 80 electronic and metal-plating manufacturing plants. Initial symptoms
occurred within 3 months of exposure. The comparison group was drawn from the same plants
as the cases, and had worked for more than 3 months without development of skin or other
symptoms. Mean age in both groups was approximately 23 years. A blood sample was obtained
from study participants for genotyping of tumor necrosis factor (TNF)-a, TNF-P, and interleukin
(IL)-4 genotypes. The genes were selected based on the role of TNF and of interleukin-4 in
hypersensitivity and inflammatory responses. The specific analyses included two
polymorphisms in the promoter region of TNF-a (G —~ A substitution at position -308)
designated as TNF All, with wildtype designated TNFAI; and aG->A substitution at position -
238), a polymorphism at the first intron on TNF-P, and a polymorphism in the promoter region
of IL-4 (C —~ T substitution at -590). There was no difference in the frequency of the TNF-a"238,
TNF-P, or IL-4 polymorphisms between cases and controls, but the wildtype TNF-a"308 genotype
was somewhat more common among cases (TNF A I/I genotype 94% in cases and 86% in
controls).
Kamijima et al. note the similarities, particular with respect to specific skin
manifestations, of the case presentations of trichloroethylene-related generalized skin diseases to
conditions that have been linked to specific medications (e.g., carbamezepine, allupurinol,
antibacterial sulfonamides), possibly in conjunction with reactivation of specific latent herpes
viruses (Kamijima et al., 2007). A previous review by these investigators discusses insights with
respect to drug metabolism that may be useful in developing hypotheses regarding susceptibility
to trichloroethylene-related generalized skin disorders (Nakajima et al., 2003). Based on
consideration of metabolic pathways and intermediaries, variability in CYP2E1, UDP-
glucoronyltransferase, glutathione-S transferase, and N-acetyl transferase (NAT) activities could
be hypothesized to affect the toxicity of trichloroethylene. NAT2 is most highly expressed in
liver, and the "slow" acetylation phenotype (which arises from a specific mutation) has been
associated with adverse effects of medications, including drug-induced lupus (Lemke and
6/22/2009
613
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
McQueen, 1995) and hypersensitivity reactions (Spielberg, 1996). There are limited data
pertaining to genetic or other sources of variability in these enzymes on risk of trichloroethylene-
related generalized skin diseases, however. In a study in Guangdong province, CYP1A1,
GSTM1, GSTP1, GSTT1, and NAT2 genotypes in 43 cases of trichloroethylene-related
generalized skin disease were compared to 43 healthy trichloroethylene-exposed workers (Huang
et al., 2002). The authors reported that the NAT2 slow acetylation genotype was associated with
disease, but the data pertaining to this finding was not presented.
4.5.1.1.3 Cytokine profiles
Cytokines are produced by many of the immune regulatory cells (e.g., macrophages,
dendritic cells), and have many different effects on the immune system. The Thl cytokines, are
characterized as "pro-inflammatory" cytokines, and include TNF-a and interferon (IFN)-y).
Although this is a necessary and important part of the innate immune response to foreign
antigens, an aberrant pro-inflammatory response may result in a chronic inflammatory condition
and contribute to development of scarring or fibrotic tissue, as well as to autoimmune diseases.
Th2 cytokines are important regulators of humoral (antibody-related) immunity. IL-4 stimulates
production of IgE and thus influences IgE-mediated effects such as allergy, atopy, and asthma.
Th2 cytokines can also act as "brakes" on the inflammatory response, so the balance between
different types of cytokine production is also important with respect to risk of conditions
resulting from chronic inflammation. Several studies have examined cytokine profiles in relation
to occupational or environmental TCE exposure (Table 4.5-2).
The 2001 Lehmann et al. study of 36 month old children (described above) also included
a blood sample taken at the 3-year study visit, which was used to determine the percentages of
specific cytokine producing T cells in relation to the indoor volatile organic compounds
exposures measured at birth. There was no association between trichloroethylene exposure and
either IL-4 CD3+ or IFN-y CD8+ T cells (Lehmann et al., 2001).
Another study by Lehmann et al. examined the relationship between indoor exposures to
volatile organic compounds and T-cell subpopulations measured in cord blood of newborns
(Lehmann et al., 2002). The study authors randomly selected 85 newborns (43 boys and 42
girls) from a larger cohort study of 997 healthy, full-term babies, recruited between 1997 and
1999 in Germany. Exclusion criteria included a history in the mother of an autoimmune disease
or infectious disease during the pregnancy. Twenty-eight volatile organic compounds were
measured via passive indoor sampling in the child's bedroom for a period of 4 weeks after the
birth (a period which is likely to reflect the exposures during the prenatal period close to the time
of delivery). The levels were generally similar or slightly higher than the levels seen in the
previous study using samples from the bedrooms of the 36-month-old children. The highest
6/22/2009
614
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
levels of exposure were seen for limonene (median 24.3 (j,g/m3), a-pinene (median 19.3 (J,g/m3)
and toluene (median 18.3 |ig/m3), and the median exposure of trichloroethylene was 0.6 [j,g/m3
(0.2 [j,g/m3 and 1.0 |ig/m3 for the 25th and 75 percentiles, respectively). Flow cytometry was
used to measure the presence of CD3 T-cells obtained from the cord blood labeled with
antibodies against IFN-y, tumor necrosis factor-a, IL-2 and IL-4. There was some evidence of a
decreased level of IL-2 with higher trichloroethylene exposure in the univariate analysis, with
median percentage of IL-2 cells of 46.1% and 33.0% in the groups that were below the 75th
percentile and above the 75th percentile of trichloroethylene exposure, respectively. In analyses
adjusting for family history of atopy, gender and smoking history of the mother during
pregnancy, there was little evidence of an association with either IL-2 or IFN-y, but there was a
trend of increasing trichloroethylene levels associated with decreased IL-4 and increased IFN-y.
Iavicoli et al. examined cytokine levels in 35 trichloroethylene-exposed workers (Group
A) from a printing area of a factory in Italy. Their work involved use of trichloroethylene in
degreasing (Iavicoli et al., 2005). Two comparison groups were included. Group B consisted of
30 other factory workers who were not involved in degreasing activities and did not work near
this location, and Group C consisted of 40 office workers at the factory. All study participants
were male and had worked at their present position for at least 3 years, and all were considered
healthy. Personal breathing zone air samples from 4 workers in Group A and 4 workers in
Group B were obtained in three consecutive shifts (24 total samples) to determine air
concentration of trichloroethylene. A urine sample was obtained from each Group A and Group
B worker (end of shift at end of work week) for determination of trichloroacetic acid
concentrations (corrected for creatinine), and blood samples were collected for assessment of IL-
2, IL-4, and IFN-y concentrations in serum using enzyme-linked immunosorbent assays. Among
exposed workers, the mean trichloroethylene concentration was approximately 35 mg/m3 (30.75
± sd 9.9, 37.75 ± 23.0, and 36.5 ± 8.2 mg/m3 in the morning, evening, and night shifts,
respectively). The urinary trichloroacetic acid concentrations were much higher in exposed
workers compared with nonexposed workers (mean ± sd, Group A 13.3 ± 5.9 mg/g creatinine;
Group B 0.02 ± 0.02 mg/g creatinine). There was no difference in cytokine levels between the
two control groups, but the exposed workers differed significantly (all p-values < 0.01 using
Dunnett's test for multiple comparisons) from each of the two comparison groups. The observed
differences were a decrease in IL-4 levels (mean 3.9, 8.1, and 8.1 pg/mL for groups A, B, and C,
respectively), and an increase in IL-2 levels (mean 798, 706, and 730 pg/mL for groups A, B, and
C, respectively) and in IFN-y levels (mean 37.1, 22.9, and 22.8 pg/mL for groups A, B, and C,
respectively).
The available data from these studies (Lehmann et al., 2001, 2002; Iavicoli et al., 2005)
provide some evidence of an association between increased trichloroethylene exposure
6/22/2009
615
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
modulation of immune response involving an increase in pro-inflammatory cytokines (IL-2, IFN-
y) and a decrease in Th2 (allergy-related) cytokines (e.g., IL-4). These observations add support
to the influence of trichloroethylene in immune-related conditions affected by chronic
inflammation.
6/22/2009
616
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-1. Studies of immune parameters (IgE antibodies and cytokines) and trichloroethylene in humans
Parameter, Reference, Location, Diagnosis
Source of Data Results Period, Sample Size, Age
IgE antibodies Lehmann et al., 2001
blood sample, indoor air Trichloroethylene exposure not associated with sensitization to indoor Germany 1997-1999. n = 121 36-
sampling of 28 volatile organic or outdoor allergens month old children
chemicals in child's bedroom
Cytokine secreting CD3+ T cell
populations
cord blood, indoor air sampling
of 28 volatile organic chemicals
in child's bedroom 4 weeks after
birth
In CD3+ cord blood cells, some evidence of association between
increasing trichloroethylene levels and
decreased IL-4 > 75th percentile OR 0.6 (95% CI 0.2, 2.1),
< 25th percentile OR 4.4 (95% CI 1.1, 17.8)
increased IFN-y > 75th percentile OR 3.6 (95% CI 0.9, 14.9)
< 25th percentile OR 0.7 (95% CI 0.2, 2.2)
Similar trends not seen with tumor necrosis factor-a or IL-2
Cytokine secreting CD3+ and
CD8+ T cell populations
blood sample, indoor air
sampling of 28 volatile organic
chemicals in child's bedroom
Cytokine concentration - serum
urine sample (trichloroacetic acid Non-exposed workers similar to office controls for all cytokine
concentration), blood sample, measures. Compared to non-exposed workers, the trichloroethylene
questionnaire (smoking history, exposed workers had:
age, residence), workplace TCE decreased IL-4 (mean 3.9 versus 8.1 pg/mL)
Trichloroethylene exposure not associated with percentages of IL-4
CD3+ or IFN-y CD8+ T cells
Lehmann et al., 2002
Germany. 1995-1996. n = 85
newborns
Lehmann et al., 2001
Germany. 1995-1999. n = 200 36-
month old children.
Iavicoli et al., 2005
Italy, n = 35 printers using TCE, 30
non-exposed workers (in same
factory, did not use or were not near
TCE), 40 office worker controls.
6/22/2009
617
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-1. Studies of immune parameters (IgE antibodies and cytokines) and trichloroethylene in humans
Parameter,
Reference, Location, Diagnosis
Source of Data
Results
Period, Sample Size, Age
measures (personal samples, 4
increased IL-2 (mean 798 versus 706 pg/mL)
All men. Mean age -33 years.
exposed and 4 non-exposed
increased IFN-y (mean 37.1 versus 22.9 pg/mL)
workers)
6/22/2009
618
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.5.1.1.4 Autoimmune disease
4.5.1.1.4.1 Disease Clusters and Geographic-Based Studies
Reported clusters of diseases have stimulated interest in environmental influences on
systemic autoimmune diseases. These descriptions include investigations into reported clusters
of systemic lupus erythematosus (Balluz et al., 2001; Dahlgren et al., 2007) and Wegener
granulomatosis (Albert et al., 2005). Wegener granulomatosis, an autoimmune disease involving
small vessel vasculitis, usually with lung or kidney involvement, is a very rare condition, with an
incidence rate of 3-14 per million per year (Mahr et al., 2006). Trichloroethylene was one of
several ground water contaminants identified in a recent study investigating a cluster of 7 cases
of Wegener granulomatosis around Dublin, Pennsylvania. Because of the multiple contaminants,
it is difficult to attribute the apparent disease cluster to any one exposure.
In addition to the study of asthma and infectious disease history among residents of
Woburn, Massachusetts (Lagakos, 1986) (see section 4.5.1.1.1), Byers et al. provide data
pertaining to immune function from 23 family members of leukemia patients in Woburn,
Massachusetts (Byers et al., 1988). Serum samples were collected in May and June of 1984 and
in November of 1985 (several years after 1979, when the contaminated wells had been closed).
Total lymphocyte counts and lymphocyte subpopulations (CD3, CD4, and CD8) and the
CD4/CD8 ratio were determined in these samples, and in samples from a combined control
group of 30 laboratory workers and 40 residents of Boston selected through a randomized
probability area sampling process. The study authors also assessed the presence of antinuclear
antibodies (ANA) or other autoantibodies (antismooth muscle, antiovarian, antithyroglobulin,
and antimicrosomal antibodies) in the family member samples and compared the results with
laboratory reference values. The age distribution of the control group, and stratified analyses by
age, are not provided. The lymphocyte subpopulations were higher and the CD4/CD8 ratio was
lower in the Woburn family members compared to the controls in both of the samples taken in
1984. In the 1985 samples, however, the subpopulation levels had decreased and the CD4/CD8
ratio had increased; the values were no longer statistically different from the controls. None of
the family member serum samples had antithyroglobulin or antimicrosomal antibodies, but 10
family member serum samples (43%) had ANA (compared to < 5% expected based on the
reference value). Because the initial blood sample was taken in 1984, it is not possible to
determine the patterns at a time nearer to the time of the exposure. The coexposures that
occurred also make it difficult to infer the exact role of trichloroethylene in any alterations of the
immunologic parameters.
6/22/2009
619
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Kilburn and Warshaw reported data from a study of contamination by metal-cleaning
solvents (primarily trichloroethylene) and heavy metals (e.g., chromium) of the aquifer of the
Santa Cruz River in Tucson, Arizona (Kilburn and Warshaw, 1992). Exposure concentrations
above 5 ppb (6-500 ppb) had been documented in some of the wells in this area. A study of
neurological effects was undertaken between 1986 and 1989 (Kilburn and Warshaw, 1993), and
two of the groups within this larger study were also included in a study of symptoms relating to
systemic lupus erythematosus. Residents of Tucson (n = 362) were compared to residents of
southwest Arizona (n = 158) recruited through a Catholic parish. The Tucson residents were
selected from the neighborhoods with documented water contamination (>5ppb trichloroethylene
for at least one year between 1957 and 1981). Details of the recruitment strategy are not clearly
described, but the process included recruitment of patients with lupus or other rheumatic diseases
(Kilburn and Warsaw 1993, 1992). The prevalence of some self-reported symptoms (malar rash,
arthritis/arthalgias, Raynaud syndrome, skin lesions, and seizure or convulsion was significantly
higher in Tucson, but there was little difference between the groups in the prevalence of oral
ulcers, anemia, low white blood count or low platelet count, pleurisy, alopecia, or proteinuria.
The total number of symptoms reported was higher in Tucson than in the other southwest
Arizona residents (14.3% versus 6.4% reported four or more symptoms, respectively). Low-titer
(1:80) ANA were seen in 10.6% and 4.7% of the Tucson and other Arizona residents,
respectively (p = 0.013). However, since part of the Tucson study group was specifically
recruited based on the presence of rheumatic diseases, it is difficult to interpret these results.
4.5.1.1.4.2 Case-Control Studies
Interest in the role of organic solvents, including trichloroethylene, in autoimmune
diseases was spurred by the observation of a scleroderma-like disease characterized by skin
thickening, Raynaud's phenomenon, and acroosteolysis and pulmonary involvement in workers
exposed to vinyl chloride (Gama and Meira, 1978). A case report in 1987 described the
occurrence of a severe and rapidly progressive case of systemic sclerosis in a 47 year old woman
who had cleaned x-ray tubes in a tank of trichloroethylene for approximately 2.5 hours (Lockey
etal., 1987).
One of the major impediments to autoimmune disease research is the lack of disease
registries, which make it difficult to identify incident cases of specific diseases (NIAMS, 2007).
There are no cohort studies of the incidence of autoimmune diseases in workers exposed to
trichloroethylene. Most of the epidemiologic studies of solvents and autoimmune disease rely on
general measures of occupational exposures to solvents, organic solvents, or chlorinated solvents
exposures. A two- to three-fold increased risk of systemic sclerosis (scleroderma) (Aryal et al.,
6/22/2009
620
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
2001; Garabrant et al., 2003; Maitre et al., 2004), rheumatoid arthritis (Lundberg et al., 1994;
Sverdrup et al., 2005), undifferentiated connective tissue disease (Lacey et al., 1999), and
antineutrophil-cytoplasmic antibody (ANCA)-related vasculitis (Beaudreuil et al., 2005; Lane et
al., 2003) has generally been seen in these studies, but there was little evidence of an association
between solvent exposure and systemic lupus erythematosus in two recent case-control studies
(Cooper et al., 2004; Finckh et al., 2006).
Two case-control studies of scleroderma (Bovenzi et al., 2004; Maitre et al., 2004) and
two of rheumatoid arthritis (Olsson et al., 2004, 2000) provide data concerning solvent exposure
that occurred among metal workers or in jobs that involved cleaning metal (i.e., types of jobs
which were likely to use trichloroethylene as a solvent). There was a two-fold increased risk
among male workers in the two studies of rheumatoid arthritis from Sweden (Olsson et al., 2004,
2000). The results from the smaller studies of scleroderma were more variable, with no exposed
cases seen in one study with 93 cases and 206 controls (Maitre et al., 2004), and an odds ratio of
5.2 (95% CI 0.7, 37) seen in a study with 56 cases and 171 controls (Bovenzi et al., 2004).
Five other case-control studies provide data specifically about trichloroethylene exposure,
based on industrial hygienist review of job history data (Table 4.5-1). Three of these studies are
of scleroderma (Diot et al., 2002; Garabrant et al., 2003; Nietert et al., 1998), one is of
undifferentiated connective tissue disease (Lacey et al., 1999), and one is of small vessel
vasculitidies involving antineurophil-cytoplasmic antibodies (ANCA) (Beaudreuil et al., 2005).
These studies included some kind of expert review of job histories, but only two studies included
a quantification of exposure (e.g., a cumulative exposure metric, or a "high" exposure group)
(Diot et al., 2002; Nietert et al., 1998). Most of the studies present data stratified by sex, and as
expected, the prevalence of exposure (either based on type of job or on industrial hygienist
assessment) is considerably lower in women compared with men. In men the studies generally
reported odds ratios between 2.0 and 8.0, and in women, the odds ratios were between 1.0 and
2.0. The incidence rate of scleroderma in the general population is approximately 5-10 times
higher in women compared with men, which may make it easier to detect large relative risks in
men.
The EPA conducted a meta-analysis of the three scleroderma studies with specific
measures of trichloroethylene (Diot et al., 2002; Garabrant et al., 2003; Nietert et al., 1998),
examining separate estimates for males and for females. The resulting combined estimate for
"any" exposure, using a random effects model to include the possibility of non-random error
between studies (DerSimonian and Laird, 1986), was OR = 2.5 (95% CI 1.1, 5.4) for men and
OR =1.2 (95%) CI 0.58, 2.6) in women. (Because the "any" exposure variable was not included
in the published report, Dr. Paul Nietert provided the EPA with a new analysis with these results,
e-mail communication from Paul Nietert to Glinda Cooper, November 28, 2007.)
6/22/2009
621
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Specific genes may influence the risk of developing autoimmune diseases, and genes
involving immune response (e.g., cytokines, major histocompatibility complex, B and T cell
activation) have been the focus of research pertaining to the etiology of specific diseases. The
metabolism of specific chemical exposures may also be involved(Cooper et al., 1999). Povey et
al. (2001) examined polymorphisms of two cytochrome P450 genes, CYP2E1 and CYP2C19, in
relation to solvent exposure and risk of developing scleroderma. These specific genes were
examined because of their hypothesized role in metabolism of many solvents, including
trichloroethylene. Seven scleroderma patients who reported a history of solvent exposure were
compared to 71 scleroderma patients with no history of solvent exposure and to 106 population-
based controls. The CYP2E1*3 allele and the CYP2E1*4 allele were more common in the 7
solvent-exposed patients (each seen in 2 of the 7 patients; 29%) than in either of the comparison
groups (approximately 5% for CYP2E1*3 and 14% for CYP2E1*4). The authors present these
results as observations that require a larger study for corroboration and further elucidation of
specific interactions.
6/22/2009
622
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-2. Case-control studies of autoimmune diseases with measures of trichloroethylene exposure
Disease, Source of Data
Results:
Exposure Prevalence, Odds Ratios (OR), 95% Confidence Intervals
Reference, Location, Sample
Size, Age
Scleroderma
Structured interview (specific
jobs and materials; jobs held 1
or more years). Exposure
classified by self-report and by
expert review (job exposure
matrix).
Men
Maximum intensity 30% cases, 10% controls OR 3.3 (1.0, 10.3)
Cumulative intensity 32% cases, 21% controls OR 2.0 (0.7, 5.3)
Maximum probability 16% cases, 3% controls OR 5.1 (not calculated)
Women:
Maximum intensity 6% cases, 7% controls OR 0.9 (0.3, 2.3)
Cumulative intensity 10% cases, 9% controls OR 1.2 (0.5, 2.6)
Maximum probability 4% cases, 5% controls OR 0.7 (0.2, 2.2)
Nietert et al., 1998
South Carolina. Prevalent cases,
178 cases (141 women, 37 men),
200 hospital-based controls. Mean
age at onset 45.2 years.
or more months). Exposure
classified by expert review
Structured interview (specific Men and women
jobs and materials; jobs held 6 any exposure: cases 16%, controls 8% OR 2.4 (95% CI 1.0, 5.4)
high exposure:3 cases 9%, controls 1% OR 7.6 (95% CI 1.5, 37.4)
Men
any exposure: cases 64%, controls 27% OR 4.7 (95% CI 0.99, 22.0
Women
any exposure: cases 9%, controls 4% OR 2.1 (95% CI 0.65, 6.8)
Structured interview (specific Women
jobs and materials; jobs held 3 Self report: cases 1.3%, controls 0.7% OR 2.0 (95% CI 0.8, 4.8)
or more months). Exposure Expert review: cases 0.7%, controls 0.4% OR 1.9 (95% CI 0.6, 6.6)
classified by self-report and by
expert review
Diot et al., 2002
France. Prevalent cases, 80 cases
(69 women, 11 men), 160 hospital
controls. Mean age at diagnosis
48 years.
Garabrant et al., 2003
Michigan and Ohio. Prevalent
cases, 660 cases (all women),
2,227 population controls.13 Ages
18 and older.
6/22/2009
623
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-2. Case-control studies of autoimmune diseases with measures of trichloroethylene exposure
Results: Reference, Location, Sample
Disease, Source of Data Exposure Prevalence, Odds Ratios (OR), 95% Confidence Intervals Size, Age
Undifferentiated connective tissue disease
Structured interview (specific Lacey et al., 1999
jobs and materials; jobs held Women Michigan and Ohio. Prevalent
3 or more months). Exposure Self report: cases 0.5%, controls 0.7% OR 0.88 (95% CI 0.11, 6.95) cases, 205 cases (all women),
classified by self-report and Expert review: cases 0.5%, controls 0.4% OR 1.67 (95% CI 0.19, 14.9) 2,095 population controls,
by expert review. Ages 18 and older.
ANCA-related diseases0
Men and women (data not presented separately by sex) Beaudreuil et al., 2005
Structured interview (specific
cases 18.3%, controls 17.5% OR 1.1 (0.5, 2.4) France. Incident cases, 60 cases
jobs and materials; jobs held
(~ 50% women), 120 hospital
6 or more months). Exposure
controls. Mean age 61 years.
classified by expert review.
a Cumulative exposure defined as product of probability x intensity x frequency x duration scores, summed across all jobs; scores
of >1 classified as "high".
b Total n; n with TCE data: self -report 606 cases, 2,138 control; expert review 606 cases, 2,137 controls.
c ANCA = antineutrophil-cytoplasmic antibody. Diseases included Wegener glomerulonephritis (n = 20), microscopic
polyangiitis (n = 8), pauci-immune glomerulonephritis (n = 10), uveitis (n = 6), Churg-Strauss syndrome (n = 4), stroke
(n = 4) and other diseases (no more than 2 each).
6/22/2009
624
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.5.1.2 Cancers of the Immune System, Including Childhood Leukemia
4.5.1.2.1 Description of Studies
Human studies have reported cancers of the immune system resulting from TCE
exposure. Lymphoid tissue neoplasms arise in the immune system and result from events that
occur within immature lymphoid cells in the bone marrow or peripheral blood (leukemias), or
more mature cells in the peripheral organs (non-Hodgkin lymphoma or NHL). As such, the
distinction between lymphoid leukemia and NHL is largely distributional with overlapping
entities, such that a particular lymphoid neoplasm may manifest both lymphomatous and
leukemic features during the course of the disease (Weisenberger, 1992). Lymphomas are
grouped according to the World Health Organization classification as B-cell neoplasms, T-
cell/NK-cell neoplasms, and Hodgkin's lymphoma, formerly known as Hodgkin's disease
(Harris et al., 2000).
Numerous studies are found in the published literature on lymphoma and either broad
exposure categories or occupational title. Most of these studies evaluate NHL, specifically. The
NHL studies generally report positive associations with organic solvents or job title as aircraft
mechanic, metal cleaner or machine tool operator, and printers, although associations are not
observed consistently across all studies, specific solvents are not identified, and different
lymphoma classifications are adopted (Alexander et al., 2007; Blair et al., 1993; Boffetta and de
Vocht, 2007; Chiu and Weisenburger, 2003; Dryver et al., 2004; Figgs et al., 1995;
Karunanayake et al., 2008; Lynge et al., 1997; Richardson et al., 2008; Seidler et al., 2007;
Mannetje et al., 2008; Tatham et al., 1997; Vineis et al., 2007; Wang et al., 2009). Although a
major use of TCE is the degreasing of metal and other products with potential exposure in jobs in
the metal industry, printing industry and aircraft maintenance or manufacturing industry (Bakke
et al., 2007), job title as a surrogate for TCE exposure is uncertain for identifying hazard. One
study, a NHL case-control study of Perdue et al. (in press), examined degreasing tasks and
reported an increasing positive increasing trend between NHL risk and three degreasing exposure
surrogates, average frequency (hours/year), maximal frequency (hours/year), or cumulative
number of hours.
As described in Appendix B, the EPA conducted a thorough and systematic search of
published epidemiological studies of cancer risk and trichloroethylene exposure using the
PubMed bibliographic database. The EPA also requested unpublished data pertaining to
trichloroethylene from studies that may have collected this data but did not include it in their
published reports. ATSDR and state health department peer-reviewed studies were also
reviewed. Information from each of these studies relating to specified design and analysis
criteria was abstracted. These criteria included aspects of study design, representativeness of
6/22/2009
625
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
study subjects, participation rate/loss to follow-up, latency considerations, potential for biases
related to exposure misclassification, disease misclassification, and surrogate information,
consideration of possible confounding, and approach to statistical analysis. All studies are
considered for hazard identification but those studies more fully meeting the objective criteria
provided the greater weight for identifying a cancer hazard.
The body of evidence on lymphoma and trichloroethylene is comprised of occupational
cohort studies, population-based case-control studies and geographic studies. Four case-control
studies and four geographic studies also examine childhood leukemia and trichloroethylene.
Most studies report observed risk estimates and associated confidence intervals for non-Hodgkin
lymphoma for overall TCE exposure. Fewer studies presented in published papers this
information for leukemia, cell-specific leukemia, or multiple myeloma.
The seven cohort studies with data on the incidence of lymphopoietic and hematopoietic
cancer in relation to trichloroethylene exposure range in size [803 (Hansen et al., 2001) to 86,868
(Chang et al., 2005)], and were conducted in Denmark, Sweden, Finland, Taiwan and the United
States (Table 4.5-3) (for additional study descriptions, see Appendix B). Some subjects in the
Hansen et al. study are also included in a study reported by Raaschou-Nielsen et al. (2003);
however, any contribution from the former to the latter are minimal given the large differences in
cohort sizes of these studies (Hansen et al., 2001; Raaschou-Nielsen et al., 2003). The exposure
assessment techniques used in all studies except Chang et al. (2005) included a detailed job
exposure matrix (Zhao et al., 2005; Blair et al., 1998) and biomonitoring data (Anttila et al.,
1995; Axelson et al., 1994; Hansen et al., 2001) with assignment of TCE exposure to individual
subjects. Subjects in Chang et al. (2005) have potential exposure to several solvents including
TCE; all subjects are presumed as "exposed" because of employment in the plant although
individual subjects would be expected to have differing exposure potentials. The lack of
attribution of exposure intensity to individual subjects in Chang et al. (2005) yields a greater
likelihood for exposure misclassification compared to the six other studies with exposure
assessment approaches supported by information on job titles, tasks, and industrial hygiene
monitoring data. Incidence ascertainment in two cohorts began 21 (Blair et al., 1998) and 38
years (Zhao et al., 2005) after the inception of the cohort. Specifically, Zhao et al. (2005) note
"results may not accurately reflect the effects of carcinogenic exposure that resulted in non-fatal
cancers before 1988." Because of the issues concerning case ascertainment raised by this
incomplete coverage, observations must be interpreted in light of possible bias reflecting
incomplete ascertainment of incident cases.
Fifteen cohort studies describing mortality risks from lymphopoietic and hematopoietic
cancer are summarized in Table 4.5-4 (for additional study descriptions, see Appendix B). Two
studies examined cancer incidence and are identified above (Blair et al., 1998; Zhao et al., 2005).
6/22/2009
626
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
In 8 of the 15 studies presenting mortality risks (Chang et al., 2003; Costa et al., 1989; Garabrant
et al., 1988; Henschler et al., 1995; Sinks et al., 1992; Wilcosky et al., 1984; ATSDR, 2004;
Clapp and Hoffman, 2008), a relatively limited exposure assessment methodology was used,
study participants may not represent the underlying population, or there was a low exposure
prevalence of TCE exposure. For reasons identified in the systematic review, these studies are
given less weight in the overall evaluation of the literature than the seven other cohort studies
that better met the ideals of evaluation criteria (Blair et al., 1998; Boice et al., 2006; Boice et al.,
1999; Greenland et al., 1994; Morgan et al., 1998; Ritz, 1999; Zhao et al., 2005).
Case-control studies of lymphoma or hairy cell leukemia [a lymphoma according to the
World Health Organization's lymphoma classification system (Morton et al., 2007, 2006) from
United States (Connecticut), Germany, Italy, Sweden, and Canada were identified, and are
summarized in Table 4.5-5 (for additional study descriptions, see Appendix B). These studies
identified cases from hospital records (Costantini et al., 2008; Hardell et al., 1994; Mester et al.,
2006; Miligi et al., 2006; Persson and Fredrikson, 1999; Seidler et al., 2007; Siemiatycki et al.,
1991); the Connecticut Tumor Registry (Wang et al., 2009); or the Swedish Cancer Registry
(Nordstrom et al., 1998), and population controls. These studies assign potential occupational
TCE exposure to cases and controls using self-reported information obtained from a mailed
questionnaire (Hardell et al., 1994; Nordstrom et al., 1998; Persson and Fredrikson, 1999) or
from direct interview with study subjects, with industrial hygienist ratings of exposure potential
and a job exposure matrix (Siemiatycki et al., 1991; Miligi et al., 2006; Seidler et al., 2007;
Costantini et al., 2008; Wang et al., 2009). Additionally, three of these large multiple center
lymphoma case-control studies examine specific types of NHL (Miligi et al., 2006; Seidler et al.,
2007; Wang et al., 2009) or leukemia (Costantini et al., 2008).
Four geographic based studies on lymphoma in adults are summarized in Table 4.5-6 (for
additional study descriptions, see Appendix B) and subjects in three studies are identified based
upon their residence in a community where TCE was detected in water serving the community
(Vartianen et al., 1993; Cohn et al., 1994; ATSDR, 2006). Both Cohn et al. (1994) and ATSDR
(2006) also present estimates for childhood leukemia and these observations are discussed below
with other studies reporting on childhood leukemia. A subject is assumed to have a probability
of exposure due to residence likely receiving water containing TCE. These studies do not
include statistical models of water distribution networks, which may influence TCE
concentrations delivered to a home, nor a subject's ingestion rate to estimate TCE exposure to
individual study subjects. Rather, one level of exposure to all subjects in a geographic area is
assigned, although there is some inherent measurement error and misclassification bias because
not all subjects are exposed uniformly.
6/22/2009
627
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
NHL risk is statistically significantly elevated in four high-quality studies [7.2, 95% CI:
1.3, 42 (Hardell et al., 1994); 3.1, 95% CI: 1.3, 6.1 (Hansen et al., 2001); 1.5, 95% CI: 1.2, 2.0,
subcohort with higher exposure (Raaschou-Nielsen et al., 2003), 2.1, 95% CI: 1.0, 4.8, >35 ppm-
years cumulative TCE exposure (Seidler et al., 2007)]. Two of these incidence studies report
statistically significantly associations for all lymphopoietic and hematopoietic cancer,
specifically NHL, for subjects with longer employment duration as a surrogate of TCE exposure
[^6.25 year, 4.2, 95% CI; 1.1, 11 (Hansen et al., 2001); >5 year, 1.6, 95% CI: 1.1, 2.2,
(Raaschou-Nielsen et al., 2003)]. Hansen et al. (2001) also examined cumulative exposure and
exposure intensity with estimated risk larger in low exposure groups than for high exposure
groups. Blair et al. (1998) observed a doubling of NHL mortality risk (SMR 2.0, 95% CI: 0.9,
4.5) in a cohort of aircraft maintenance workers with a stronger exposure assessment compared
to approaches adopted in the aerospace cohort studies of (Boice et al., 2006, 1999; Garabrant et
al., 1988; Morgan et al., 1998; Zhao et al., 2005) and the nested case-control study of Greenland
et al. (1994) where exposure misclassification and bias is more likely (NRC, 2006). The
association seen with TCE among men in Blair et al. (1998), all 8 deaths (RR = 2.3, 95% CI: 0.7,
7.5) was among the highest seen in the analyses of individual solvent exposures, and was higher
than the estimate for males with "any solvent" exposure (RR =1.6, 95% CI: 0.6, 4.1). NHL risk
among TCE exposure subjects in Blair et al. (1998) remained elevated but of a lower magnitude
(RR = 1.36, 95% CI: 0.77, 2.39) with an additional 10 years of follow-up (Radican et al., 2008).
Four high-quality population case-control studies observed a 10% to 50% increased risk
between NHL and any TCE exposure [1.1, 95% CI: 0.6, 2.3 (Siemiatycki, 1991); 1.5, 95% CI:
0.7, 3.3 (Nordstrom et al., 1998); 1.2, 95% CI: 0.5, 2.4 (Persson and Fredrikson, 1999); 1.2, 95%
CI: 0.9, 1.8 (Wang et al., 2009)]. Observed risks for overall TCE exposure in population case-
control studies are lower than those observed in cohort studies and this observation may argue
against association between TCE and NHL due to apparent inconsistency or heterogeneity.
However, a consequence of low exposure prevalence in population case-control studies is lower
average exposure compared to cohort studies, which assigned TCE exposure to individual study
subjects and lower expected risk.
Odds ratios are higher for diffuse NHL, primarily a B-cell lymphoma, than for all non-
Hodgkin lymphomas in both studies which examine forms of lymphoma (Miligi et al., 2006;
Seidler et al., 2007) (Table 4.5-6). Observations in the two other studies of B-cell lymphomas
(Persson and Fredrikson, 1999; Wang et al., 2009), appear consistent with Miligi et al. (2006)
and Seidler et al. (2007). Together, these observations suggest that the associations between
trichloroethylene and diffuse NHL are stronger than the associations seen with other forms of
lymphoma, and that disease misclassification may be introduced in studies examining
trichloroethylene and NHL as a broader category. Mortality observations in other occupational
6/22/2009
628
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
cohorts (Wilcosky et al., 1984; Garabrant et al., 1988; Costa et al., 1989; Ritz, 1999; Henschler
et al., 1995; Chang et al., 2003; ATSDR, 2004) included a risk estimate of 1.0 in 95% confidence
intervals; these studies neither add to nor detract from the overall weight of evidence given their
lower likelihood for TCE exposure due to inferior exposure assessment approaches, lower
prevalence of exposure, lower statistical power, and fewer exposed deaths.
6/22/2009
629
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-3. Incidence cohort studies of TCE exposure and lymphopoietic and hematopoietic cancer risk
Population
Exposure Group
Lymphopoietic
Cancer
Relative Risk
(95% CI)a
Non-Hodgkin
Lymphoma
Relative Risk
(95% CI)a
Leukemia
Relative Risk
(95% CI)a
Reference(s) and Study Description13
Aerospace workers (Rocketdyne), California
Any TCE exposure Not reported
Low cumulative TCE score
Medium cumulative TCE score
High cumulative TCE score
(p for trend)
Not reported
1.0 (referent)
0.88 (0.47, 1.65)
0.20 (0.03, 1.46)
(0.097)
Electronic workers (Taiwan)
All employees
0.67 (0.42, 1.01)
Males 0.73 (0.27, 1.60)
22
6 Not reported
Females 0.65 (0.37,1.05) 16 Not reported
Females
Blue-collar workers, Denmark
Any exposure
Subcohort w/higher exposured
Employment duration
1-4.9 years
> 5 years
1.1 (1.0, 1.6)
Not reported
229 1.2(1.0,1.5)
1.5(1.2,2.0)
1.5(1.1,2.1)
1.6(1.1,2.2)
28
16
1
Not reported
Not reported
0.78 (0.49, 1.17)
96 1.2 (0.9, 1.4)
65 Not reported
35
30
23
82
Zhao et al., 2005
n = 5,049 (2,689 with high cumulative
TCE exposure), began work before 1980,
worked at least 2 years, alive with no
cancer diagnosis in 1988, follow-up from
1988-2000, job exposure matrix
(intensity), internal referents (workers
with no TCE exposure). Leukemia
observations included in non-Hodgkin
lymphoma category
Chang et al., 2005; Sung et al., 2007
n = 88,868 (n = 70,735 female), follow-
up 1979-1997, does not identify TCE
exposure to individual subjects (Chang et
al., 2005)
n = 63,982 females, follow-up
1979-2001, dose not identify TCE
exposure to individual subjects (Sung et
al., 2007)
Raaschou-Nielsen et al., 2003
n = 40,049 (14,360 with presumed higher
level exposure to TCE), worked for at
least 3 months, follow-up from
1968-1997, documented TCE use0. EPA
based the lymphopoietic cancer category
on summation of ICD codes 200-204.
Biologically-monitored workers, Denmark
Any TCE exposure
2.0(1.1,3.3)
15 3.1(1.3,6.1)
2.0 (0.7, 4.4)
Hansen etal., 2001
6 n = 803, urinary-TCA or air TCE
6/22/2009
630
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Cumulative exp (Ikeda), males Not reported
<17 ppm-yr
>17 ppm-yr
Mean concentration (Ikeda),
males Not reported
<4 ppm
4+ ppm
Employment duration, males Not reported
< 6.25 yr
> 6.25 yr
Aircraft maintenance workers, Hill Air Force Base, Utah)
TCE Subcohort Not reported
Males, Cumulative exp
0 1.0 (referent)
< 5 ppm-yr 0.8 (0.4, 1.7)
5-25 ppm-yr 0.7(0.3,1.8)
>25 ppm-yr 1.4(0.6,2.9)
Females, Cumulative exp
0 1.0 (referent)
< 5 ppm-yr 1.2(0.3,4.4)
5-25 ppm-yr 1.9(0.4,8.8)
>25 ppm-yr 0.9(9.2,3.3)
36
12
7
17
3.9(0.8, 11)
3.1 (0.6, 9.1)
3.9(1.1, 10)
3.2(1.1, 10)
2.5 (0.3, 9.2)
4.2(1.1, 11)
Not reported
1.0 (referent)
0.9 (0.3,2.6)
0.7 (0.2, 2.6)
1.0 (0.4, 2.9)
1.0 (referent)
0.6 (0.1,5.0)
0.9 (0.2, 4.5)
Biologically-monitored workers, Finland 1.51 (0.92,2.33) 20 1.81 (0.78,3.56)
Mean air-TCE (Ikeda
extrapolation)
<6 ppm 1.36 (0.65,2.49) 10 2.01 (0.65,4.69)
6+ppm 2.08 (0.95,3.95) 9 1.40 (0.17,5.04)
Biologically-monitored workers, Sweden
Males, 2+years exposure 1.17(0.47,2.40) 7 1.56 (0.51,3.64)
duration
6/22/2009
631
Not reported
3
3
Not reported
4
4
Not reported
2
4
Not reported
19
1.0 (referent)
8 0.4(0.1,2.0)
4
7 0.9 (0.2, 3.7)
1.0 (referent)
1
0 2.4(0.3,21.8)
2
8 1.08 (0.35,2.53)
5 0.39 (0.01,2.19)
2 2.65 (0.72, 6.78)
samples, follow-up 1968-1996 (subset of
Raaschlou-Nielsen et al. [2003] cohort).
EPA based the lymphopoietic cancer
category on summation of ICD codes
200-204
Blair etal., 1998)
n = 10,461 men and 3,605 women (total n
1 = 14,066, n = 7,204 with TCE exposure),
employed at least 1 year from 1952 to
2 1956, follow-up 1973-1990, job exposure
0 matrix (intensity), internal referent
4 (workers with no chemical exposures)
0
1
0
5 Anttila et al., 1995
n = 3,089 men and women, urinary-TCA
samples, follow-up 1967-1992
1
4
5
Not reported
Axelson et al., 1994
n = 1,421 men and 249 women (total
1,670), urinary-TCA samples, follow-up
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
0-17 ppm (Ikeda extrapolation) Not reported
18-35 ppm (Ikeda extrapolation)
>36 ppm (Ikeda extrapolation)
Females Not reported
1.44 (0.30, 4.20) 3
(0,8.58) 0
6.25 (0.16,34.8) 1
Not reported
Not reported
Not reported
1958-1987. EPA based the
lymphopoietic cancer category includes
ICD-7 200-203; ICD-7 204 (leukemia)
not reported.
a n = number of observed cases.
b Standardized incidence ratios using an external population referent group unless otherwise noted.
0 Companies included iron and metal (48%), electronics (11%), painting (11%), printing (8%), chemical (5%), dry cleaning (5%), and other industries (13%).
Defined as at least 1 year duration and first employed before 1980.
6/22/2009
632
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-4. Mortality cohort studies of TCE exposure and lymphopoietic and hematopoietic cancer risk
Lymphopoietic
Cancer
Non-Hodgkin
Lymphoma
Leukemia
Population,
Exposure Group
Relative Risk
(95% CI)
Relative Risk
(95% CI)
Relative Risk Reference(s) and Study
(95% CI) na Description15
Computer manufacturing workers (IBM), NY
Males
Females
2.24(1.01,4.19)
Clapp and Hoffman, 2008
ii= 115 cancer deaths from
1969-2001, proportional mortality
ratio, does not identify TCE
exposure to individual subjects.
EPA based the lymphopoietic
cancer category on "all lymphatic
cancers".
Aerospace workers (Rocketdyne), California
Any TCE (utility/eng flush)
0.74 (0.34, 1.40)
0.21 (0.01, 1.18)
1.08 (0.35,2.53)
5 Boice et al., 2006
n = 41,351 (1,111 Santa Susana
workers with TCE exposure),
employed on or after 1948-1999,
worked >6 months, follow-up to
1999, job exposure matrix without
quantitative estimate of TCE
intensity.
Any TCE exposure
Low cumulative TCE score
Medium cumulative TCE score
High TCE score
Not reported
Not reported
Not reported 60
1.0 (referent) 27
1.49 (0.86, 2.57) 27
1.30 (0.52,3.23) 6
Not reported Zhao et al., 2005
n = 6,044 (n = 2,689 with high
cumulative level exposure to TCE),
began work and worked at least 2
6/22/2009
633
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(p for trend) (0.370)
View-Master employees, Oregon
Males 0.58 ( ,) 3 0.69 ( ,)
Females 0.64 ( ,) 8 0.52 ( ,)
Electronic workers, Taiwan
All employees
Males
Females
Not reported
Not reported
1.27 (0.41, 2.97)
1.14(0.55,2.10)
Aerospace workers (Lockheed), California
Routine TCE, any exposure
Routine-Intermittent
Any TCE exposure
Duration of exposure
O years
<1 year
1-4 years
>5 years
p for trend
1.5 (0.81, 1.60)
Not reported
Not reported
36 1.19 (0.65,1.99)
Not reported
1.0 (referent)
0.74 (0.32, 1.72)
1.33 (0.64, 2.78)
1.62 (0.82, 3.22)
0.20
6/22/2009
634
years in 1950 or later - 1993,
follow-up to 2001, job exposure
matrix (intensity), internal referents
(workers with no TCE exposure).
Leukemia observations included in
non-Hodgkin lymphoma category.
2 0.50 (0.5,2.79) 1
4 0.67(0.14,1.96) 3
ATSDR, 2004
n = 430 deaths from 1989-2001,
proportional mortality ratio, does
not identify TCE exposure to
individual subjects. EPA based the
non-Hodgkin lymphoma cancer
category on "other lymphopoietic
tissue".
5 0.44 (0.05, 1.59) 2
10 0.54 (0.23, 1.07) 8
Chang et al., 2003
n = 88,868 (n = 70,735 female),
began work 1978-1997, follow-up
1985-1997, does not identify TCE
exposure to individual subjects.
14 1.05 (0.54, 1.84)
Not reported
Not reported
32
7
10
14
12 Boice et al., 1999
n = 77,965 (n = 2,267 with routine
TCE exposure and n = 3.016 with
intermittent-routine TCE exposure),
began work >1960, worked at least
1 year, follow-up from 1960-1996,
job exposure matrix without
quantitative estimate of TCE
intensity.
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Uranium-processing workers (Fernald), Ohio
Any TCE exposure
No TCE exposure
Light TCE exposure, >2 yrs
Moderate TCE exposure, >2 yrs
Not reported
1.0 (referent)
1.45 (0.68, 3.06)° 18
1.17(0.15,9.00)c 1
Not reported
Not reported
Not reported
Not reported
Ritz, 1999
Not reported n = 3,814 (n = 2,971 with TCE),
Not reported began work 1951-1972, worked > 3
months, follow-up to 1989, internal
Not reported referents (workers with no TCE
Not reported exposure).
Aerospace workers (Hughes), California
TCE Subcohort
TCE Subcohort
Low Intensity (<50 ppm)
High Intensity (>50 ppm)
TCE Subcohort (Cox Analysis)
Never exposed
Ever exposed
Peak
Cumulative
No/Low
Med/Hi
0.99 (0.64, 1.47) 25
1.07 (0.51, 1.96) 10
0.95 (0.53, 1.57) 15
1.0 (referent) 82
1.05 (0.67, 1.65)f 25 1.36 (0.35,5.22)
Referent
Low
High
Aircraft maintenance workers, Hill Air Force Base,
TCE subcohort
Males, Cumulative exp
0
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
Females, Cumulative exp
1.0 (referent) 90
1.08 (0.64, 1.82) 17
1.0 (referent) 82
1.09(0.56,2.14) 10
1.03 (0.59, 1.79) 15
Utah
1.1 (0.7, 1.8)8
1.0 (referent)
1.1 (0.6, 2.1)
1.0 (0.4,2.1)
1.3 (0.7, 2.5)
66
21
11
21
0.96 (0.20, 2.81) 3
1.01 (0.46, 1.92)e 9
1.79 (0.22, 6.46)d 2
0.50 (0.01, 2.79)d 1
1.0 (referent) 8
d,f
1.0 (referent)
1.31 (0.28, 6.08)d
1.0 (referent)
1.8 (0.6, 5.4)
1.9 (0.6,6.3)
1.1 (0.3,3.8)
1.05 (0.50, 1.93) 10
0.85 (0.17,2.47)
1.17(0.47,2.41)
1.0 (referent)
3 0.99 (0.48,2.03)'
1.0 (referent) 8
2.25 (0.46, 11.l)d 2
0.81 (0.10, 6.49)d 1
2.0 (0.9, 4.6)8 28
10
6
5
1.0 (referent)
0.69 (0.21, 2.32)
1.14 (0.5,2.60)
0.6 (0.3, 1.2)s
3
7
32
10
Morgan etal., 1998
n = 20,508 (4,733 with TCE
exposure), worked > 6 months
1950-1985, follow-up to 1993,
external and internal (all non-TCE
exposed workers) workers referent,
job exposure matrix (intensity)
9 1.0 (referent) 35
2 1.10 (0.49,2.49) 7
32
3
1.0 (referent)
1.0 (0.3,3.2)
1.2 (0.4, 3.6)
16
Blair et al., 1998; Radican et al.,
2008
n = 10,461 men and 3,605 women
(total n = 14,066), employed at least
1 year from 1952 to 1956, follow-up
to 1990 (Blair etal., 1998) or to
2000 (Radican et al., 2008), job
exposure matrix, internal referent
(workers with no chemical
6/22/2009
635
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
0
1.0 (referent)
< 5 ppm-yr
1.5 (0.6,4.0)
6
3.8(0.8, 18.9)
5-25 ppm-yr
0.7 (0.1,4.9)
1
>25 ppm-yr
1.1 (0.4,3.0)
6
3.6 (0.8, 16.2)
TCE subcohort
1.06 (0.75, 1.51) h
106
1.36 (0.77, 2.39)h
Males, Cumulative exp
1.12(0.72, 1.73)
88
1.56 (0.79,4.21)
0
1.0 (referent)
1.0 (referent)
< 5 ppm-yr
1.04 (0.63, 1.74)
34
1.83 (0.79, 4.21)
5-25 ppm-yr
1.06 (0.49, 1.88)
21
1.17 (0.42, 3.24)
>25 ppm-yr
1.25 (0.75, 2.09)
33
1.50 (0.61, 3.69)
Females, Cumulative exp
1.00 (0.55, 1.83)
18
1.18 (0.49,2.85)
0
1.0 (referent)
1.0 (referent)
< 5 ppm-yr
1.10(0.48, 2.54)
7
1.48 (0.47, 4.66)
5-25 ppm-yr
0.38 (0.05,2.79)
1
>25 ppm-yr
1.11 (0.53,2.31)
10
1.30 (0.45, 3.77)
Cardboard manufacturing workers, Arnsburg, Germany
TCE-exposed subjects 1.10(0.03,6.12) 1
Unexposed subjects from same 1.11(0.03,6.19) 1
factory
General Electric plant, Pittsfield, Massachusetts 0.76 (0.24, 2.42)1J
6/22/2009
636
1.0 (referent)
3
0.4 (0.1,3.2)
1
0
0
4
0.3 (0.1,2.4)
1
46
0.64 (0.35, 1.18)h
27
37
0.77 (0.37, 1.62)
24
1.0 (referent)
18
0.86 (0.36, 2.02)
11
7
0.51 (0.16, 1.63)
4
12
0.87 (0.35,2.14)
9
9
0.36 (0.10, 1.32)
3
1.0 (referent)
4
0.35 (0.05, 2.72)
1
0
0
5
0.48 (0.10,2.19)
2
exposures)
15 1.1 (0.46,2.66)1 22
Henschleretal., 1995
n = 169 TCE exposed and n = 190
unexposed men, employed >1 year
from 1956-1975, follow-up to
1992, local population referent,
qualitative exposure assessment
Greenland et al., 1994
Nested case-control study, n = 512
cancer [cases] and 1,202 non-cancer
[controls] male deaths reported to
pension fund between 1969-1984
among workers employed <1984
and with job history record, job
exposure matrix-ever held job with
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Cardboard manufacturing workers, Atlanta, Georgia
0.3 (0.0, 1.6)
Not reported
Not reported
Aircraft manufacturing employees, Italy
All male subjects
0.80 (0.41, 1.40)
12
Not reported
Not reported
Workbench job title
3/1.27
Aircraft manufacturing, San Diego, California
All employees 0.82 (0.56,1.15) 32
0.82 (0.44, 1.41) 13
0.65 (0.21, 1.52)k 5
0.82 (0.47, 1.32) 10
Solvent-exposed rubber workers 2.41 3 0.81
TCE exposure.
Sinks et al., 1999
n = 2,050, employed on or before
1957-1988, follow-up to 1985 (or
1989 by current mailing address),
Material Data Safety Sheets used to
identify chemicals used in work
areas.
Costa et al., 1989
n = 7,676, employed on or before
1954-1981, followed to 1981, job
titles of white- and blue-collar
workers, technical staff, and
administrative clerks, does not
identify TCE exposure to individual
subjects.
Garabrant et al., 1988
n = 14,067, employed at least 4
years with company and >1 day at
San Diego plant from 1958-1982,
followed to 1982, does not identify
TCE exposure to individual
subjects.
Wilcosky et al., 1984
Nested case-control study, n = 9
lymphosarcoma and 10 leukemia
[cases] and 20% random sample of
6/22/2009
637
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
all other deaths [controls] between
1964-1973 in cohort of n = 6,678,
exposure assessment by company
record for use in work area
a n = number of observed cases
b Unless otherwise noted, all studies reported standardized mortality ratios using an external population referent group.
0 Logistic regression analysis with 15 lag for TCE exposure (Ritz, 1999)
d In Morgan et al. (1998) and Garabrant et al. (1988), this category was based on lymphosarcoma and reticulosarcoma.
e As presented in Mandel et al. (2006), this category defined as ICD -7, ICDA-8, and ICD-9 codes of 200 and 202.
f Risk ratio from Cox Proportional Hazard Analysis, stratified by age and sex, from Environmental Health Strategies (1997) Final Report to Hughes Corporation
(Communication from Paul A. Cammer, President, Trichloroethylene Issues Group to Cheryl Siegel Scott, U.S. EPA, December 22, 1997).
8 Estimated relative risks from Blair et al. (1998) from Poisson regression models adjusted for date of hire, calendar year of death and sex.
h Estimated relative risks from Radican et al. (2008) from Cox proportional hazard models adjusted for age and sex.
1 Odds ratio from nested case-control analysis
J Lymphomas, lymphosarcomas, and reticulosarcomas (ICDA8 200-202) in Greenland et al. (1994)
k Other lymphatic and hematopoietic tissue neoplasms (Garabrant et al., 1988)
6/22/2009
638
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-5. Case-control studies of TCE exposure and lymphopoietic cancer or leukemia
Population
Cancer Type and Exposure Group
Odds Ratio
(95% CI)
n
exposed
cases
Reference(s)
Women aged 21-84 in
CT, USA
Non-Hodgkin Lymphoma
Wang et al., 2009
Any TCE exposure
1.2 (0.9, 1.8)
77
Low intensity TCE exposure
1.1 (0.8, 1.6)
64
Medium-high intensity TCE exposure
2.2 (0.9, 5.4)
13
(p for linear trend)
0.06
Low probability TCE exposure
1.1 (0.7, 1.8)
43
Medium-high probability TCE exposure
1.4 (0.9, 2.4)
34
(p for linear trend)
0.37
Low intensity TCE exposure/ Low probability
0.9 (0.6, 1.5)
30
Low intensity /Medium-high probability
1.4 (0.9, 2.4)
34
Medium-high intensity/Low probability
2.2 (0.9, 5.4)
13
Medium-high intensity/Medium-high probability
0
Population in 6
German Regions
Non-Hodgkin Lymphoma
Seidler et al., 2007; Mester et al., 2006
Any TCE exposure
Not reported
Cumulative TCE
0 ppm-years
1.0
610
>0-<4 ppm-years
0.7 (0.4, 1.1)
40
4.4-<35 ppm-years
0.7 (0.5, 1.2)
32
High exposure, >35 ppm-years
2.1 (1.0,4.8)
21
(p for linear trend)
0.14
>35 ppm-years, 10 year lag
2.2(1.0,4.9)
B-cell NHL
Cumulative TCE
0 ppm-years
1.0
47
6/22/2009
639
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
>0-<4 ppm-years 0.7(0.5,1.2)
4.4-<35 ppm-years 0.8(0.5,1.3)
High exposure, >35 ppm-years 2.3 (1.0, 5.3)
(p for linear trend) 0.08
Diffuse B-cell NHL
Cumulative TCE
0 ppm-years
1.0
>0-<4 ppm-years
0.5 (0.2,
1.2)
4.4-<35 ppm-years
0.8 (0.3,
1.8)
High exposure, >35 ppm-years
2.6 (0.7,
3.0)
(p for linear trend)
0.03
Chronic Lymphocytic Leukemia
Cumulative TCE
0 ppm-years
1.0
>0-<4 ppm-years
1.1 (0.5,
2.4)
4.4-<35 ppm-years
0.7 (0.3,
1.7)
High exposure, >35 ppm-years
0.9 (0.2,
4.5)
(p for linear trend)
0.46
Population in 8 Italian
Regions
Non-Hodgkin lymphoma
Any TCE exposure Not reported
TCE exposure intensity
very low/low 0.8 (0.5, 1.3)
medium/high 1.2 (0.7, 2.0)
(p for linear trend) 0.8
Duration exposure, Med/High TCE intensity
< 15 yr 1.1(0.6,2.1)
>15 1.0(0.5,2.6)
(p for linear trend) 0.72
6/22/2009
640
32
27
17
139
6
7
4
610
10
6
2
Miligi et al., 2006
35
35
22
12
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Other non-Hodgkin lymphoma
TCE exposure intensity, Medium/High
Small lymphocytic NHL
Follicular NHL
Diffuse NHL
Other NHL
Leukemia
Any TCE exposure
TCE exposure intensity
Acute myeloid leukemia
Any TCE exposure
TCE exposure intensity
Chronic lymphocytic leukemia
Any TCE exposure
TCE exposure intensity
very low/low
medium/high
very low/low
medium/high
very low/low
medium/high
0.9(0.4,2.1)
Not presented
1.9 (0.9, 3.7)
1.2 (0.6, 2.4)
Not reported
1.0 (0.5, 1.8)
0.7 (0.4, 1.5)
Not reported
1.0 (0.4, 2.5)
1.1 (0.5,2.9)
Not reported
1.2 (0.5, 2.7)
0.9 (0.3, 2.6)
Population of Orebro and Linkoping, Sweden
B-cell non-Hodgkin lymphoma
Any TCE exposure
1.2 (0.5, 2.4)
Population of Sweden Hairy cell lymphoma
Any TCE exposure
1.5 (0.7, 3.3
Population of Umea,
Sweden
Non-Hodgkin lymphoma
6/22/2009
641
Costantini et al., 2008
Persson and Fredrikson, 1999
Nordstrom etal., 1998
Hardell et al., 1994
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Any exposure to TCE 7.2(1.3,42)
Population of
Montreal, Canada Non-Hodgkin lymphoma
Any TCE exposure 1.1 (0.6, 2.3)
Substantial TCE exposure 0.8 (0.2, 2.5)
a 90% Confidence Interval
6/22/2009
642
4
Siemiatycki et al., 1991
6
2
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5.-6. Geographic-based Studies of TCE and Non-Hodgkin Lymphoma or Leukemia in Adults
non-Hodgkin Lymphoma Leukemia
n
n
Relative Risk
exposed
Relative Risk
exposed
Population
Exposure Group
(95% CI)
cases
(95% CI)
cases
Reference
Two study areas in Endicott, NY
0.54 (0.22, 1.12)
7
0.79 (0.34, 1.55)
8
ATSDR, 2006
Residents of 13 census tracts in Redland, CA
1.09 (0.84, 1.38)
111
1.02 (0.74, 1.35)
77
Morgan and Cassady, 2002
Population in
Males, maximum estimated TCE
New Jersey
concentration (ppb) in municipal
drinking water
Cohnetal., 1994
<0.1
1.00
493
1.00
438
0.1-0.5
1.28(1.10, 1.48)
272
0.85 (0.71, 1.02)
162
>5.0
1.20 (0.94, 1.52)
78
1.10 (0.84, 1.90)
63
Females, maximum estimated TCE
concentration (ppb) in municipal
drinking water
<0.1
1.00
504
1.00; 315
0.1-0.5
1.02 (0.87, 1.2)
26
1.13 (0.93, 1.37)
156
>5.0
1.36 (1.08, 1.70)
87
1.43 (1.43, 1.90)
56
Population in
Finland
Residents of Hausjarvi
0.6 (0.3, 1.1)
14
1.2 (0.8, 1.7)
33
Vartiainen et al., 1993
Residents of Huttula
1.4(1.0,2.0)
13
0.7 (0.4, 1.1)
19
6/22/2009
643
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
6/22/2009 644
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
The number of studies of childhood lymphoma including acute lymphatic leukemia and
trichloroethylene is much smaller than the number of studies of trichloroethylene and adult
lymphomas, and consists of four case-control studies (Costas et al., 2002; Lowengart et al., 1987;
McKinney et al., 1991; Shu et al., 1999) and four geographic based studies(Aickin et al., 1992;
ADHS, 1990, 1995; AT SDR, 2006, 2008; Cohn et al., 1994) (Table 4.5-7). An additional
publication, focusing on ras mutations, based on one of the case-control studies is also available
(Shu et al., 2004). All four case-control studies evaluate maternal exposure, and three studies
also examine paternal occupational exposure (Lowengart et al., 1987; McKinney et al., 1991;
Shu et al., 2004, 1999). There are relatively few cases with maternal exposure (range 0 to 16) in
these case-control studies, and only Shu et al. have a large number (n = 136) of cases with
paternal exposure (Shu et al., 2004, 1999). The small numbers of exposed case parents limit
examination of possible susceptibility time windows. Overall, evidence for association between
parental trichloroethylene exposure and childhood leukemia is not robust or conclusive.
The results from the studies of Costas et al. (2002) and Shu et al. (1999, 2002) suggest a
fetal susceptibility to maternal exposure during pregnancy, with relative risks observed for this
time period equal or higher than the relative risks observed for periods before conception or after
birth (Table 4.5-7). The studies by Lowengart et al. (1987) and McKinney et al. (1991) do not
provide informative data pertaining to this issue due to the small number (n = <3) of exposed
case mothers. A recent update of a cohort study of electronics workers at a plant in Taiwan
(Chang et al., 2003, 2005) reported a four-fold increased risk [3.93; 95% CI; 1.17, 12.55(Sung et
al., 2008)] for childhood leukemia risk among the offspring of female workers employed during
the three months before to three months after conception. Exposures at this factory included
trichloroethylene, perchloroethylene, and other organic solvents (Sung et al., 2008) and the lack
of TCE assignment to individual subjects in this study decrease its weight in the overall analysis.
The evidence for an association between childhood leukemia and paternal exposure to
solvents is quite strong (Colt and Blair, 1998); however, for studies of TCE exposure, the small
numbers of exposed case fathers in two studies (McKinney et al., 1991; Lowengart et al., 1987)
and, for all three studies, likelihood of misclassification resulting from a high percentage of
paternal occupation information obtained from proxy interviews, limits observation
interpretations. Both Lowengart et al. (1987) and McKinney et al. (1991) provide some evidence
for a two- to four-fold increase of childhood leukemia risk and paternal occupational exposure
although the population study of Shu et al. (1999, 2002), with 13% of case father's occupation
reported by proxy respondents, does not appear to support the earlier and smaller studies.
The geographic based studies for adult lymphopoietic (Table 4.5-6) or childhood
leukemias (Table 4.5-7) do not greatly contribute to the overall weight of evidence. While some
studies observed statistically significantly elevated risks for NHL or childhood cancer, these
6/22/2009
645
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
studies generally fulfilled only the minimal of evaluation criteria with questions raised about
subject selection (Morgan and Cassady, 2002), their use of less sophisticated exposure
assessment approaches and associated assumption of an average exposure to all subjects (all
studies), and few cases with high level parental exposure (all studies).
6/22/2009
646
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5.-7, Selected Results from Epidemiologic Studies of TCE Exposure and Childhood Leukemia
Relative Risk
(95% CI)
n
observed
events
Cohort Studies (solvents)
Childhood leukemia among offspring of electronic workers
Nonexposed
Exposed to organic solvents
Case-control Studies
Children's Cancer Group Study (children <15 years age)
Acute lymphocytic leukemia
Maternal occupational exposure to TCE
1.01
3.83 (1.17, 12.55)
Anytime
1.8 (0.8,4.1)
15
Preconception
1.8 (0.8,5.2)
9
During pregnancy
1.8 (0.5,6.4)
6
Postnatal
1.4 (0.5,4.1)
9
Paternal occupational exposure to TCE
Anytime
1.1 (0.8, 1.5)
136
Preconception
1.1 (0.8, 1.5)
100
During pregnancy
0.9 (0.6, 1.4)
56
Postnatal
1.0 (0.7, 1.3)
77
K-ras + acute lymphocytic leukemia
Maternal occupational exposure to TCE
Anytime
1.8 (0.6,4.8)
5
Preconception
2.0 (0.7,6.3)
4
During pregnancy
3.1 (1.0,9.7)
4
Postnatal
0
Paternal occupational exposure to TCE
Anytime
0.6 (0.3, 1.4)
9
Preconception
0.6 (0.3, 1.5)
8
During pregnancy
0.3 (0.1, 1.2)
2
Postnatal
0.4 (0.1, 1.4)
3
Residents of ages < 19 in Woburn, MA
Maternal exposure 2 years before conception
to diagnosis
Never
1.00
3
Least
5.00 (0.75,33.5)
9
Most
3.56 (0.51,24.8)
7
(p for linear trend)
>0.05
Maternal exposure 2 years before conception
Never
1.00
11
Least
2.48 (0.42, 15.2)
4
Sung et al., 2008
Shu et al., 1999
Shu et al., 2004
Costas et al., 2002
6/22/2009
647
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Most
(p for linear trend)
Birth to diagnosis
Never
Least
Most
(p for linear trend)
Maternal exposure during pregnancy
Never
Least
Most
(p for linear trend)
2.82 (0.30,26.4)
>0.05
1.00
1.82 (0.31, 10.8)
0.90 (0.18,4.56)
>0.05
1.00
3.53 (0.22, 58.1)
14.3 (0.92, 224)
<0.05
Population <14 years of age in 3 areas north England, United
Kingdom
Acute lymphocytic leukemia and NHL
Maternal occupation exposure to TCE
Preconception
Paternal occupational exposure to TCE
Preconception
Periconception and gestation
Postnatal
McKinney et al., 1991
1.16(0.13,7.91)
2.27 (0.84,6.16)
4.49(1,15,21)
2.66 (0.82, 9.19)
Los Angeles Cancer Surveillance Program
Acute lymphocytic and nonlymphocytic
leukemia, < 10 years of age
Maternal occupational exposure to TCE
Paternal occupational exposure to TCE
One year before pregnancy
During pregnancy
After delivery
Lowengart et al., 1987
2.0 (p = 0.16)
2.0 (p =0.16)
2.7 (0.64, 15.6)
6/3
6/32
8/32
Geographic Based Studies
Two study areas in Endicott, NY
Leukemia, < 19 years of age
Not reported
<6
ATSDR, 2006
Population in New Jersey
Acute lymphocytic leukemia
Maximum estimated TCE concentration
(ppb) in municipal drinking water
Males
<0.1
0.1-0.5
1.00
0.91(0.53, 1.57)
45
16
Cohn et al., 1994
6/22/2009
648
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
>5.0 0.54(0.17,17.7) 3
Females
<0.1 1.00 25
0.1-0.5 1.85 (1.03,3.70) 22
>5.0 2.36 (1.03,5.45) 7
Resident of Tucson Airport Area, AZ ADHS, 1990, 1995
Leukemia, < 19 years of age
1970-1986 1.48(0.74,2.65) 11
1987-1991 0.80 (0.31,2.05) 3
Resident of West Central Phoenix, AZ Aickin et al., 1992
Leukemia, < 19 years of age 1.95 (1.43,2.63) 38
1 Internal referents, live born children among female workers not exposed to organic solvents
2 Discordant pairs
6/22/2009
649
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.5.1.2.2 Meta-analysis
Meta-analysis is adopted as a tool for examining the body of epidemiologic evidence on
NHL and TCE exposure and to identify possible sources of heterogeneity. The meta-analysis of
lymphoma examines 15 cohort and case-control studies identified through a systematic review
and evaluation of the epidemiologic literature on TCE exposure (Siemiatycki et al., 1991;
Axelson et al., 1994; Hardell et al., 1994; Anttila et al., 1995; Blair et al., 1998; Greenland et al.,
1994; Morgan et al., 1998; Nordstrom et al., 1998; Boice et al., 1999; Persson and Fredrikson,
1999; Hansen et al., 2001; Raaschou-Nielsen et al., 2003; Zhao et al., 2005; Miligi et al., 2006;
Seidler et al., 2007). These 15 studies of lymphoma and TCE had high likelihood of exposure,
were judged to have met, to a sufficient degree, the stands of epidemiologic design and analysis,
and reported estimated risks for overall TCE exposure; 11 of these studies, also, presented
estimated lymphoma risk with high level TCE exposure (Siemiatycki et al., 1991; Axelson et al.,
1994; Anttila et al., 1995; Blair et al., 1998; Morgan et al., 1998; Boice et al., 1999; Hansen et
al., 2001; Raaschou-Nielsen et al., 2003; Zhao et al., 2005; Miligi et al., 2006; Seidler et al.,
2007). Full details of the systematic review and meta-analysis of the TCE studies is discussed in
Appendices B and C.
The meta-analyses of the overall effect of TCE exposure on lymphoma suggest a small,
robust, and statistically significant increase in NHL risk. The pooled estimate from the primary
random effect meta-analysis (pooled relative risk estimate, RRp) was 1.27 (95% CI: 1.04, 1.53)
(Figure 4.5 - 1). This result and its statistical significance were not overly influenced by most
individual studies, though the removal of Hansen et al. (2001) resulted in the RRp just missing
statistical significance, with a RRp of 1.17 (95% CI: 1.00, 1.38). The result is similarly not
sensitive to most individual risk ratio estimate selections, except that the RRp is no longer
statistically significant when the Zhao et al. (2005) mortality results are substituted by either the
study's incidence results [RRp of 1.22 (95% CI: 0.99, 1.49)] or the Boice et al. (2006) results
[RRp of 1.24 (95% CI: 1.00, 1.54).
Meta-analysis of the highest exposure groups, either duration, intensity, or their product,
cumulative exposure, results in an RRp of 1.50 (95% CI: 1.20, 1.88), which is greater than the
RRp from the overall exposure analysis, and provides additional support for an association
between NHL and TCE (Figure 4.5 - 2). The highest exposure category groups have a reduced
likelihood for exposure misclassification because they are believed to represent a greater
differential TCE exposure compared to people identified with overall TCE exposure.
Observation of greater risk associated with higher exposure category compared to overall
(typically any versus none) exposure comparison additionally suggests an exposure-response
gradient between NHL and TCE, although estimation of a level of exposure associated with the
meta-relative risk is not possible.
6/22/2009
650
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Heterogeneity in RRp is observed across the results of the 15 studies in the analysis
(p = 0.048), with difference between cohort and case-control studies explaining much of the
observed heterogeneity, and some evidence of publication bias. Increased risk of lymphoma was
strengthened in analysis limited to cohort studies and virtually eliminated in the case-control
study analysis. Examination of heterogeneity in cohort and case-control studies separately was
not statistically significant in either case although some may be present given that statistical tests
of heterogeneity are generally insensitive in cases of minor heterogeneity. Sources of
heterogeneity are uncertain and may reflect several features known to influence epidemiologic
studies. One reason may be differences in exposure assessment and in overall TCE exposure
concentration between cohort and case-control studies. Several cohort studies (Anttila et al.,
1995; Axelson et al., 1994; Blair et al., 1998; Hansen et al., 2001; Raaschou-Nielsen et al., 2003)
adopt exposure assessment approaches that are expected to reduce potential for bias (NRC,
2006). Exposure misclassification bias due to random or measurement error and recall bias
amore likely in three case-control studies (Hardell et al., 1994; Nordstrom et al., 1998; Persson
and Fredrikson, 1999) with self-reported TCE exposure compared to Siemiatycki (1991), Miligi
et al. (2006), Seidler et al. (2007). In addition, a low overall TCE exposure prevalence is
anticipated in population case-control studies which would typically assess a large number of
workplaces and operations, where exposures are less well defined, and where case and control
subjects identified as exposed to TCE probably have minimal contact (NRC, 2006). Observed
higher risk ratios with higher exposure categories in NHL case-control studies support exposure
differences as a source of heterogeneity.
Diagnostic inaccuracies are likely another source of heterogeneity in the meta-analysis
through study differences in lymphoma groupings and in lymphoma classification schemes. All
studies include a broad but slightly different group of lymphosarcoma, reticulum-cell sarcoma,
and other lymphoid tissue neoplasms (Codes 200 and 202), except Nordstrom et al. (1998)
whose case-control study examined hairy cell leukemia, now considered a lymphoma. Cohort
studies have some consistency in coding NHL, with NHL defined as lymphosarcoma and
reticulum-cell sarcoma (200) and other lymphoid tissue neoplasms (202) using the International
Disease Classification (ICD), Revision 7, 200 and 202 - four studies (Axelson et al., 1994;
Anttila et al., 1995; Hansen et al., 2001; Raaschou-Nielsen et al., 2003), ICD-Adapted, Revision
8 (Blair et al., 1998), and ICD-7, 8, and 9, per the version in use at the time of death (Morgen et
al., 1997, as presented in Mandel et al., 2006; Boice et al., 1999), as does the case-control study
of Siemiatycki (1991) whose coding scheme for NHL is consistent with ICD 9, 200 and 202.
Case-control studies, on the other hand, have adopted other classification systems for defining
NHL including the NCI Working Formulation (Miligi et al., 2006), World Health Organization
6/22/2009
651
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(Seidler et al., 2007), Rappaport (Hardell et al., 1994), or else do not identify the classification
system for defining NHL (Persson and Fredrikson, 1999).
NRC (2006) deliberations on trichloroethylene commented on two prominent evaluations
of the then-current TCE epidemiologic literature using meta-analysis techniques. These studies
were by Wartenberg et al. (2000), and by Kelsh et al. (2005), submitted by Exponent-Health
Sciences to NRC during their deliberations and subsequently published in a paper on NHL
(Mandel et al., 2006) and a paper on multiple myeloma and leukemia (Alexander et al., 2006).
The NRC found weaknesses in the techniques used in each of these studies, and suggested that
EPA conduct a new meta-analysis of the epidemiologic data on trichloroethylene using objective
and transparent criteria so as to improve on the past analyses. EPA staff conducted their analysis
according to NRC (2006) suggestions for transparency, systematic review criteria, and
examination of both cohort and case-control studies. The EPA analysis of NHL analysis
considered a larger number of studies than in the previous analyses (Mandel et al., 2006;
Wartenberg et al., 2000), and includes recently published studies (Boice et al., 2006; Miligi et al.,
2006; Seidler et al., 2007; Zhao et al., 2005). Despite the weaknesses in Wartenberg et al.
(2000), Kelsh (2005) and Mandel et al. (2006), pooled NHL risk for overall TCE exposure in
these analyses is of a similar magnitude as that observed in EPA's updated analysis [1.5, 95%
CI: 0.9, 2.3, Tier 1 incidence; 1.2, 95% CI: 0.9, 1.7, Tier 1 mortality (Wartenberg et al., 2000);
1.59, 95% CI: 1.21, 2.08, Group I, TCE Subcohorts, 1.39, 95% CI: 0.62, 3.10, case-control
studies (Kelsh, 2005; Mandel et al, 2006)].
EPA did not perform a pooled analysis of leukemia observations. Seven studies
presented estimated risks for leukemia and overall TCE exposure (Anttila et al., 1995; Blair et
al., 1998; Morgan et al., 1998; Boice et al., 1999, 2006; Hansen et al., 2001; Raachou-Nielsen et
al., 2003); only three studies also presented estimated risks for a high exposure category (Anttila
et al., 1995; Morgan et al., 1998; Blair et al., 1998). Two case-control studies presented
estimated risk for leukemia categories and low or high TCE exposure category (Seidler et al.,
2007; Costantini et al., 2008); however, neither study presented estimated risk for overall TCE
exposure. In spite of the fewer number of studies with information on leukemia compared to
NHL, Alexander et al. (2006) present an estimated of the pooled relative risk (RRp) for leukemia
of 1.11 (95%) CI: 0.93, 1.32). Sensitivity analysis of leukemia observation was not included in
Alexander et al. (2006), as was recommended by NRC (2006).
6/22/2009
652
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE and lymphoma
Study name
Statistics for each study
Rate ratio and 95% CI
Rate
Lower
Upper
ratio
limit
limit
p-Value
Anttila 1995
1.810
0.905
3.619
0.093
Axelson 1994
1.520
0.633
3.652
0.349
Blair 1998
2.000
0.885
4.521
0.096
Boice 1999
1.190
0.705
2.009
0.515
Greenland 1994
0.760
0.239
2.413
0.642
Hansen 2001
3.100
1.550
6.199
0.001
Morgan 1998
1.010
0.526
1.941
0.976
Raaschou-Nielsen 2003
1.240
1.011
1.521
0.039
Zhao 2005 moil
1.437
0.899
2.297
0.130
Hardell 1994
7.200
1.267
40.923
0.026
Miligi 2006
0.933
0.671
1.298
0.682
Nordstrom 1998
1.500
0.691
3.257
0.305
Persson&Fredrikson 19991.200
0.548
2.629
0.649
Seidler 2007
0.800
0.566
1.131
0.207
Siemiatycki 1991
1.100
0.479
2.525
0.822
1.266
1.045
1.533
0.016
0.1 0.2
0.5
10
random effects model
Figure 4.5-1. Meta-analysis of lymphoma and overall TCE exposure. The pooled estimate is
in the bottom row. Symbol sizes reflect relative weights of the studies. The horizontal midpoint
of the bottom diamond represents the pooled RR estimate and the horizontal extremes depict the
95% CI limits.
6/22/2009
653
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE and lymphoma - highest exposure groups
Study name
Statistics for each study
Rate ratio and 95% CI
Rate
Lower
Upper
ratio
limit
limit
p-Value
Anttila 1995
1.400
0.350
5.598
0.634
Axelson 1994
6.250
0.880
44.369
0.067
Blair 1998 inc
0.970
0.421
2.237
0.943
Boice 1999
1.620
0.818
3.210
0.167
Hansen 2001 cum exp
2.700
0.871
8.372
0.085
Morgan 1998
0.810
0.101
6.525
0.843
Raaschou-Nielsen 2003
1.600
1.119
2.288
0.010
Zhao 2005 mort
1.300
0.522
3.240
0.573
Miligi 2006
1.200
0.709
2.028
0.497
Seidler 2007
2.300
1.008
5.250
0.048
Siemiatycki 1991
0.800
0.195
3.275
0.756
1.502
1.201
1.879
0.000
0.1 0.2
0.5
10
random effects model
Figure 4.5 - 2. Meta-analysis of lymphoma and TCE exposure - highest exposure groups. The
pooled estimate is in the bottom row. Symbol sizes reflect relative weights of the studies. The
horizontal midpoint of the bottom diamond represents the pooled RR estimate and the horizontal
extremes depict the 95% CI limits.
6/22/2009
654
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.5.2 Animal Studies
The immunosuppressive and immunomodulating potential of TCE has not been fully
evaluated in animal models across various exposure routes, over various relevant durations of
exposure, across representative life stages, and/or across a wide variety of endpoints.
Nevertheless, the studies that have been conducted indicate a potential for TCE-induced
immunotoxicity, both following exposures in adult animals and during immune system
development (i.e., in utero and pre-weaning exposures).
4.5.2.1 Immunosuppression
A number of animal studies have indicated that moderate to high concentrations of TCE
over long periods have the potential to result in immunosuppression in animal models, dependant
on species and gender. These studies are described in detail below and summarized in Table
4.5-8
4.5.2.1.1 Inhalation exposures
Mature cross-bred dogs (5/group) were exposed to 0, 200, 500, 700, 1,000, 1,500, or
2,000 ppm TCE for 1-hour or to 700 ppm TCE for 4 hours, by tracheal intubation under
intravenous sodium pentobarbital anesthesia. An additional group of dogs was exposed by
venous injection of 50 mg/kg TCE administered at a rate of 1 mL/min (Hobara et al., 1984).
Blood was sampled pre- and post-exposure for erythrocyte and leukocyte counts. Marked,
transient decreases in leukocyte counts were observed at all exposure levels 30 minutes after
initiation of exposure. At the end of the exposure period, all types of leukocytes were decreased
(by 85%); neutrophils were decreased 33%, and lymphocytes were increased 40%. There were
no treatment-related changes in erythrocyte counts, hematocrit values, or thrombocyte counts.
In a study that examined the effects of a series of inhaled organic chemical air
contaminants on murine lung host defenses, Aranyi et al. exposed female CD1 mice to single 3-
hour exposures of TCE at time-weighted concentrations of 0, 2.6, 5.2, 10.6, 25.6, or 48 ppm
(Aranyi et al., 1986). Additionally, at the dose at which no adverse treatment-related effect
occurred with a single exposure (i.e., 2.6 ppm), a multiple exposure test (5 days, 3 hr/day) was
conducted. Susceptibility to Streptococcus zooepidimicus aerosol infection and pulmonary
bactericidal activity to inhaled Klebsiella pneumoniae were evaluated. There was a significant
(p < 0.0001) treatment by concentration interaction for mortality, with the magnitude of the
6/22/2009
655
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
effect increasing with concentration. A significant (p < 0.0001) treatment by concentration
interaction was also found for bactericidal activity. Single 3-hr exposures at 10.6, 25.6, and 48
ppm resulted in significant increases in mortality, although increases observed after single
exposures at 5.2 or 2.6 ppm or five exposures at 2.6 ppm were not significant. Pulmonary
bactericidal activity was significantly decreased after a single exposure at 10.6 ppm, but single
exposures to 2.6 or 5.2 ppm resulted in significant increases.
In a host-resistance assay, CD-I mice (sex and number/group not specified) exposed to
TCE by inhalation for 3 hours at 50-200 ppm were found to be more susceptible to increased
infection following challenge with Streptococcus zooepidimicus administered via aerosol (Park
et al., 1993). Dose-related increases in mortality, bacterial antiphagocytic capsule formation, and
bacterial survival were observed. Alveolar macrophage phagocytosis was impaired in a dose-
responsive manner, and an increase in neutrophils in bronchoalveolar lavage fluid was observed
in exposed mice 3 days post infection.
A guideline (OPPTS 870.3800) 4-week inhalation immunotoxicity study was conducted
in female Sprague-Dawley rats (Woolhiser et al., 2006). The animals (16/group) were exposed
to TCE at nominal levels of 0, 100, 300, or 1,000 ppm for 6 hours/day, 5 days/week. Effects on
the immune system were assessed using an antigen response assay, relevant organs weights,
histopathology of immune organs, and hematology parameters. Four days prior to study
termination, the rats were immunized with sheep red blood cells (SRBC), and within 24 hours
following the last exposure to TCE, a plaque forming cell assay was conducted to determine
effects on splenic anti-SRBC IgM response. Minor, transient effects on body weight and food
consumption were noted in treated rats for the first 2 weeks of exposure. Mean relative liver and
kidney weights were significantly (p = 0.05) increased at 1,000 ppm as compared to control,
while lung, spleen, and thymus weights were similar to control. No treatment-related effects
were observed for hematology, WBC differential counts, or histopathological evaluations
(including spleen, thymus, and lung-associated lymph nodes). At 1,000 ppm, rats demonstrated
a 64% decrease in plaque forming cell assay response. Lactate dehydrogenase, total protein
levels, and cellular differentiation counts evaluated from bronchoalveolar lavage (BAL) samples
were similar between control and treated groups. A phagocytic assay using BAL cells showed
no alteration in phagocytosis, although these data were not considered fully reliable since 1) the
number of retrieved macrophage cells was lower than expected and pooling of samples was
conducted and 2) samples appear to have been collected at 24-hours after the last exposure
(rather than within approximately 2 hours of the last exposure), thereby allowing for possible
macrophage recovery. The NOAEL for this study was considered by the study authors to be 300
ppm, and the LOAEL was 1,000 ppm; however, the effect level may have actually been lower.
It is noted that the outcome of this study does not agree with the studies by Aranyi et al. (1986)
6/22/2009
656
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
and Park et al. (1993), both of which identified impairment of macrophage phagocytic activity in
BAL following inhalation TCE exposures.
4.5.2.1.2 Oral exposures
In a study by Sanders et al., TCE was administered to male and female CD-I mice for 4
or 6 months in drinking water at concentrations of 0, 0.1, 1, 2.5, or 5 mg/mL (Sanders et al.,
1982). In females, humoral immunity was suppressed at 2.5 and 5 mg/mL, while cell-mediated
immunity and bone marrow stem cell activity were inhibited at all dose levels. Male mice were
relatively unaffected either at 4 or 6 months, even though a preliminary study in male CD-I mice
(exposed to TCE for 14 days by gavage at 0, 24, or 240 mg/kg-day) had demonstrated a decrease
in cell-mediated immune response to SRBC in male mice at both treatment levels.
A significant decrease in humoral immunity (as measured by plasma hemagglutination
titers and the number of spleen antibody producing cells of mice sensitized to sheep
erythrocytes) was observed by Kaufmann et al. (1982) in female CD-I mice (15-20/group)
following a 90-day drinking water exposure to 0, 0.07, or 0.7 mg/mL (equivalent to 0, 18, or 173
mg/kg) chloral hydrate, a metabolite of TCE. Similar responses were not observed in male CD-I
mice exposed for 90 days in drinking water (at doses of 0, 16, or 160 mg/kg-day), or when
administered chloral hydrate by gavage to 12/group for 14 days at 14.4 or 144 mg/kg-day.
The potential for developmental immunotoxicity was assessed in B6C3F1 mice
administered TCE in drinking water at dose levels of 0, 1,400 or 14,000 ppb from gestation day 0
to either 3 or 8 weeks of age (Adams et al., 2003 [preliminary data]; Peden-Adams et al., 2006).
At 3 and 8 weeks of age, offspring lymphocyte proliferation, NK cell activity, SRBC-specific
IgM production (PFC response), splenic B220+ cells, and thymus and spleen T-cell
immunophenotypes were assessed. Delayed-typed hypersensitivity and autoantibodies to ds-
DNA were evaluated in offspring at 8 weeks of age. Observed positive responses consisted of
suppressed PFC responses in males at both ages and both TCE treatment levels, and in females at
both ages at 14,000 ppb and at 8 weeks of age at 1,400 ppb. Spleen numbers of B220+ cells
were decreased in 3-week old pups at 14,000 ppb. Pronounced increases in all thymus T-cell
subpopulations (CD4+, CD8+, CD4+/CD8+, and CD4-/CD8-) were observed at 8-weeks of age.
Delayed hypersensitivity response was increased in 8-week old females at both treatment levels
and in males at 14,000 ppb only. No treatment-related increase in serum anti-ds-DNA antibody
levels was found in the offspring at 8 weeks of age.
In a study designed to examine potential susceptibility of the young (Blossom and Doss,
2007), TCE was administered to groups of pregnant MRL +/+ mice in drinking water at
occupationally-relevant levels of 0, 0.5, or 2.5 mg/mL. A total of 3 litters per treatment group
were maintained following delivery (i.e., a total of 11 pups at 0 mg/mL TCE, 8 pups at 0.5
6/22/2009
657
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
mg/mL TCE, and 12 pups at 2.5 mg/mL TCE), and TCE was continuously administered to the
offspring until young adulthood (i.e., 7-8 weeks of age). Although there were no effects on
reproduction, offspring post-weaning body weights were significantly decreased in both treated
groups. Additionally, TCE exposure was found to modulate the immune system following
developmental and early life exposures. Decreased spleen cellularity and reduced numbers of
CD4+, CD8+, and B220+ lymphocyte subpopulations were observed in the post-weaning
offspring. Thymocyte development was altered by TCE exposures, as evidenced by significant
alterations in the proportions of double-negative subpopulations and inhibition of in vitro
apoptosis in immature thymocytes. TCE was also shown to induce a dose-dependent increase in
CD4+ and CD8+ T-lymphocyte IFNy in peripheral blood by 4-5 weeks of age, although these
effects were no longer observed at 7-8 weeks of age. Serum anti-histone autoantibodies and
total IgG2a were significantly increased in treated offspring; however, no histopathological signs
of autoimmunity were observed in the liver and kidneys at sacrifice.
This increase in T-cell hyperactivity was further explored in a study by Blossom et al.
(2008). In this study, MRL +/+ mice were treated in the drinking water with 0 or 0.1 mg/mL
TCE. Based on drinking water consumption data, average maternal doses of TCE were 25.7
mg/kg-day, and average offspring (PND 24-42) doses of TCE were 31.0 mg/kg-day. Treatment
was initiated at the time of mating, and continued in the females (8/group) throughout gestation
and lactation. Pups were weaned at PND 24, and the offspring were continued on drinking water
treatment in a group-housed environment until study termination (PND 42). Subsets of offspring
were sacrificed at PND 10 and 20, at which time developmental and functional endpoints in the
thymus were evaluated (i.e., total cellularity, CD4+/CD8+ ratios, CD24 differentiation markers,
and double-negative subpopulation counts). Indicators of oxidative stress were measured in the
thymus at PND 10 and 20, and in the brain at PND 42. Mitogen-induced intracellular cytokine
production by splenic CD4+ and CD8+ T cells was evaluated in juvenile mice and brain tissue
was examined at PND 42 for evidence of inflammation. Behavioral testing was also conducted;
these methods and results are described in Section 4.2. TCE treatment did not affect
reproductive capacity, parturition, or ability of dams to maintain litters. The mean body weight
of offspring was not different between the control and treated groups. Evaluation of the thymus
identified a significant treatment-related increase in cellularity, accompanied by alterations in
thymocyte subset distribution, at PND 20 (sexes combined). TCE treatment also appeared to
promote T cell differentiation and maturation at PND 42, and ex vivo evaluation of cultured
thymocytes indicated increased reactive oxygen species (ROS) generation. Evaluation of
peripheral blood indicated that splenic CD4+ T cells from TCE-exposed PND 42 mice produced
significantly greater levels of IFN-y and IL-2 in males and TNF-a in both sexes. There was no
effect on cytokine production on PND 10 or 20. The dose of TCE that resulted in adverse
6/22/2009
658
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
offspring outcomes in this study (i.e., 0.1 mg/mL, equivalent to 25.7-31.0 mg/kg-day) is
comparable to that which has been previously demonstrated to result in immune system
alterations and autoimmunity in adult MRL +/+ mice (i.e., 0.1 mg/mL, equivalent to 21 mg/kg-
day; Griffin et al., 2000b).
Another study that examined the effects of developmental exposure to TCE on the
MRL+/+ mouse was conducted by Peden-Adams et al. (2008). In this study, MRL/MpJ (i.e.,
MRL +/+) mice (unspecified number of dams/group) were exposed to TCE (solubilized with 1%
emulphore) in drinking water at levels of 0, 1,400, or 14,000 ppb from gestation day (GD) 0 and
continuing until the offspring were 12 months of age. TCE concentrations in the drinking water
were reported to be analytically confirmed. Endpoints evaluated in offspring at 12 months of age
included final body weight; spleen, thymus, and kidney weights; spleen and thymus lymphocyte
immunophenotyping (CD4 or CD8); splenic B-cell counts; mitogen-induced splenic lymphocyte
proliferation; serum levels of autoantibodies to dsDNA and glomerular antigen (GA),
periodically measured from 4 to 12 months of age; and urinary protein measures. Reported
sample sizes for the offspring measurements varied from 6 to 10 per sex per group; the number
of source litters represented within each sample was not specified. The only organ weight
alteration was an 18% increase in kidney weight in the 1,400 ppb males. Splenic CD4-/CD8-
cells were altered in female mice (but not males) at 1,400 ppm only. Splenic T-cell populations,
numbers of B220+ cells, and lymphocyte proliferation were not affected by treatment.
Populations of thymic T-cell subpopulations (CD8+, CD4-/CD8-, and CD4+) were significantly
decreased in male but not female mice following exposure to 14,000 ppb TCE, and CD4+/CD8+
cells were significantly reduced in males by treatment with both TCE concentrations.
Autoantibody levels (anti-dsDNA and anti-GA) were not increased in the offspring over the
course of the study, indicating that TCE did not contribute to the development of autoimmune
disease markers following developmental exposures that continued into adult life.
Overall, the studies by Peden-Adams et al. (2006, 2008 in press), Blossom and Doss
(2007), and Blossom et al. (2008), which examined various immunotoxicity endpoints following
exposures that spanned the critical periods of immune system development in the rodent, were
generally not designed to assess issues such as posttreatment recovery, latent outcomes, or
differences in severity of response that might be attributed to the early life exposures.
4.5.2.1.3 Intraperitoneal administration
Wright et al. reported that following 3 days of single intraperitoneal injections of TCE in
Sprague-Dawley rats at 0, 0.05, 0.5, or 5 mmol/kg/day and B6C3F1 mice at 0 or 10 mmol/kg/day,
natural killer (NK) cell activity was depressed in the rats at the mid- and high-dose levels, and in
the mice at the high dose level(Wright et al., 1991). Also at the highest dose levels tested,
6/22/2009
659
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
decreased splenocyte counts and relative spleen weight were observed in the rats and mice,
respectively. In vitro assays demonstrated treatment-related decreases in splenocyte viability,
inhibition of lipopolysaccharide-stimulated lymphocyte mitogenesis, and inhibited NK cell
activity suggesting the possibility that compromised immune function may play a role in
carcinogenic responses of experimental animals treated with TCE.
6/22/2009
660
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-8 Summary of TCE immunosuppression studies
Exposure Route/vehicle,
Duration, Dose
NOAEL; LOAEL3 Results
Reference, Species/strain
sex/number
Inhalation Exposure Studies
Single 1-hr exposure to all LOAEL: 200 ppm
dose groups; plus single 4-
hr exposure at 700 ppmb
0, 200, 500, 700, 1,000,
1,500, or 2,000 ppm
Single 3 hr exposure. Also, NOAEL: 2.6 ppm
3 hr/day on 5 days at lowest LOAEL: 5.2 ppm
dose
0,2.6,5.2, 10.6, 25.6, or 48
ppm
Marked transient -1 leukocyte counts at all exposure
levels 30-min after initiating exposure. At end of
exposure, 85% -1 leukocyte counts (33% -1
neutrophils, 40% -1 lymphocytes).
Challenged with Streptococcus zooepidemicus to
assess susceptibility to infection and Klebsiella
pneumoniae to assess bacterial clearance. For single
exposure: dose-related sig. | mortality at >5.2 ppm
over 14 days. Sig. [ in bactericidal activity at 10.6
ppm.
Hobaraetal., 1984
Dog, cross-bred, both sexes,
5/group
Aranyi et al., 1986
Mouse, CD 1 females, 4-5 wk
old, approx. 30 mice/group,
5-10 replications; for
pulmonary bactericidal
activity assay, 17-24
mice/group.
Single 3-hr exposure,
50-200 ppmc
4-wk, 6 hr/day, 5 days/wk NOAEL: 300 ppm
Challenged with Streptococcus zooepidemicus.
Dose-related t mortality, bacterial antiphagocytic
capsule formation, and bacterial survival. Dose-
related impairment of alveolar macrophages;
increased neutrophils in bronchoalveolar fluid at 3
days post-infection.
At 1,000 ppm, 64% -1 plaque-forming cell assay
Park et al., 1993 (abstract)
Mouse, CD1, (sex and
#/group not specified)
Woolhiser et al., 2006
6/22/2009
661
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
0, 100, 300, or 1,000 ppm LOAEL: 1,000
ppm
Oral Exposure Studies
Gavage in 10% emulphor, LOAEL: 24 mg/kg-
14 days, daily, 0, 24, or 240 day
mg/kg-day
Drinking water with 1% LOAEL: 0.1
emulphor, 4-6 months mg/kg-day
0,0.1, 1.0, 2.5, or
5.0 mg/mL
response
Sig. -I cell-mediated immune response to SRBC at
both dose levels
In females, humoral immunity at 2.5 and 5 mg/mL
TCE, whereas cell-mediated immunity -1 and bone
marrow stem cell colonization -1 at all four
concentrations. The males were relatively unaffected
after both 4 and 6 months.
Rat, Sprague-Dawley,
female, 16/group
Sanders et al., 1982
Mouse, CD-I, male,
9-12/group
Sanders et al., 1982
Mouse, CD-I, male and
female, 7-25/group
Gavage, 14 days, 0, 14.4, or NOAEL: 144
144 mg/kg-day chloral
hydrate
Drinking water, 90 days, 0,
0.07, or 0.7 mg/mL chloral
hydrate. (M: 0, 16, or 160
mg/kg-day; F: 0, 18, or 173
mg/kg-day)
mg/kg-day
NOAEL: 0.07
mg/mL
LOAEL: 0.7
mg/mL
No treatment-related effects
Sig. -I cell-mediated immune response (plasma
hemagglutination titers and spleen antibody-
producing cells of mice sensitized to SRBC) in
females at 0.7 mg/mL
Kauffmann et al., 1982
Mouse, CD1, male, 12/group
Kauffmann et al., 1982
Mouse, CD-I, male and
female, 15-20/group
Drinking water, From
mating to PND 21 or PND
56, (emulphor conc. not
provided)
0 (emulphor), 1, or 10 ppm
LOAEL: 1 ppm
At 10 ppm, J, body weight & length at PND 21. IgM
antibody response to SRBC challenge suppressed in
both and 9 pups at 10 ppm, and S pups at 1 ppm,
I in splenic CD4+CD8-T-cells. At 56 PND, striking
t in natural killer cell activity seen at both doses.
Adams et al., 2003 (abstract)
Mouse, B6C3F1, both sexes,
numbers of pups not stated.
Drinking water, From GD0 LOAEL: 1,400 ppb Suppressed PFC responses in both sexes and ages at Peden-Adams et al., 2006
6/22/2009
662
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
to 3 or 8 wks of age, 0,
1,400, or 14,000 ppb
Drinking water, From GD 0 LOAEL: 0.5
to 7-8 wks of age; 0, 0.5, or mg/mL
2.5 mg/mL
14,000 ppb, in males at both ages at 1,400 ppb, and in
females at 8 wks at 1,400 ppb. Numbers of spleen
B220+ cells -l at 3-wks at 14,000 ppb. Pronounced |
thymus T-cell populations at 8 wks.
At 0.5 mg/mL: Sig j postweaning weight; sig.f IFNy
produced by splenic CD4+ cells at 5-6 wks; sig j
splenic CD8+and B220+ lymphocytes; sig.t IgG2a
and histone; sig. altered CD4-/CD8- and CD4+/CD8+
thymocyte profile
At 2.5 mg/mL: Sig j postweaning weight; sig.t IFNy
produced by splenic CD4+ and CD8+ cells at 4-5
and 5-6 wks; sig J, splenic CD4+, CD8+, and B220+
lymphocytes; sig. altered CD4+/CD8+thymocyte
profile
Mouse, B6C3F1, dams and
both sexes offspring, 5
litters/group; 5-7 pups/group
at 3 wks; 4-5 pups/sex/
group at 8 weeks
Blossom and Doss, 2007
Mouse, MRL +/+, dams and
both sexes offspring, 3
litters/group; 8-12
pups/group;
Drinking water, From GD 0
toPND 42; 0 or 0.1
mg/mL; maternal dose =
25.7 mg/kg-day; offspring
PND 24-42 dose = 31.0
mg/kg-day
LOAEL: 0.1
mg/mL
At 0.1 mg/mL: at PND 20, sig. | thymocyte
cellularity and distribution, associated with sig. | in
thymocyte subset distribution; sig. | reactive oxygen
species generation in total thymocytes; sig. | in
splenic CD4+ T cell production of IFN-y and IL-2 in
females and TNF-a in males at PND 42
Blossom et al., 2008
Mouse, MRL +/+, dams and
both sexes offspring, 8
litters/group; 3-8 pups/group
Drinking water, From GD 0 LOAEL: 1,400 ppb At 1,400 ppb: splenic CD4-/CD8- cells sig.f in
to 12 months of age; 0 (1% females; thymic CD4+/CD8+ cells sig. -I in males;
emulphore), 1,400, or 18% \ in male kidney weight
14,000 ppb At 14,000 ppb: thymic T-cell subpopulations (CD8+,
CD4/CD8-, CD4+) sig. -I in males
Peden-Adams et al., 2008 (in
press)
Mouse, MRL +/+, dams and
both sexes offspring,
unknown # litters/group,
6/22/2009
663
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Intraperitoneal Injection Exposure Studies
3 days, single daily
injection, 0, 0.05, 0.5, or 5
mmol/kg/day
3 days, single daily
injection, 0 or 10
mmol/kg/day
NOAEL: 0.05
mmol/kg/day
LOAEL: 0.5
mmol/kg/day
LOAEL: 10
mmol/kg/day
6-10 offspring/sex/group
natural killer cell activity at 0.5 and 5 Wright et al., 1991
mmol/kg/day. -I splenocyte counts at 5 mmol/kg/day Rat, Sprague-Dawley,
-1 natural killer cell activity and -1 spleen weights at
10 mmol/kg/day.
Wright etal., 1991
Mouse, B6C3F1
Abbreviations: j, T = decreased, increased., sig. = statistically significant, GD = gestational day(s), PND = postnatal day(s)
a NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level) are based upon reported study findings.
b Inhalation, tracheal intubation under anesthesia
c Exact dose levels not specified.
6/22/2009
664
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.5.2.2 Hypersensitivity
Evidence of a treatment-related increase in delayed hypersensitivity response has been
observed in guinea pigs following dermal exposures with TCE and in mice following exposures
that occurred both during development and postnatally (Table 4.5-9).
In a modified guinea pig maximization test, Tang et al. evaluated the contact allergenicity
potential of TCE and three metabolites (trichloroacetic acid, trichloroethanol, and chloral
hydrate) in 4 animals (FMMU strain, sex not specified) per group (Tang et al., 2002). Edema
and erythema indicative of skin sensitization (and confirmed by histopathology) were observed.
Sensitization rates were reported to be 71.4% for TCE and 58.3% for trichloroacetic acid, as
compared to a reference positive control response rate (i.e., 100% for 2,4-dinitrochlorobenzene).
In this study, the mean response scores for TCE, trichloroacetic acid, and 2,4-
dinitrochlorobenzene were 2.3, 1.1, and 6.0, respectively. TCE was judged to be a strong
allergen and TCA was a moderate allergen, according to the criteria of Magnusson and Kligman
(Magnusson and Kligman, 1969). Trichloroethanol and chloral hydrate were not found to elicit a
dermal hypersensitivity response.
Immune-mediated hepatitis associated with dermal hypersensitivity reactions in the
guinea pig following TCE exposures was characterized by Tang et al. (2008). In this study,
FMMU strain female guinea pigs (5-6/group) were treated with intradermal injection of 0, 167,
500, 1,500, or 4,500 mg/kg TCE or with a dermal patch containing 0 or 900 mg/kg TCE and
sacrificed at 48 hours posttreatment. At the intradermal dose of 1,500 mg/kg, a significant
increase (p < 0.05) in serum aspartate aminotransferase (AST) level was observed. At 4,500
mg/kg, significantly (p < 0.01) increased alanine aminotransferase (ALT) and AST levels were
reported, and total protein and globulin decreased significantly (p < 0.05). Histopathological
examination of the liver revealed fatty degeneration, hepatic sinusoid dilation, and inflammatory
cell infiltration. No changes were observed at the intradermal doses of 500 mg/kg or below, or
the dermal patch dose of 900 mg/kg. A Guinea Pig Maximization Test was also conducted
according to the procedures of Magnusson and Kligman on 10 FMMU females/group, in which
the total TCE dosage from induction through challenge phases was below 340 mg/kg. TCE
treatment resulted in dermal erythema and edema, and the sensitization rate was 66% (i.e.,
classified as a strong sensitizer). Significant increases (p < 0.05) in ALT, AST, lactate
dehydrogenase (LH), and relative liver weight, and significant decreases (p < 0.05) in albumin,
IgA, and y-glutamyl transpeptidase (GGT) were observed. Additionally, hepatic lesions (diffuse
ballooning changes without lymphocyte infiltration and necrotic hepatocytes) were noted. It was
concluded that TCE exposure to guinea pigs resulted in delayed type hypersensitivity reactions
6/22/2009
665
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
with hepatic injury, that was similar to occupational medicamentosa-like dermatitis (OMLD)
disorders observed in human occupational studies.
Also, as indicated in Section 4.5.2.1.2 above, in a developmental immunotoxicity-type
study in B6C3F1 mice, administration of TCE in drinking water at dose levels of 0, 1,400, or
14,000 ppb from gestation day 0 through to 8 weeks of age resulted in an increased delayed
hypersensitivity response in 8-week old female offspring at both treatment levels and in males at
the high dose of 14,000 ppb (Peden-Adams et al., 2006).
In an in vitro study that evaluated a number of chlorinated organic solvents, non-purified
rat peritoneal mast (NPMC) cells and rat basophilic leukemia (RBL-2H3) cells were sensitized
with anti-dinitrophenol (DNP) monoclonal IgE antibody and then stimulated with DNP-
conjugated bovine serum albumin plus TCE (Seo et al., 2008). TCE enhanced antigen-induced
histamine release from NPMC and RBL-2H3 cells in a dose-related manner, and increased IL-4
and TNF-a production from the RBL-2H3 cells. In an in vivo study, i.p.-injected TCE was found
to markedly enhance passive cutaneous anaphylaxis reaction in antigen-challenged rats. These
results suggest that TCE increases histamine release and inflammatory mediator production from
antigen-stimulated mast cells via the modulation of immune responses; TCE exposure may lead
to the enhancement of allergic disease through this response.
6/22/2009
666
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-9 Summary of TCE hypersensitivity studies
Exposure Route/vehicle,
NOAEL; LOAEL1
Results
Reference, Species/strain
Duration, Dose
sex/number
Induction by single intradermal
Edema and erythema (confirmed by
Tang et al., 2002
injection, then challenge by dermal
histopathology) indicative of skin
Guinea pig, FMMU strain, sex
application at 21 days
sensitization for TCE (strong sensitizer)
not specified, 4/group
0 or 0.1 mL induction; 0 or 0.2 mL
and TCA (moderate sensitizer)
challenge
TCE, TCA, TCOH, and chloral
hydrate
Intradermal injection, 0, 167, 500,
Intradermal NOAEL: 500 mg/kg
Intradermal injection: At 1,500 mg/kg: Sig.
Tang et al., 2008
1,500, or 4,500 mg/kg
Intradermal LOAEL: 1,500
t AST; at 4,500 mg/kg, sig. | ALT and
Guinea pig, FMMU strain,
mg/kg
AST, sig. i total protein and globulin; fatty
female, 5-6/group for
Dermal patch, 0 or 900 mg/kg
degeneration of liver
intradermal/dermal patch study,
Dermal patch NOAEL: 900
10/group for hypersensitivity
Hypersensitivity: total dose from
mg/kg
Dermal patch: no effects of treatment
study, female
induction through challenge <340
mg/kg.
Hypersensitivity: sensitization rate of 66%
(strong sensitizer), with edema and
erythema; sig. | ALT, AST, and LH; sig. |
relative liver weight; sig. j albumin, IgA,
and GGT; hepatic lesions (ballooning
changes)
Drinking water, from GD0 to 8
LOAEL: 1,400 ppb
Sig. | swelling of foot pad in females at
Peden-Adams et al., 2006
wks of age
1,400 and in both sexes at 14,000 ppb.
Mouse, B6C3F1, both sexes, 5
6/22/2009
667
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
0, 1,400, or 14,000 ppb litters/ group; 4-5
pups/sex/group at 8 weeksb
Abbreviations: j, T = decreased, increased, sig. = statistically significant, GD = gestational day(s)
a NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level) are based upon reported study findings.
b Subset of immunosuppression study.
6/22/2009
668
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.5.2.3 Autoimmunity
A number of studies have been conducted to examine the effects of TCE exposure in
mouse strains (i.e., MRL +/+, MRL -lpr, or NZB x NZW) which are all known to be genetically
susceptible to autoimmune disease. The studies have demonstrated the potential for TCE to
induce autoimmune disease (as demonstrated in Table 4.5-10 which summarizes those studies
which assessed serology, ex vivo assays of cultured splenocytes, and/or clinical or
histopathology). These and other studies conducted in susceptible mouse strains have proven to
be useful tools in exploring various aspects of the mode of action for this response.
Khan et al. used the MRL +/+ mouse model to evaluate the potential for TCE and one of
its metabolites, dichloroacetyl chloride (DCAC) to elicit an autoimmune response(Khan et al.,
1995). Female mice (4-5/group) were dosed by intraperitoneal injection with 10 mmol/kg TCE
or 0.2 mmol/kg DCAC every 4th day for 6 weeks and then sacrificed. Spleen weights and IgG
were increased. ANA and anti-ssDNA antibodies were detected in the serum of TCE- and
DC AC-treated mice; anti-cardiolipin antibodies were detected in the serum of DC AC-treated
mice. A greater magnitude of response observed with DCAC treatment suggested that the
metabolite may be important to the mechanism of TCE-induced autoimmunity.
Other studies in female MRL +/+ mice (8/group) examined exposure via drinking water. In one
of these studies, mice were treated with 2.5 or 5.0 mg/mL (455 or 734 mg/kg-day) TCE in
drinking water for up to 22 weeks (Gilbert et al., 1999; Griffin et al., 2000a). Serial sacrifices
were conducted at weeks 4, 8, and 22. Significant increases in ANA and total serum
immunoglobulin were found at 4 weeks of TCE treatment (indicating an autoimmune response),
but not at 32 weeks. Increased expression of the activation marker C44 on splenic CD4+ cells
was observed at 32 weeks. In addition, at 4 and 32 weeks, splenic T cells from treated mice
secreted more IFN-y than control T cells (significant at 0.5 and 2.5 mg/mL), consistent with a T-
helper type 1 (Thl) immune or inflammatory response. By 22 weeks of TCE treatment, a
specific immune serum antibody response directed against dichloroacetylated proteins was
activated n hepatic tissues, indicating the presence of protein adducts. There was a slight, but
significant, increase in serum alanine aminotransferase levels at 32 weeks at 0.5 mg/mL.
Histopathological evaluation at 32 weeks revealed extensive hepatic lymphocytic cell infiltration
at 0.5 and 2.5 mg/mL; all treated groups contained significantly more hepatocyte reactive
changes (i.e., presence of multinucleated hepatocytes, variations in hepatocyte morphology, and
hepatocytes in mitosis) than controls.
In a subsequent study which assessed occupationally relevant concentrations, TCE was
administered to female MRL +/+ mice (8/group) in drinking water at treatment levels of 0.1, 0.5,
or 2.5 mg/mL (21, 100, or 400 mg/kg-day) for 4 and 32 weeks (Griffin et al., 2000b). At 4
6/22/2009
669
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
weeks, significant increases in serum antinuclear antibody levels were observed at 0.1 and 0.5
mg/kg-day; at 32 weeks, the effects were observed at all three treatment levels. A dose-related
increase in the percentage of activated CD4+ T cells in spleens and lymph nodes of treated mice
was observed at 32 weeks, and the CD4+ T cells were found to secrete Thl-type cytokines at 4
and 32 weeks.
A similar response was observed by Cai et al. following chronic (48 weeks) exposure of
TCE to female MRL +/+ mice (5/group) in drinking water at 0 or 0.5 mg/mL (approximately 60
(j,g/g/day) (Cai et al., 2008). After 11 weeks of treatment, a statistically significant decrease in
body weight gain was observed. After 24 weeks of exposure serum ANA were consistently
elevated in treated mice as compared to control, although statistical significance was not
achieved. Apparent treatment-related effects on serum cytokines included decreased IL-6 after
36 and 48 weeks, decreased TNF-a after 48 weeks, and increased G-CSF after 36 weeks of
treatment. After 36 weeks of treatment, ex vivo cultured splenocytes secreted higher levels of
IFN-y than control splenocytes. Although there were no observed effects on serum
aminotransferase liver enzymes at termination, statistically significant incidences of hepatocytic
necrosis and leukocyte infiltration (including CD3+ T lymphocytes) into liver lobules were
observed in treated mice after 48 weeks of exposure. Hepatocyte proliferation was also
increased. TCE treatment for 48 weeks also induced necrosis and extensive infiltration of
leukocytes in the pancreas, infiltration of leukocytes into the perivascular and peribronchial
regions of the lungs, and thickening of the alveolar septa in the lungs. At 36 and 48 weeks of
exposure, massive perivascular infiltration of leukocytes (including CD3+ T lymphocytes) was
observed in the kidneys, and immunoglobulin deposits were found in the glomeruli.
To examine the role of metabolic activation in the autoimmune response, Griffin et al.
(2000c) treated MRL +/+ mice with 2.5 mg/mL (300 mg/kg-day) TCE in drinking water for 4
weeks (Griffin et al., 2000c). Immune responses were examined in the presence or absence of
subcutaneous doses of 200 mg/kg-day diallyl sulfide, a specific inhibitor of CYP2E1 which is
known to be a primary P450 cytochrome that is active in TCE metabolism. With diallyl sulfide
co-treatment that resulted in a decreased level of CYP2E1 apoprotein in liver microsomes, the
enhanced mitogen-induced proliferative capacity of T cells was inhibited and the reduction in IL-
4 levels secreted by CD4+ T cells was reversed for TCE-treated MRL +/+ mice. This study
suggests that metabolism of TCE by CYP2E1 is responsible, at least in part, for the treatment-
related CD4+ T cell alterations.
The TCE metabolite, trichloroacetaldehyde (TCAA) or trichloroacetaldehyde hydrate
(TCAH), was also evaluated in MRL +/+ mice (Blossom et al., 2007; Blossom and Gilbert,
2006; Gilbert et al., 2004) in order to determine if outcomes similar to the immunoregulatory
effects of TCE would be observed, and to attempt to further characterize the role of metabolism
6/22/2009
670
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
in the mode of action for TCE. At concentrations ranging from 0.04 to 1 mM TCAA stimulated
proliferation of murine Thl cells treated with anti-CD3 antibody or antigen in vitro. At similar
concentrations, TCAA induced phenotypic alterations consistent with upregulation of CD28 and
downregulation of CD62L in cloned memory Thl cells and DC4+ T cells from untreated MRL
+/+ mice. Phosphorylation of activating transcription factor 2 (ATF-2) and c-Jun (two
components of the activator protein-a transcription factor) was, also, observed with TCAA-
induced Thl cell activation. Higher concentrations of TCAA formed a Schiff base on T cells,
which suppressed the ability of TCAA to phosphorylate ATF-2. These findings suggested that
TCAA may promote T-cell activation by stimulating the mitogen-activated protein (MAP)
kinase pathway in association with Schiff base formation on T-cell surface proteins (Gilbert et
al., 2004).
In order to determine whether metabolites of TCE could mediate the immunoregulatory
effects previously observed with TCE treatment (i.e., the generation of lupus and autoimmune
hepatitis, associated with activation of IFN-y-producing CD4+ T cells), Blossom et al. (2004)
administered TCE metabolites, TCAH and TCA, to MRL +/+ mice (6-8/group) in drinking
water for 4 weeks. Drinking water concentrations were 0, 0.1, or 0.9 mg/mL; average daily
doses were calculated as 0, 24, or 220 mg/kg-day for TCAH and 0, 27, or 205 mg/kg-day for
TCA. These treatment levels were considered to be physiologically relevant and to reflect
occupational exposure. A phenotypic analysis of splenic and lymph node cells, cytokine profile
analysis, evaluation of apoptosis in CD4+ T cells, and examination of serum markers of
autoimmunity (anti-ssDNA, anti-histone, or ANA) were conducted. Exposure to TCAH or TCA
at both treatment levels was found to promote CD4+ T cell activation, as shown by significant
(p < 0.05) increases in the percentage of CD62L10 CD4+ T cells in the spleens and lymph nodes
of the MRL +/+ mice. Increased levels of IFN-y were secreted by CD4+ T cells from mice
treated by TCAH and TCA. No significant changes in body weight were observed; spleen
weights were similar between control and treated mice with the exception of a significant
decrease in spleen weight from mice treated with 0.9 mg/mL TCA. Liver and kidney histology
were not affected, and serum alanine aminotransferase levels were similar for control and treated
mice. A generalized trend towards an increase in serum autoantibodies (anti-ssDNA) was
observed in TCAH-treated mice, and slight but significant increases in anti-histone and anti-
nuclear antibody production were observed in mice treated with 0.9 mg/mL-day TCAH.
The autoimmune response of female MRL +/+ mice to dichloroacetyl chloride (DCAC),
a metabolite of TCE, and to dichloroacetic anhydride (DCAA) a similar acylating agent, was
evaluated by Cai et al. (2006). Six mice/group were injected intraperitoneally, twice weekly for
6 weeks, with 0.2 mmol/kg DCAC or DCAA in corn oil. Body weight gain was significantly
decreased after 5 or 6 weeks treatment with DCAC and DCAA. DCAC treatment resulted in
6/22/2009
671
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
significant increases in total serum IgG (77% increase over control) and IgGl (172% increase
over control), as well as the induction of DC AC-specific IgG and IgGl. Serum IgM levels were
significantly decreased by 25% and 18% in DCAC and DCAA-treated mice, respectively. IgE
levels were increased 100% over controls in DCDC-treated mice. Of eight Thl/Th2 cytokines
measured, only IL-5 was decreased in DCAC- and DCAA-treated mice. Serum ANA were
detected in both DCAC- and DCAA-treated mice. Treatment-related increases in cytokine and
chemokine secretion in cultured splenocytes were observed for DCAC and DCAA (IL-1, G-CSF,
KC, IL-3, and IL-6). DCAC-treated splenocytes also secreted more IL-17 and IFN-a than
controls. Histopathological changes were observed in the spleens of DCAC and DCAA-treated
mice (lymphocyte population increases in the red pulp). With both DCAC and DCAA treatment,
the alveolar septa were thickened in the lungs, moderate levels of lymphocytic interstitial
infiltrates were present in tissues, and alveolar capillaries were clogged with erythrocytes. These
findings were attributed both to the predisposition of the MRL +/+ mice towards autoimmune
disease, and to the treatment-related induction of autoimmune responses.
Fas-dependant activation-induced cell death leading to autoimmune disease has been
shown to be related to impaired Fas or FasL ligand expression in humans and mice, and defects
in the Fas-signaling pathways have been described in autoimmune disease models. The study by
Blossom and Gilbert examined the effects of TCAH on Fas-dependent autoimmune cell death
(Blossom and Gilbert, 2006). In this study, TCAH 1) inhibited apoptosis of antigen-activated
cells, 2) did not protect CD4+ T cells from Fas-independent apoptosis, 3) did not inhibit
autoimmune cell death induced by direct engagement of the Fas receptor, 4) inhibited the
expression of FasL but not Fas on the surface of activated CD4+ T cell, 5) increased release of
FasL from CD4+ cells in a metalloprotein-dependent manner, and 6) increased metalloprotein
MMP-7 expression.
Gilbert et al. (2006) studied the effect of treatment on apoptosis in CD4+ T-lymphocytes
isolated from MRL +/+ female mice that had been exposed to TCE (0, 0.1, 0.5, or 2.5 mg/mL) in
the drinking water for 4 or 32 weeks or to TCAH (0.1, 0.3, or 0.9 mg/mL) in drinking water for 4
or 40 weeks. After only 4 weeks, decreased activation-induced apoptosis was associated with
decreased FasL expression in the CD4+ T-cells, suggesting that TCE- and TCAH-induced
autoimmune disease was promoted through suppression of the process that would otherwise
delete activated self-reactive T-lymphocytes. By 32 weeks of treatment, TCE had induced
autoimmune hepatitis, which was associated with the promotion of oxidative stress, the
formation of liver protein adducts, and the stimulated production of antibodies to those adducts.
TCAH-treated mice did not exhibit autoimmune hepatitis by 40 weeks, but developed a dose-
dependant alopecia and skin inflammation (Blossom et al., 2007). TCAH appeared to modulate
the CD4+ T-cell subset by promoting the expression of an activated/effector phenotype with an
6/22/2009
672
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
increased capacity to secrete the proinflammatory cytokine IFN-y. A 4-week exposure to TCAH
attenuated activation-induced cell death and the expression of the death receptor Fas in CD4+
cells; which a 40-week exposure did not. Differences in response were tentatively attributed to
higher levels of metalloproteinases (specifically MMP-7) at 4-weeks of treatment, suggesting a
possible mechanism for the promotion of skin pathology by TCAH.
The role of protein adduct formation in autoimmune response has been pursued by
various researchers. Halmes et al. administered a single i.p. dose of TCE in corn oil to male
Sprague-Dawley rats (2/group) at 0 or 1,000 mg/kg (Halmes et al., 1997). Using antiserum that
recognizes TCE covalently bound to protein, a single 50 kDa microsomal adduct was detected by
Western blot in livers of treated rats. Using affinity chromatography, a 50 kDa dichloroacetyl
protein was also isolated from rat plasma. The protein was reactive immunochemically with
anti-CYP2El antibodies. The data suggest that the protein adduct may be CYP2E1 that has been
released from TCE-damaged hepatocytes.
Cai et al. examined the role of protein haptenization in the induction of immune
responses (Cai et al., 2007). In this study, MRL +/+ mice were immunized with albumin adducts
of various TCE reactive intermediates of oxidative metabolism. Serum immunoglobulins and
cytokine levels were measured to evaluate immune responses against the haptenized albumin.
Antigen-specific IgG responses (subtypes: IgGl, IgG2a, and IgG2b) were found. Serum levels
of G-CSF were increased in immunized mice, suggesting macrophage activation. Following
immunization with formyl-albumin, lymphocyte infiltration in the hepatic lobule and portal area
was increased. This study suggests that proteins that are haptenized by metabolites of TCE may
act as antigens to induce humoral immune responses and T cell-mediated hepatitis.
A possible role for oxidative stress in inflammatory autoimmune disease was proposed by
Khan et al. (2001). A study was performed in which female MRL +/+ mice were treated with 10
mmol/kg TCE or 0.2 mmol/kg dichloroacetyl chloride (DCAC) via intraperitoneal injection
every 4th day for 2, 4, 6, or 8 weeks. Anti-malondialdehyde serum antibodies, a marker of lipid
peroxidation and oxidative stress, were measured and were found to increase by 4 weeks of
treatment, marginally for TCE and significantly for DC AC. It was reported that anti-
malondialdehyde antibodies has also been found to be present in the serum of systemic lupus
erythematosus-prone MRL-lpr/lpr mice.
In another study that addressed the association of oxidative and nitrosative stress, and the
role of lipid peroxidation and protein nitration, in TCE-mediated autoimmune response, Wang et
al. treated female MRL +/+ mice with 0.5 mg/mL TCE in drinking water for 48 weeks (Wang et
al., 2007b). The formation of antibodies in the serum to lipid peroxidation-derived aldehyde
(LPDA) protein adducts was evaluated. With TCE treatment, the serum levels of anti-
malondialdehyde and anti-4-hydroxynonenal protein adduct antibodies, iNOS, and nitrotyrosine
6/22/2009
673
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
were increased. These were associated with increases in anti-nuclear-, anti-ssDNA- and anti-
dsDNA antibodies. The involvement of lipid peroxidation-derived aldehyde protein adducts in
TCE autoimmunity was further explored, using female MRL +/+ mice that were administered
i.p. injections of TCE at 10 mmol/kg, either every 4th day for 6 or 12 weeks (Wang et al., 2007a)
or once per week for 4 weeks (Wang et al., 2008). Significant increases in malondialdehyde and
4-hydroxynonenal protein adducts, as well as significant induction of specific antibodies directed
against these antigens were observed in both studies. Wang et al. also demonstrated a significant
proliferation of CD4+ T cells in TCE-treated mice, and splenic lymphocytes from TCE-treated
mice released more IL-2 and IFN-y when stimulated with MDA- or HNE-adducted mouse serum
albumin (Wang et al., 2008). Overall, the result of these studies suggest a role for lipid
peroxidation aldehydes in the induction and/or exacerbation of autoimmune response in the MRL
+/+ animal model, and the involvement of Thl cell activation.
In studies conducted in other rodent strains, less consistent outcomes have been observed.
Inhalation exposure of an autoimmune-prone strain of male mice (MRL-lpr/lpr) to 0, 500, 1,000,
or 2,000 ppm TCE for 4 hr/day, 6 days/week, for 8 weeks resulted in depressed serum IgG levels
and increased numbers of lymphoblastoid cells (Kaneko et al., 2000). Also at 2,000 ppm,
changes in T-cell helper to suppressor cell ratios were observed. At histopathological evaluation,
dose-dependent inflammation and associated changes were noted in the liver at >500 ppm,
hyperplasia of the lymphatic follicles of the spleen and splenomegaly were observed at >500
ppm, and the spleen exhibited the development of an immunoblastic-cell-like structure at 1,000
ppm.
A 26-week drinking water study of TCE in NZB x NZW (NZBWF1) autoimmune-prone
mice demonstrated an increase in anti-dsDNA antibodies at 19 weeks and at 32 and 34 weeks in
the 1,400 ppb group, and increased kidney disease at 14,000 ppb (i.e., increased proteinuria at 20
weeks; increased renal pathology scores at termination, based upon glomerular proliferation,
inflammation, and necrosis) (Gilkeson et al., 2004).9 Also in that study, a small increase in anti-
dsDNA antibody production, without kidney disease, was observed in B6C3F1 mice, with
statistically significant (p < 0.05) or borderline (p = 0.07) effects seen in the 1400 ppb group at
observations between 32 and 39 weeks of age, and in the 14,000 ppb group at observations
between 26 and 39 weeks of age.
Keil et al. (2009) also assessed the effects of TCE exposure on NZWBF1 mice,
comparing the responses to those of TCE-exposed B6C3F1 mice, which are not autoimmune
9 The study was reported in symposium proceedings Dose levels cited in the proceedings were incorrect; however,
corrections were provided by personal communication from Margie Peden-Adams [Medical University of South
Carolina] to Glinda Cooper [U.S. EPA] on 13 August 2008, and dose levels are correctly reported here.
6/22/2009
674
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
prone (Keil et al. 2009). In this study, groups of NZWBF1 and B6C3F1 female mice (10/dose
level) were administered 0, 1400, or 14,000 ppb TCE in the drinking water. Treatment was
initiated at 9 weeks of age and continued until 36 weeks of age for the NZBWF1 and until 39
weeks of age for the B6C3F1 mice. Body weight; spleen, thymus, liver, and kidney weight;
spleen and thymus cellularity; and renal pathology were assessed. Splenic lymphocyte
proliferation, autoantiboidy production (anti-dsDNA, anti-ssDNA, and anti-glomerular), total
serum IgG, NK cell activity, and mitogen-induced lymphocyte proliferation were conducted.
Administration of TCE did not result in alterations to NK cell activity or to T- or B-cell
proliferation in either strain of mice. In the NZBWF1 mice, there was little evidence of an
increase or of an acceleration in ss-DNA antibody production with TCE exposure, but as was
seen in the earlier study by these investigators (Gilkeson et al., 2004), ds-DNA antibodies were
increased at 19 weeks and at 32-34 weeks in the 1,400 ppb group. However, anti-GA levels
were increased in NZBWF1 mice early in the study, returning to control levels by 23 weeks of
age. In the B6C3F1 mice the number of activated T-cells (CD4++/CD44+) was increased
(significantly at 14000 ppm; p<0.05) and thymus weights were significantly decreased (p<0.05)
in a dose-responsive manner. Renal pathology (as indicated by renal score based on assessment
of glomerular inflammation, proliferation, crescent formation and necrosis) was significantly
increased (p<0.05) at 1400 ppm. Also in the B6C3F1 mice, autoantibodies to dsDNA were
increased relative to controls beginning at 26 weeks in the 14,000 ppb group and at 32 weeks of
age in the 1,400 ppb group; increases in anti-ssDNA antibodies were seen in both groups at 32
weeks. Anti-glomerular autoantibodies (anti-GA) were not affected in B6C3F1 mice. In
summary, the authors concluded that this study showed that 27-30 weeks of TCE drinking water
administration to NZBWF1 (autoimmune-prone) mice did not contribute to the progression of
autoimmune disease, while similar administration to B6C3F1 (non-autoimmune-prone) mice
increased the expression of a number of markers that are associated with autoimmune disease.
This study is important in that it demonstrates that autoimmune responses to TCE exposure in
animal models are not solely dependant upon a genetic predisposition to autoimmune disease.
White et al. conducted a study in female Brown Norway rats, which have been shown to
be susceptible to development of chemically-induced IgE mediated glomerulonephritis that is
similar to the nephritic damage seen in systemic lupus erythematosus (White et al., 2000). TCE
administered by gavage 5 days/week at 100, 200, or 400 mg/kg did not increase in IgE levels
after 6 weeks exposure, or after an additional challenge with 1 mg/kg mercuric chloride (HgCl2).
Several studies have examined the potential for autoimmune response following oral
exposures during pre- and postnatal immune system development, as described in Section
4.5.2.1.2 above. Peden-Adams et al. conducted two such studies. In the first study, B6C3F1
mice were treated with either 1,400 or 14,000 ppb TCE in drinking water from gestation day 0 to
6/22/2009
675
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
postnatal week 8 (Peden-Adams et al., 2006). No treatment-related increases in serum anti-ds-
DNA antibody levels were observed in the 8-week old offspring, although it is noted that the
mouse strain used in the experiment is not an autoimmune-prone animal model. A more recent
study (Peden-Adams et al., 2008) exposed pregnant MRL +/+ mice to TCE in drinking water at
levels of 0, 1,400, or 14,000 ppb from GD 0 and continued the exposures until the offspring were
12 months of age. Consistent with the findings of the 2006 publication, autoantibody levels
(anti-dsDNA and anti-GA) were not increased in the offspring over the course of the study.
Contrasting with these negative studies, the lupus-prone MRL +/+ mouse model was utilized in
two additional drinking water studies with developmental exposures in which there was some
indication of a positive association between developmental exposures to TCE and the initiation
of autoimmune disease. Blossom and Doss (2007) administered TCE to pregnant MRL +/+ mice
in drinking water at levels of 0, 0.5, or 2.5 mg/mL and continued administration to the offspring
until approximately 7-8 weeks of age. TCE exposure induced a dose-dependent increase in T-
lymphocyte IFN-y in peripheral blood at 4-5 weeks of age, but this effect was not observed in
splenic T-lymphocytes at 7-8 weeks of age. Serum anti-histone autoantibodies and total IgG2a
were significantly increased in the TCE-treated offspring; however, histopathological evaluation
of the liver and kidneys did not reveal any treatment-related signs of autoimmunity. In a study
by Blossom et al. (2008), pregnant MRL +/+ mice were administered TCE in the drinking water
at levels of 0 or 0.1 mg/mL from GD 0 through lactation, and continuing postweaning in the
offspring until postnatal day 42. Significant treatment-related increases in pro-inflammatory
cytokines (IFN-y and 11-2 in males and TNF-a in both sexes) produced by splenic CD4+ T-cells
were observed in PND 42 offspring.
In summary, TCE treatment induces and exacerbates autoimmune disease in genetically
susceptible strains of mice, and has also been shown to induce signs of autoimmune disease in a
non-genetically predisposed strain. Although the mechanism for this response is not fully
understood, a number of studies have been conducted to examine this issue. The primary
conclusion to date is that metabolism of the TCE to its chloral or dichloroacetic acid metabolites
is at least partially responsible for activating T cells or altering T cell regulation and survival
associated with polyclonal disease in susceptible mice strains.
6/22/2009
676
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-10. Summary of autoimmune-related studies of TCE and metabolites in mice and rats (by sex, strain, and route of
exposure)3
Results
No./group, Vehicle, Dose, NOAEL;
Duration LOAELb
Serology
Ex vivo Assays of
Cultured Splenocytes
Clinical and
Histopathology
Reference
Autoimmune-prone: Female MRL +/+ Mice, Drinking Water
8 per group, 0, 2.5, or
5 mg/mL TCE (average 0,
455, or 734 mg/kg-day), 4,
8 or 22 weeks
LOAEL: 2.5
mg/mL
Increased ANA at 4 and
8 weeks, no difference
between groups at 22
weeks
Increased activated CD4+
T cells and IFN-y
secretion across doses at 4
weeks, these effects were
reversed at 22 weeks;
decreased IL-4 secretion
(4 and 22 weeks)
No evidence of liver or
renal damage, based on
serum alanine
aminotransferase, sorbitol
dehydrogenase, and blood
urea nitrogen.
Griffin et al.
(2000a)
8 per group, 0, 0.1, 0.5, or
2.5 mg/mL TCE (0,21,
100, or 400 mg/kg-day), 4
or 32 weeks
LOAEL: 0.1
mg/mL
Increased ANA in all
treated groups at 4
weeks, but not at 32
weeks
Increased activated CD4+
T cells (32 weeks), IFN-y
secretion (4 and 32
weeks), no effect on IL-4
secretion
Extensive hepatic
mononuclear cellular
infiltrate in 0.5 and 2.5
mg/mL groups, and
hepatocyte reactive
changes in all treated
groups at 32 weeks
Griffin et al.
(2000b)
6-8 per group, 0, 0.1, or
0.9 mg/mL
trichloroacetaldehyde
hydrate (0, 24, or 220
mg/kg-day) or
trichloroacetic acid (0, 27,
or 205 mg/kg-day), 4
weeks
LOAEL: 0.1
mg/mL
Increased ANA and
anti-histone antibodies
at 0.9 mg/mL
trichloroacetaldehyde
hydrate0
Increased activated CD4+
T cells at 0.1 and 0.9 g/mL
doses of both metabolites.
At 0.9 mg/mL, increased
IFN-y secretion, no effect
on IL-4 secretion
No evidence of liver of
kidney damage, based on
serum alanine
aminotransferase, liver
and kidney histology.
Blossom et
al. (2004)
6/22/2009
677
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-10. Summary of autoimmune-related studies of TCE and metabolites (by sex, strain, and route of exposure),
continued3
Results
No./group, Vehicle, Dose, NOAEL;
Duration LOAELb
Serology
Ex vivo Assays of
Cultured Splenocytes
Clinical and
Histopathology
Reference
8 per group, 0, 0.1, 0.3, or
0.9 mg/mL
trichloroacetaldehyde
hydrate (0, 13, 46, or 143
mg/kg-day), 40 weeks
5 per group, 0 or 0.5
mg/mL TCE (mean 60
pg/g-day), 48 weeks
LOAEL: 0.9
mg/mL
LOAEL: 0.5
mg/mL
Slightly suppressed
anti-ssDNA, anti-
dsDNA, and anti-
histone antibody
expression; differences
not statistically
significant
Increased activated CD4+
T cells and increased INF-
y secretion, no effect on
IL-4 secretion
Increased ANA after 24
weeks but not
statistically significant
Increased INF-y secretion
after 36 weeks but not
statistically significant
Autommune-prone: Male and Female Offspring MRL +/+ Mice, Drinking Water
3 litters/group, 8-12 LOAEL: 0.5 Increased anti-histone Dose-dependant increase
offspring/group; 0, 0.5, or mg/mL antibodies and total
2.5 mg/mL, GD 0 to 7-8 IgG2a in treated groups
wks of age
in IFN-y secretion at 4-5
weeks of age but not 7-8
weeks of age
Diffuse alopecia, skin
inflammation and
ulceration, mononuclear
cell infiltration, mast cell
hyperplasia, dermal
fibrosis. Statistically
significant increase at 0.9
mg/mL dose group, but
also increased at lower
doses. No liver or kidney
histopathology effects
seen.
Hepatic necrosis;
hepatocyte proliferation;
leucocyte infiltrate in the
liver, lungs, and kidneys;
no difference in serum
aminotransferase liver
enzymes
No histopathological
effects in liver or kidneys
Blossom et
al. (2007)
Cai et al.
(2008)
Blossom and
Doss
(2007)
6/22/2009
678
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-10. Summary of autoimmune-related studies of TCE and metabolites (by sex, strain, and route of exposure),
continued3
Results
No./group, Vehicle, Dose, NOAEL;
Duration LOAELb
Serology
Ex vivo Assays of
Cultured Splenocytes
Clinical and
Histopathology
Reference
8 litters/group, 8-12
offspring/group; 0 or 0.1
mg/mL; maternal dose =
25.7 mg/kg-day; offspring
PND 24-42 dose = 31.0
mg/kg-day; GD 0 to PND
42
LOAEL: 0.1
mg/mL
Not evaluated
Increased IFN-y and IL-2
in females, increased
TNF- a in both sexes
Not evaluated
Blossom et
al. (2008)
Unknown # litters/group, NOAEL: 1400
6-10 offspring/sex/group; ppb
0 (1% emulphore), 1400,
or 14,000 ppb; GD 0 to 12
months of age
Autoimmune-prone: Female MRL +/+ Mice,
4-5 per group, 0 (corn oil), LOAEL: 10
10 mmol/kg TCE, or 0.2 mmol/kg TCE,
mmol/kg dichloroacetyl 0.2 mmol/kg
chloride, every 4th day for dichloroacetyl
6 weeks chloride
No increase in
autoantibody levels
Not evaluated
Not evaluated
Intraperitoneal Injection
In both groups, not evaluated
increased ANA and
anti-ssDNA antibodies.
In dichloroacetyl
chloride group, anti-
cardiolipin antibodies.
No difference in anti-
histone, -Sm, or -DNA
antibodies.
not evaluated
Peden-
Adams et
al. (2008)
Khan et al.
(1995)
6/22/2009
679
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-10. Summary of autoimmune-related studies of TCE and metabolites (by sex, strain, and route of exposure),
continued3
Results
No./group, Vehicle, Dose, NOAEL;
Duration LOAELb
Serology
Ex vivo Assays of
Cultured Splenocytes
Clinical and
Histopathology
Reference
6 per group, 0 (corn oil),
0.2 mmol/kg
dichloroacetyl chloride, or
0.2 mmol/kg
dichloroacetic anhydride, 2
times per week for 6
weeks
LOAEL: 0.2
mmol/kg TCE,
0.2 mmol/kg
dichloroacetic
anhydride
In both treated groups,
increased ANA
Autoimmune-prone: Female NZB x NZW Mice, Drinking Water
6 per group, 0, 1400, or LOAEL: 1400 Increased anti-dsDNA
ppb
14,000 ppb TCE£
weeks exposure
27
10 per group, 0, 1400, or
14,000 ppb TCEf, 27
weeks exposure
LOAEL: 1400
ppb
antibodies at 19 weeks
and at 32-32 weejks in
the 1,400 ppb group
Increased anti-dsDNA
antibodies at 19 weeks
and at 32-32 weejks in
the 1,400 ppb group
In both treated groups,
increased IL-la, IL-1B, IL-
3, IL-6, IFN-y,
granulocyte colony
stimulating factor (G-CSF)
and keratinocyte-derived
chemokine (KC) secretion;
decreased IL-5. In
dichloroacetyl chloride
group, increased IL-17 and
INF-a.d
Not evaluated
No effect on splenocyte
NK activity
In both treated groups,
increased lymphocytes in
spleen, thickening of
alveolar septa with
lymphocytic interstitial
infiltration
Cai et al.
(2006)
At 14,000 ppb, proteinuria
increased beginning at 20
weeks; renal pathology
scores increased, no
evidence of liver disease
No effect on renal
pathology score; liver
disease not examined
Gilkeson et
al. (2004)
Kiel et al.
(2009)
6/22/2009
680
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-10. Summary of autoimmune-related studies of TCE and metabolites (by sex, strain, and route of exposure),
continued3
Results
No./group, Vehicle, Dose, NOAEL;
Duration LOAELb
Serology
Ex vivo Assays of
Cultured Splenocytes
Clinical and
Histopathology
Reference
Autoimmune-prone: Male MRL - IprApr Mice, Inhalation
5 per group, 0, 500, 1000, LOAEL: 500
or 2000 ppm TCE, 4 hours ppm
per day, 6 days per week, 8
weeks
Autoimmune-inducible: Female Brown Norway Rat, Gavage
6-8 per group, 0, 100, 200, NOAEL 500 Not reported8
400 mg/kg, 5 days per mg/kg
week, 6 weeks followed by
1 mg/kg HgCl2 challenge
Non-autoimmune-prone: Female B6C3F1 Mice, Drinking Water
Not evaluated
6 per group, 0, 1400, or
14,000 ppb TCEef, 30
weeks exposure
LOAEL: 1400
ppb
Anti-dsDNA
increased in 1400 ppb
group beginning at
age 32 weeks and in
the 14,000 ppb group
beginning at age 26
weeks
No effect on splenocyte
NK activity
At > 500 ppm, dose-
related liver inflammation,
splenomegaly and
hyperplasia of lymphatic
follicles; at 1000 ppm,
immunoblastic cell
formation in lymphatic
follicles, no changes in
thymus
Not evaluated
No renal disease observed
Kaneko et al.
(2000)
White et al.
(2000)
Gilkeson et
al. (2004)
6/22/2009
681
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5-10. Summary of autoimmune-related studies of TCE and metabolites (by sex, strain, and route of exposure),
continued3
Results
No./group, Vehicle, Dose, NOAEL;
Duration LOAELb
Serology
Ex vivo Assays of
Cultured Splenocytes
Clinical and
Histopathology
Reference
10 per group, 0, 1400,
14,000 ppb TCEf, 30
weeks exposure
or
LOAEL: 1400
ppb
Anti-dsDNA
increased beginning
at 26 weeks in the
14,000 ppb group and
at 32 weeks of age in
the 1,400 ppb group;
increases in anti-
ssDNA antibodies
seen in both groups at
32 weeks. Anti-GA
were not affected
No effect on splenocyte
NK activity
Increased renal pathology
scores in 1400 ppb group;
Significant decrease in
thymus weight in both
groups
Kiel et al.
(2009)
a Selected endpoints, based on those reported across the majority of studies. Lupus-prone mouse strains develop lupus-like condition spontaneously,
with virtually complete penetrance. The autoimmune-inducible (Brown Norway) rat has been used as a model of mercuric chloride induced
glomerulonephritis and experimental autoimmune myasthenia gravis.
b NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level) are based upon reported study findings.
c No difference reported in anti- ds-DNA, -ss-DNA, -riboneucleosome, -SSA, -SSB, -Sm, -Jo-1, or -Scl-70 antibodies.
d No difference in secretion of other cytokines measured: IL-2, IL-4, IL-10, IL-12, TNF-a, granulocyte monocyte colony stimulating factor,
macrophage inflammatory protein-la, and RANTES (CCL-5)
e Dose levels cited in the report (Gilkeson et al., 2004) were incorrect; corrections provided by personal communication from Margie Peden-Adams
[Medical University of South Carolina] to Glinda Cooper [U.S. EPA] on 13 August 2008; dose levels in this table are correctly report.
f Dose in mg/kg-day not given
8 Anti-dsDNA tests were described in the methods section; no effect of TCE on serum IgE levels was seen, and it is not clear if the additional
serological tests were conducted in the TCE portion of this study or if they were conducted but not reported because no effect was seen.
6/22/2009
682
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.5.2.4 Cancers of the immune system
Cancers of the immune system that have been observed in animal studies and are
associated with TCE exposure are summarized in Tables 4.5-9 and 4.5-10. The specific tumor
types observed are malignant lymphomas, lymphosarcomas, and reticulum cell sarcomas in mic
and leukemias in rats.
In the NCI (1976) study, the results for Osborne-Mendel rats were considered
inconclusive due to significant early mortality, but exposure to B6C3F1 mice were also
analyzed. Limited increases in lymphomas over controls were observed in both sexes of mice
exposed (Table 4.5-11). The NCI study (1976) used technical grade TCE which contained two
known carcinogenic compounds as stabilizers (epichlorohydrin and 1,2-epoxybutane), and a later
study (Henschler et al., 1984) in which mice were given TCE that was pure, industrial, and
stabilized with one or both of these stabilizers did not find significant increases in lymphomas
over historical controls. A later gavage study by NTP (1988), which used TCE stabilized with
diisopropylamine, did not see an increase in lymphomas in all four strains of rats (ACI, August,
Marshall and Osborne-Mendel). The final NTP study (1990) in male and female F344 rats and
B6C3F1 mice, using epichlorohydrin-free TCE, again experienced early mortality in male rats.
This study did not observe significant increase in lymphomas over that of controls. Henschler et
al. (1980) tested NMRI mice, WIST rats and Syrian hamsters of both sexes, and observed a
variety of tumors in both sexes (Henschler et al., 1980), consistent with the spontaneous tumor
incidence in this strain (Deerberg and Muller-Peddinghaus, 1970; Deerberg et al., 1974).
Henschler et al. did not show an increase in lymphomas in rats or hamsters of either sex
(Henschler et al., 1980). Background levels of lymphomas in this mouse strain are high, making
it difficult to determine if the increased lymphomas in female mice is a treatment effect. In a
follow-up study, Henschler et al. (1984) examined the role of stabilizers of TCE in the
lymphomas demonstrated in female mice in the 1980 paper. Each exposure group had -50 SPF-
bred ICR/HA-Swiss mice and exposure was for 18 months. Background incidence of tumors
was high in all groups. Focusing just on malignant lymphomas (Table 4.5-11), the high
background incidence in unexposed animals again makes it difficult to determine if there is TCE
and/or stabilizer-related incidence of lymphomas. There is no data at any other timepoint than
18 months. A high mortality rate in all animals as well as the increased incidence of
'background' lymphomas in that report was also a problem and may have been related to the
shorter time frame.
6/22/2009
683
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5.-11. Malignant lymphomas incidence in mice exposed to TCE in Gavage and Inhalation Exposure
Studies
Cancer Type, Species and Sex
Exposure Groups
Reference
Gavage Exposure
Malignant lymphomas
Prevalence in: (n affected/total)
B6C3F1 mice, male
B6C3F1 mice, female
NTP, 1990a
Vehicle control
11/50 (22%)
7/48 (15%)
Lymphosarcomas and reticulum cell sarcomas
Prevalence in: (n affected/total)
B6C3F1 mice, male
B6C3F1 mice, female
Malignant lymphomas
Vehicle control
1/20 (5%)
1/20 (5%)
1,000 mg/kg-day
13/50 (26%)
13/49 (27%)
Low dose
4/50 (8%)
5/50 (10%)
NCI, 1976
High dose
2/48 (4%)
5/47(11%)
Henschler et al..
1984°
TCE-
TCE-
TCE-
TCE-
TCE-
Prevalence in: (n affected/total)
Control
pure
indust
EPC
BO
EPC-BO
Swiss (ICR/HA) mice, male
19/50
16/50
17/49
11/49
11/49
12/49
(38%)
(32%)
(35%)
(22%)
(22%)
(24%)
Swiss (ICR/HA) mice, female
28/50
21/50
19/50
20/50
23/48
18/50
(56%)
(42%)
(38%)
(40%)
(48%)
(36%)
Henschler et al..
1980d
Inhalation Exposure
Malignant lymphomas Control 96 480
Prevalence in: (n affected/total)
Han:NMRI mice, male 7/30 (23%) 7/29 (24%) 6/30 (20%)
Han:NMRI mice, female6 9/29(31%) 17/30 (57%) 18/28(64%)
a after 103 weeks gavage exposure, beginning at 8 weeks of age
b after 90 weeks gavage exposure, beginning at 5 weeks of age. Low dose is 1,200 mg/kg-d for male mice, 900 mg/kg-
d for female mice (5 d/wk). High dose is 2,400 mg/kg-d for male mice, 1,800 mg/kg-d for female mice (5 d/wk).
0 after 72 weeks gavage exposure (corn oil), beginning at 5 weeks of age. Male mice received 2,400 mg/kg-d, female
mice received 1,800 mg/kg-d. Stabilizers were added in the percent w/w: TCE-EPC, 0.8%, TCE-BO, 0.8%, TCE-
EPC-BO, 0.25% and 0.25%.
dafter 78 weeks inhalation exposure. Administered daily concentration: low dose is 96 (mg/m3) and high dose is 480
(mg/m3), equivalent to 100 and 500 ppm (100 ppm = 540 mg/m3), adjusted for 6 hr/d, 5 d/wk exposure.
e Statistically significant by Cochran-Armitage trend test (p < 0.05).
Sources: NTP (1990) Tables 8, 9; NCI (1976) Table XXXa; Henschler et al. (1980) Table 3a.
Maltoni et al reported a non-significant increase leukemias in male rats exposed in
inhalation (Matoni et al., 1988, 1986). Maltoni et al. (1986) demonstrates a borderline higher
frequency of leukemias in male Sprague Dawley rats following exposure by ingestion for 52
6/22/2009
684
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
weeks, believed by the authors to be related to an increase in lymphoblastic lymphosarcomas
(Table 4.5-12). The gavage study by NTP (1988), which used TCE stabilized with
diisopropylamine, observed leukemia in female August rats with a positive trend, but was not
significantly greater than the vehicle controls.
6/22/2009
685
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.5.-12. Leukemia incidence in rats exposed to TCE in Gavage and Inhalation Exposure Studies
Species and Sex
Exposure Groups
Reference
Gavage Exposure
Prevalence in: (n affected/total)
Control
50mg/kg
250mg/kg
Maltoni et al., 198611
Sprague-Dawley rats, male
0/30
2/30
3/30
(0 %)
(6.7%)
(10.0%)
Sprague-Dawley rats, female
1/ 30
0/30
0/30
(3.3 %)
(0%)
(0%)
Control
500mg/kg
l,000mg/kg
NTP, 1988 b
August rats, female
0/50
1/50
5/50
(0%)
(2%)
(10%)
Inhalation Exposure
Prevalence in: (n affected/total)
Control
100 ppm
300 ppm
600 ppm
Maltoni et al., 19880
Sprague-Dawley rats, male
9/135
13/130
14/130
15/130
(6.7)
(10.0)
(10.8)
(11.5)
Sprague-Dawley rats, female
7/145
9/130
2/130
11/130
(4.8)
(6.9)
(1.5)
(8.5)
1 after 52 weeks gavage exposure, beginning at 13 weeks of age, olive oil vehicle. Percent affected and starting n
given in reported; EPA calculated n affected.
3 after 104 weeks gavage exposure, beginning at 6.5-8 weeks of age, corn oil vehicle.
3 after 104 weeks inhalation exposure, BT304 and BT304bis. Percent affected and starting n given in reported; EPA
calculated n affected.
In summary, overall there is limited available data on the role of TCE in lymphomas and
leukemias. There are few studies that analyze for lymphomas and/or leukemias. Lymphomas
were described in four studies (NTP, 1990; NCI, 1976; Henschler et al., 1980, 1984) but study
limitations (high background rate) in most studies make it difficult to determine if these are
TCE-induced. Three studies have found positive trends in leukemia in specific strains and/or
gender (Maltoni et al., 1986, 1988; NTP, 1988) but also due to study limitations can not be
determined to be TCE-induced.
6/22/2009 686
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.5.3 Summary
4.5.3.1 Noncancer Effects
The human and animal studies of TCE and immune-related effects provide strong
evidence for a role of TCE in autoimmune disease and in a specific type of generalized
hypersensitivity syndrome. The data pertaining to immunosuppressive effects is weaker.
The relation between systemic autoimmune diseases, such as scleroderma, and
occupational exposure to TCE has been reported in several recent studies. A meta-analysis of
scleroderma studies (Diot et al., 2002; Garabrant et al., 2003; Nietert et al., 1998) conducted by
the EPA resulted in a statistically significant combined odds ratio for any exposure in men (OR =
2.5, 95% CI 1.1, 5.4), with a lower relative risk seen in women in women (OR = 1.2, 95% CI
0.58, 2.6). The incidence of systemic sclerosis among men is very low (approximately 1 per
100,000 per year), and is approximately 10 times lower than the rate seen in women (Cooper and
Stroehla, 2003). Thus the human data at this time do not allow us to determine if the difference
in effect estimates between men and women reflects the relatively low background risk of
scleroderma in men, gender-related differences in exposure prevalence or in the reliability of
exposure assessment (Messing et al., 2003), a gender-related difference in susceptibility to the
effects of TCE, or chance. Changes in levels of inflammatory cytokines were reported in an
occupational study of degreasers exposed to TCE (Iavicoli et al., 2005) and a study of infants
exposed to TCE via indoor air (Lehmann et al., 2001, 2002). Experimental studies support the
biological plausibility of these effects. Numerous studies have demonstrated accelerated
autoimmune responses in autoimmune-prone mice (Cai et al., 2008; Blossom et al., 2007, 2004;
Griffin et al., 2000a, b). With shorter exposure periods, effects include changes in cytokine
levels similar to those reported in human studies. More severe effects, including autoimmune
hepatitis, inflammatory skin lesions, and alopecia, were manifest at longer exposure periods, and
interestingly, these effects differ somewhat from the "normal" expression in these mice.
Immunotoxic effects, including increases in anti-ds DNA antibodies in adult animals and
decreased plaque forming cell response with prenatal and neonatal exposure, have been also
reported in B6C3F1 mice, which do not have a known particular susceptibility to autoimmune
disease (Gilkeson et al., 2004, Peden-Adams et al., 2006). Recent mechanistic studies have
focused on the roles of various measures of oxidative stress in the induction of these effects by
TCE (Wang et al., 2008, 2007b).
There have been a large number of case reports of a severe hypersensitivity skin disorder,
distinct from contact dermatitis and often accompanied by hepatitis, associated with occupational
exposure to TCE, with prevalences as high as 13% of workers in the same location (Kamijima et
al., 2008, 2007). Evidence of a treatment-related increase in delayed hypersensitivity response
6/22/2009
687
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
accompanied by hepatic damage has been observed in guinea pigs following intradermal
injection (Tang et al., 2008, 2006), and hypersensitivity response was also seen in mice exposed
via drinking water pre- and post-natally (gestation day 0 through to 8 weeks of age) (Peden-
Adams et al., 2006).
Human data pertaining to TCE-related immunosuppression resulting in an increased risk
of infectious diseases is limited to the report of an association between reported history of
bacteria of viral infections in Woburn, Massachusetts (Lagakos, 1986). Evidence of localized
immunosuppression, as measured by pulmonary response to bacterial challenge (i.e., risk of
Streptococcal pneumonia-related mortality and clearance of Klebsiella bacteria) was seen in an
acute exposure study in CD-I mice (Aranyi et al., 1986). A 4-week inhalation exposure in
Sprague-Dawley rats reported a decrease in plaque forming cell response at exposures of 1,000
ppm (Woolhiser et al., 2006).
4.5.3.2 Cancer
The available epidemiologic studies provide limited evidence for a causal relation
between trichloroethylene exposure and non-Hodgkin lymphoma. Issues of study heterogeneity,
potential publication bias, and weaker exposure-response results contribute uncertainty to the
evaluation of the available data.
In a systematic review of the non-Hodgkin lymphoma studies, 17 studies in which there
is a high likelihood of TCE exposure in individual study subjects (e.g., based on job-exposure
matrices or biomarker monitoring) and which met, to a sufficient degree, the standards of
epidemiologic design and analysis were identified. These studies generally reported excess
relative risk estimates for non-Hodgkin lymphoma between 0.8 and 3.1 for overall TCE
exposure. Statistically significant elevated relative risk estimates were observed in two cohort
(Hansen et al., 2001; Raaschou-Nielsen et al., 2003) and one case-control (Hardell et al., 1994)
study. The other high-quality studies reported elevated relative risk estimates with overall TCE
exposure that were not statistically significant, except for two population case-control studies of
non-Hodgkin lymphoma, which did not reported relative risk estimates with overall TCE
exposure (Miligi et al., 2006; Seidler et al., 2007). Fifteen additional studies were given less
weight because of their lesser likelihood of TCE exposure and other design limitations that
would decrease study power and sensitivity. The observed lack of association with lymphoma in
these studies likely reflects study design and exposure assessment limitations and is not
considered inconsistent with the overall evidence on TCE and lymphoma.
Consistency of the association between TCE exposure and lymphoma is further
supported by the results of meta-analyses of 15 high-quality studies reporting risk estimates for
overall TCE exposure. These meta-analyses found a statistically significant increased pooled
6/22/2009
688
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
relative risk estimate for lymphoma of 1.27 (95% CI: 1.04, 1.53) for overall TCE exposure. The
analysis of non-Hodgkin lymphoma was generally robust to the removal of individual studies
and the use of alternate relative risk estimates from individual studies, though in a few cases, the
resulting pooled relative risk was no longer statistically significant (lower 95% confidence
bounds reduced to 0.99-1.00). However, some evidence heterogeneity was observed
(p = 0.048), particularly between cohort and case-control studies; and, in addition, there was
some evidence of potential publication bias. Analyzing the cohort and case-control studies
separately resolved most of the heterogeneity, but the result for the pooled case-control studies
was only a 5% increased relative risk estimate and was not statistically significant. The sources
of heterogeneity are uncertain but may be the result of some bias associated with exposure
assessment and/or disease classification, or from differences between cohort and case-control
studies in average TCE exposure.
Exposure-response relationships are examined in the TCE epidemiologic studies only to a
limited extent. Many studies examined only overall "exposed" versus "unexposed" groups and
did not provide exposure information by level of exposure. Others do not have adequate
exposure assessments to confidently distinguish between levels of exposure. The non-Hodgkin
lymphoma case-control study of Seidler et al. (2007) reported a statistically significant trend with
TCE exposure [p = 0.03 for Diffuse B-cell lymphoma trend with cumulative TCE exposure], and
NHL risk in Boice et al. (1999) appeared to increase with increasing exposure duration [p = 0.20
for routine-intermittent exposed subjects]. The borderline statistically significant trend with TCE
intensity in the case-control study of Wang et al. (2009) [p = 0.06] is consistent with Seidler et
al. (2007). Further support was provided by meta-analyses using only the highest exposure
groups, which yielded a higher pooled relative risk estimate [1.50 (95% CI: 1.20, 1.88)] than for
overall TCE exposure.
Few risk factors are recognized for non-Hodgkin lymphoma, with the exception of
viruses, immunosuppression or smoking, which are associated with specific lymphoma subtypes.
Associations between non-Hodgkin lymphoma and TCE exposure are based on groupings of
several NHL subtypes. Three of the six non-Hodgkin lymphoma case-control studies adjusted
for age, sex and smoking in statistical analyses (Miligi et al., 2006; Seidler et al., 2007; Wang et
al., 2009), the other three case-control studies presented only unadjusted estimates of the odds
ratio.
Animal studies describing rates of lymphomas and/or leukemias in relation to TCE
exposure (NTP, 1990, 1988; NCI, 1976; Henschler et al., 1980, 1984; Maltoni et al., 1986, 1988)
are available. Henschler et al. (1980) reported statistically significant increases in lymphomas in
female Han:NMRI mice treated via inhalation. While Henschler et al. (1980) suggested these
lymphomas were of viral origin specific to this strain, subsequent studies reported increased
6/22/2009
689
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
lymphomas in female B6C3F1 mice treated via corn oil gavage (NTP, 1990) and leukemias in
male Sprague-Dawley and female August rats (Maltoni et al., 1986; NTP, 1988). However,
these tumors had relatively modest increases in incidence with treatment, and were not reported
to be increased in other studies.
6/22/2009
690
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
References
Adams, C; Keil, D; Meyers, K; EuDaly, A; Smythe, J; EuDaly, J; Gilkeson, G; Peden-Adams,
MM. (2003) Lifetime exposure to trichloroethylene (TCE) modulates immune function.
Toxicologist 72(S-1):375.
ADHS (Arizona Department of Health Services). 1990. The incidence of childhood leukemia
and testicular cancer in Pima County: 1970-1986. Prepared for the Arizona Department
of Environmental Quality under the Water Quality Assurance Revolving Fund and by the
Arizona Department of Health Services, Division of Disease Prevention, Office of Risk
Assessment and Investigation, Office of Chronic Disease Epidemiology. September 17,
1990.
ADHS (Arizona Department of Health Services). 1995. Update of the incidence of childhood
cancers and testicular cancer in Southwest Tucson, 1987 - 1991. Prepared by the
Arizona Department of Health Services, Office of Risk Assessment and Investigations,
Disease Prevention Services Arizona Department of Health Services. June, 6, 1995.
Aickin M, Chapin CA, Flood TJ, Englender SJ, Caldwell GG. 1992. Assessment of the spatial
occurrence of childhood leukaemia mortality using standardized rate ratios with a simple
linear Poisson model. Int J Epidemiol 21:649-655.
Albert DA, Albert AN, Vernace M, Sebastian JK, Hsia EC. 2005. Analysis of a cluster of cases
of Wegener granulomatosis. J Clin Rheumatol 11:188-193.
Alexander DD, Mink PJ, Mandel JH, Kelsh MA. 2006. A meta-analysis of occupational
trichloroethylene exposure and multiple myeloma or leukaemia. Occup Med (Lond)
56:485-493.
Alexander DD, Mink PJ, Adami HO, Chang ET, Cole P, Mandel JS, et al. 2007. The non-
Hodgkin lymphomas: a review of the epidemiologic literature. Int J Cancer 120 Suppl
12:1-39.
Anttila A, Pukkala E, Sallmen M, Hernberg S, Hemminki K. 1995. Cancer incidence among
Finnish workers exposed to halogenated hydrocarbons. J Occup Environ Med 37:797-
806.
Aranyi C, O'Shea WJ, Graham JA, Miller FJ. 1986. The effects of inhalation of organic chemical
air contaminants on murine lung host defenses. Fundam Appl Toxicol 6:713-720.
Arif AA, Shah SM. 2007. Association between personal exposure to volatile organic compounds
and asthma among US adult population. Int Arch Occup Environ Health 80:711-719.
6/22/2009
691
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Aryal BK, Khuder SA, Schaub EA. 2001. Meta-analysis of systemic sclerosis and exposure to
solvents. Am J Ind Med 40:271-274.
ATSDR (Agency for Toxic Substance and Disease Registry). 2004. Feasibility Investigation of
Worker Exposure to Trichloroethylene at the View-Master Factory in Beaverton, Oregon.
Prepared by the Oregon Department of Human Services. Report No. PB2005-102173.
ATSDR (Agency for Toxic Substances and Disease Registry). 2006. Health Consultation.
Cancer and Birth Outcome Analysis, Endicott Area, Town of Union, Broome County,
New York. Health Statistics Review. Atlanta, GA: U.S. Department of Health and
Human Services, Public Health Service, Agency for Toxic Substances and Disease
Registry. May 26, 2006.
Axelson O, Selden A, Andersson K, Hogstedt C. 1994. Updated and expanded Swedish cohort
study on trichloroethylene and cancer risk. J Occup Med 36:556-562.
Bakke B, Stewart PA, Waters MA. 2007. Uses of and exposure to trichloroethylene in U.S.
industry: a systematic literature review. J Occup Environ Hyg 4:375-390.
Balluz L, Philen R, Ortega L, Rosales C, Brock J, Barr D, et al. 2001. Investigation of systemic
lupus erythematosus in Nogales, Arizona. Am J Epidemiol 154:1029-1036.
Beaudreuil S, Lasfargues G, Laueriere L, El Ghoul Z, Fourquet F, Longuet C, et al. 2005.
Occupational exposure in ANCA-positive patients: a case-control study. Kidney Int
67:1961-1966.
Blair A, Linos A, Stewart PA, Burmeister LF, Gibson R, Everett G, et al. 1993. Evaluation of
risks for non-Hodgkin's lymphoma by occupation and industry exposures from a case-
control study. Am J Ind Med 23:301-312.
Blair A, Hartge P, Stewart PA, McAdams M, Lubin J. 1998. Mortality and cancer incidence of
aircraft maintenance workers exposed to trichloroethylene and other organic solvents and
chemicals: extended follow up. Occup Environ Med 55:161-171.
Blossom SJ, Doss JC. (2007) Trichloroethylene alters central and peripheral immune function in
autoimmune-prone MRL+/+ mice following continuous developmental and early life
exposure. Journal of Immunotoxicology 4:129-141.
Blossom SJ, Gilbert KM. 2006. Exposure to a metabolite of the environmental toxicant,
trichloroethylene, attenuates CD4+ T cell activation-induced cell death by
metalloproteinase-dependentFasL shedding. Toxicol Sci 92:103-114.
6/22/2009
692
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Blossom SJ, Pumford NR, Gilbert KM. 2004. Activation and attenuation of apoptosis of CD4+ T
cells following in vivo exposure to two common environmental toxicants,
trichloroacetaldehyde hydrate and trichloroacetic acid. J Autoimmun 23:211-220.
Blossom SJ, Doss JC, Gilbert KM. 2007. Chronic exposure to a trichloroethylene metabolite in
autoimmune-prone MRL+/+ mice promotes immune modulation and alopecia. Toxicol
Sci 95:401-411.
Blossom SJ, Doss JC, Hennings LJ, Jernigan S, Melnyk S, James SJ. 2008. Developmental
exposure to trichloroethylene promotes CD4+ T cell differentiation and hyperactivity in
association with oxidative stress and neurobehavioral deficits in MRL+/+ mice. Toxicol
Appl Pharmacol 231:344-353.
Boffetta P, de Vocht F. 2007. Occupation and the risk of non-Hodgkin lymphoma. Cancer
Epidemiol Biomarkers Prev 16:369-372.
Boice JD, Jr., Marano DE, Fryzek JP, Sadler CJ, McLaughlin JK. 1999. Mortality among aircraft
manufacturing workers. Occup Environ Med 56:581-597.
Boice JD, Jr., Marano DE, Cohen SS, Mumma MT, Blot WJ, Brill AB, et al. 2006. Mortality
among Rocketdyne workers who tested rocket engines, 1948-1999. J Occup Environ Med
48:1070-1092.
Bovenzi M, Barbone F, Pisa FE, Betta A, Romeo L, Tonello A, et al. 2004. A case-control study
of occupational exposures and systemic sclerosis. Int Arch Occup Environ Health 77:10-
16.
Byers VS, Levin AS, Ozonoff DM, Baldwin RW. 1988. Association between clinical symptoms
and lymphocyte abnormalities in a population with chronic domestic exposure to
industrial solvent-contaminated domestic water supply and a high incidence of
leukaemia. Cancer Immunol Immunother 27:77-81.
Cai P, Konig R, Khan MF, Qiu S, Kaphalia BS, Ansari GA. 2006. Autoimmune response in
MRL+/+ mice following treatment with dichloroacetyl chloride or dichloroacetic
anhydride. Toxicol Appl Pharmacol 216:248-255.
Cai P, Konig R, Khan MF, Kaphalia BS, Ansari GA. 2007. Differential immune responses to
albumin adducts of reactive intermediates of trichloroethene in MRL+/+ mice. Toxicol
Appl Pharmacol 220:278-283.
Cai P, Konig R, Boor PJ, Kondraganti S, Kaphalia BS, Khan MF, et al. 2008. Chronic exposure
to trichloroethene causes early onset of SLE-like disease in female MRL +/+ mice.
Toxicol Appl Pharmacol 228:68-75.
6/22/2009
693
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Chang YM, Tai CF, Yang SC, Chen CJ, Shih TS, Lin RS, et al. 2003. A cohort mortality study
of workers exposed to chlorinated organic solvents in Taiwan. Ann Epidemiol 13:652-
660.
Chang YM, Tai CF, Yang SC, Lin RS, Sung FC, Shih TS, et al. 2005. Cancer incidence among
workers potentially exposed to chlorinated solvents in an electronics factory. J Occup
Health 47:171-180.
Chiu BC, Weisenburger DD. 2003. An update of the epidemiology of non-Hodgkin's lymphoma.
Clin Lymphoma 4:161-168.
Cohn M, Klotz J, Bove F, Berkowitz M, Fagliano J. 1994. Drinking water contamination and
the incidence of leukemia and non-Hodgkin's lymphoma. Environ Health Perspect
102:556-561.
Colt JS, Blair A. 1998. Parental occupational exposures and risk of childhood cancer. Environ
Health Perspect 106 Suppl 3:909-925.
Cooper GS, Miller FW, Pandey JP. 1999. The role of genetic factors in autoimmune disease:
implications for environmental research. Environ Health Perspect 107 Suppl 5:693-700.
Cooper GS, Parks CG, Treadwell EL, St Clair EW, Gilkeson GS, Dooley MA. 2004.
Occupational risk factors for the development of systemic lupus erythematosus. J
Rheumatol 31:1928-1933.
Cooper GS, Stroehla BC. 2003. The epidemiology of autoimmune diseases. Autoimmun Rev
2:119-125.
Costa G, Merletti F, Segnan N. 1989. A mortality cohort study in a north Italian aircraft factory.
Br JInd Med 46:738-743.
Costantini AS, Benvenuti A, Vineis P, Kriebel D, Tumino R, Ramazzotti V, et al. 2008. Risk of
leukemia and multiple myeloma associated with exposure to benzene and other organic
solvents: Evidence from the Italian Multicenter Case-control study. Am J Ind Med
51:803-811.
Costas K, Knorr RS, Condon SK. 2002. A case-control study of childhood leukemia in Woburn,
Massachusetts: the relationship between leukemia incidence and exposure to public
drinking water. Sci Total Environ 300:23-35.
Dahlgren J, Takhar H, Anderson-Mahoney P, Kotlerman J, Tarr J, Warshaw R. 2007. Cluster of
systemic lupus erythematosus (SLE) associated with an oil field waste site: a cross
sectional study. Environ Health 6:8.
6/22/2009
694
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Dai Y, Leng S, Li L, Niu Y, Huang H, Cheng J, et al. 2004. Genetic polymorphisms of cytokine
genes and risk for trichloroethylene-induced severe generalized dermatitis: a case-control
study. Biomarkers 9:470-478.
Deerberg F, Muller-Peddinghaus R. 1970. [Comparative studies on the occurrence of
spontaneous diseases and spontaneous deaths in conventional and specific germ-free
female NMRI-HAN mice], Z Versuchstierkd 12:341-347.
Deerberg F, Pittermann W, Rapp K. 1974. [Lung tumor in the mouse: a progressive tumor]. Vet
Pathol 11:430-441.
DerSimonian R, Laird N. 1986. Meta-analysis in clinical trials. Control Clin Trials 7:177-188.
Diot E, Lesire V, Guilmot JL, Metzger MD, Pilore R, Rogier S, et al. 2002. Systemic sclerosis
and occupational risk factors: a case-control study. Occup Environ Med 59:545-549.
Dryver E, Brandt L, Kauppinen T, Olsson H. 2004. Occupational exposures and non-Hodgkin's
lymphoma in Southern Sweden. Int J Occup Environ Health 10:13-21.
Environmental Health Strategies. (1997). Final Report. Correspondence from Paul A. Cammer,
Ph.D., Trichloroethylene Issues Group, to Cheryl Siegel Scott, US Environmental
Protection Agency. December 22, 1997.
Figgs LW, Dosemeci M, Blair A. 1995. United States non-Hodgkin's lymphoma surveillance by
occupation 1984-1989: a twenty-four state death certificate study. Am J Ind Med 27:817-
835.
Finckh A, Cooper GS, Chibnik LB, Costenbader KH, Watts J, Pankey H, et al. 2006.
Occupational silica and solvent exposures and risk of systemic lupus erythematosus in
urban women. Arthritis Rheum 54:3648-3654.
Gama C, Meira JB. 1978. Occupational acro-osteolysis. J Bone Joint Surg Am 60:86-90.
Garabrant DH, Held J, Langholz B, Bernstein L. 1988. Mortality of aircraft manufacturing
workers in southern California. Am J Ind Med 13:683-693.
Garabrant DH, Lacey JV, Jr., Laing TJ, Gillespie BW, Mayes MD, Cooper BC, et al. 2003.
Scleroderma and solvent exposure among women. Am J Epidemiol 157:493-500.
Gilbert KM, Griffin JM, PumfordNR. 1999. Trichloroethylene activates CD4+ T cells: potential
role in an autoimmune response. Drug Metab Rev 31:901-916.
Gilbert KM, Whitlow AB, Pumford NR. 2004. Environmental contaminant and disinfection by-
product trichloroacetaldehyde stimulates T cells in vitro. Int Immunopharmacol 4:25-36.
6/22/2009
695
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Gilbert KM, Pumford NR, Blossom SJ. 2006. Environmental Contaminant Trichloroethylene
Promotes Autoimmune Disease and Inhibits T-cell Apoptosis in MRL(+/+) Mice. J
Immunotoxicol. 3:263-267.
Gilkeson GS, Keil D, Peden-Adams MM. 2004. Immune effects of trichloroethylene on
autoimmune disease in mice. In: Trichloroethylene - The Scientific Basis of Risk
Assessment (Mohr LC, Hoel DG, Jollow D, eds). Charleston, S.C.: The Medical
University of South Carolina Press, 87-98.
Greenland S, Salvan A, Wegman DH, HallockMF, Smith TJ. 1994. A case-control study of
cancer mortality at a transformer-assembly facility. Int Arch Occup Environ Health
66:49-54.
Griffin JM, Blossom SJ, Jackson SK, Gilbert KM, Pumford NR. 2000a. Trichloroethylene
accelerates an autoimmune response by Thl T cell activation in MRL +/+ mice.
Immunopharmacology 46:123-137.
Griffin JM, Gilbert KM, Lamps LW, Pumford NR. 2000b. CD4(+) T-cell activation and
induction of autoimmune hepatitis following trichloroethylene treatment in MRL+/+
mice. Toxicol Sci 57:345-352.
Griffin JM, Gilbert KM, Pumford NR. 2000c. Inhibition of CYP2E1 reverses CD4+ T-cell
alterations in trichloroethylene-treated MRL+/+mice. Toxicol Sci 54:384-389.
Halmes NC, Perkins EJ, McMillan DC, Pumford NR. 1997. Detection of trichloroethylene-
protein adducts in rat liver and plasma. Toxicol Lett 92:187-194.
Hansen J, Raaschou-Nielsen O, Christensen JM, Johansen I, McLaughlin JK, Lipworth L, et al.
2001. Cancer incidence among Danish workers exposed to trichloroethylene. J Occup
Environ Med 43:133-139.
Hardell L, Eriksson M, Degerman A. 1994. Exposure to phenoxyacetic acids, chlorophenols, or
organic solvents in relation to histopathology, stage, and anatomical localization of non-
Hodgkin's lymphoma. Cancer Res 54:2386-2389.
Harris NL, Jaffe ES, Diebold J, Flandrin G, Muller-Hermelink HK, Vardiman J. 2000.
Lymphoma classification—from controversy to consensus: the R.E.A.L. and WHO
Classification of lymphoid neoplasms. Ann Oncol 11 Suppl 1:3-10.
Henschler D, Romen W, Elsasser HM, Reichert D, Eder E, Radwan Z. 1980. Carcinogenicity
study of trichloroethylene by longterm inhalation in three animal species. Arch Toxicol
43:237-248.
6/22/2009
696
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Henschler D, Elsasser H, Romen W, Eder E. 1984. Carcinogenicity study of trichloroethylene,
with and without epoxide stabilizers, in mice. J Cancer Res Clin Oncol 107:149-156.
Henschler D, Vamvakas S, Lammert M, Dekant W, Kraus B, Thomas B, et al. 1995. Increased
incidence of renal cell tumors in a cohort of cardboard workers exposed to
trichloroethene. Arch Toxicol 69:291-299.
Hobara T, Kobayashi H, Higashihara E, Kawamoto T, Sakai T. 1984. Acute effects of 1,1,1-
trichloroethane, trichloroethylene, and toluene on the hematologic parameters in dogs.
Arch Environ Contam Toxicol 13:589-593.
Huang HL, Li LY, Chen B, Huang JX, Huang SR. 2002. New problems caused by occupational
trichloroethylene exposure. Int J Immunopath Pharm 15:30-32.
Iavicoli I, Marinaccio A, Carelli G. 2005. Effects of occupational trichloroethylene exposure on
cytokine levels in workers. J Occup Environ Med 47:453-457.
Kamijima M, Hisanaga N, Wang H, Nakajima T. 2007. Occupational trichloroethylene exposure
as a cause of idiosyncratic generalized skin disorders and accompanying hepatitis similar
to drug hypersensitivities. Int Arch Occup Environ Health 80:357-370.
Kamijima M, Wang H, Huang H, Li L, Shibata E, Lin B, et al. 2008. Trichloroethylene Causes
Generalized Hypersensitivity Skin Disorders Complicated by Hepatitis. J Occup Health.
Kaneko T, Saegusa M, Tasaka K, Sato A. 2000. Immunotoxicity of trichloroethylene: a study
with MRL-lpr/lpr mice. J Appl Toxicol 20:471-475.
Karunanayake CP, McDuffie HH, Dosman JA, Spinelli JJ, Pahwa P. 2008. Occupational
exposures and non-Hodgkin's lymphoma: Canadian case-control study. Environ Health
7:44
Kauffmann BM, White KL, Jr., Sanders VM, Douglas KA, Sain LE, Borzelleca JF, et al. 1982.
Humoral and cell-mediated immune status in mice exposed to chloral hydrate. Environ
Health Perspect 44:147-151.
Keil DE, Peden-Adams MM, Wallace S, Ruiz P, Gilkeson GS. (2009) Assessment of
trichloroethylene (TCE) exposure in murine strains genetically-prone and non-prone to
develop autoimmune disease. J Environ Sci Health Part A 44:443-453.
Kelsh M, Weingart M, Mandel J, Mink P, Alexander D, Basu R, Kalmes R, Goodman M. 2005.
A meta-analysis of epidemiology studies of occupational TCE exposure and selected
cancers. Presented to National Research Council, Committee on Human Health Risks of
Trichloroethylene. June 9, 2005.
6/22/2009
697
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Khan MF, KaphaliaBS, Prabhakar BS, Kanz MF, Ansari GA. 1995. Trichloroethene-induced
autoimmune response in female MRL +/+ mice. Toxicol Appl Pharmacol 134:155-160.
Khan MF, Wu X, Ansari GA. 2001. Anti-malondialdehyde antibodies in MRL+/+ mice treated
with trichloroethene and dichloroacetyl chloride: possible role of lipid peroxidation in
autoimmunity. Toxicol Appl Pharmacol 170:88-92.
Kilburn KH, Warshaw RH. 1992. Prevalence of symptoms of systemic lupus erythematosus
(SLE) and of fluorescent antinuclear antibodies associated with chronic exposure to
trichloroethylene and other chemicals in well water. Environ Res 57:1-9.
Kilburn KH, Warshaw RH. 1993. Neurobehavioral testing of subjects exposed residentially to
groundwater contaminated from an aluminum die-casting plant and local referents. J
Toxicol Environ Health 39:483-496.
Lacey JV, Jr., Garabrant DH, Laing TJ, Gillespie BW, Mayes MD, Cooper BC, et al. 1999.
Petroleum distillate solvents as risk factors for undifferentiated connective tissue disease
(UCTD). Am J Epidemiol 149:761-770.
Lagakos SW, Wessen BJ, Zelen M. 1986. Analysis of contaminated well water and health effect
in Woburn, Massachusetts. Journal of the American Statistical Association 81:583-596.
Lane SE, Watts RA, Bentham G, Innes NJ, Scott DG. 2003. Are environmental factors important
in primary systemic vasculitis? A case-control study. Arthritis Rheum 48:814-823.
Lehmann I, Rehwagen M, Diez U, Seiffart A, Rolle-Kampczyk U, Richter M, et al. 2001.
Enhanced in vivo IgE production and T cell polarization toward the type 2 phenotype in
association with indoor exposure to VOC: results of the LARS study. Int J Hyg Environ
Health 204:211-221.
Lehmann I, Thoelke A, Rehwagen M, Rolle-Kampczyk U, Schlink U, Schulz R, et al. 2002. The
influence of maternal exposure to volatile organic compounds on the cytokine secretion
profile of neonatal T cells. Environ Toxicol 17:203-210.
Lemke LE, McQueen CA. 1995. Acetylation and its role in the mutagenicity of the
antihypertensive agent hydralazine. Drug Metab Dispos 23:559-565.
Lockey JE, Kelly CR, Cannon GW, Colby TV, Aldrich V, Livingston GK. 1987. Progressive
systemic sclerosis associated with exposure to trichloroethylene. J Occup Med 29:493-
496.
6/22/2009
698
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Lowengart RA, Peters JM, Cicioni C, Buckley J, Bernstein L, Preston-Martin S, et al. 1987.
Childhood leukemia and parents' occupational and home exposures. J Natl Cancer Inst
79:39-46.
Lundberg I, Alfredsson L, Plato N, Sverdrup B, Klareskog L, Kleinau S. 1994. Occupation,
occupational exposure to chemicals and rheumatological disease. A register based cohort
study. Scand J Rheumatol 23:305-310.
Lynge E, Anttila A, Hemminki K. 1997. Organic solvents and cancer. Cancer Causes Control
8:406-419.
Magnusson B, Kligman AM. 1969. The identification of contact allergens by animal assay. The
guinea pig maximization test. J Invest Dermatol 52:268-276.
Mahr AD, Neogi T, Merkel PA. 2006. Epidemiology of Wegener's granulomatosis: Lessons
from descriptive studies and analyses of genetic and environmental risk determinants.
Clin Exp Rheumatol 24:S82-91.
Maitre A, Hours M, Bonneterre V, Arnaud J, Arslan MT, Carpentier P, et al. 2004. Systemic
sclerosis and occupational risk factors: role of solvents and cleaning products. J
Rheumatol 31:2395-2401.
Maltoni C, Lefermine G, Cotti G. 1986. Experimental research on trichloroethylene
carcinogenesis. In: Archives of Research on Industrial Carcinogenesis, Vol 5 (Maltoni C,
Mehlman MA, eds). Princeton, NJ:Princeton Scientific Publishing, 1-393.
Maltoni C, Lefemine G, Cotti G, Perino G. 1988. Long-term carcinogenicity bioassays on
trichloroethylene administered by inhalation to Sprague-Dawley rats and Swiss and
B6C3F1 mice. AnnN Y Acad Sci 534:316-342.
Mandel JH, Kelsh MA, Mink PJ, Alexander DD, Kalmes RM, Weingart M, et al. 2006.
Occupational trichloroethylene exposure and non-Hodgkin's lymphoma: a meta-analysis
and review. Occup Environ Med 63:597-607.
Mannetje A, Dryson E, Walls C, McLean D, McKenzie F, Maule M, et al. 2008. High risk
occupations for non-Hodgkin's lymphoma in New Zealand: Case-control study. Occup
Environ Med 65:354-363.
McKinney PA, Alexander FE, Cartwright RA, Parker L. 1991. Parental occupations of children
with leukaemia in west Cumbria, north Humberside, and Gateshead. BMJ 302:681-687.
6/22/2009
699
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Messing K, Punnett L, Bond M, Alexanderson K, Pyle J, Zahm S, et al. 2003. Be the fairest of
them all: challenges and recommendations for the treatment of gender in occupational
health research. Am J Ind Med 43:618-629.
Mester B, Meters A, Deeg E, Eisner G, Becker N, Seidler A. 2006. Occupation and malignant
lymphoma: a population based case control study in Germany. Occup Environ Med
63:17-26.
Miligi L, Costantini AS, Benvenuti A, Kriebel D, Bolejack V, Tumino R, et al. 2006.
Occupational exposure to solvents and the risk of lymphomas. Epidemiology 17:552-561.
Morgan RW, Kelsh MA, Zhao K, Heringer S. 1998. Mortality of aerospace workers exposed to
trichloroethylene. Epidemiology 9:424-431.
Morgan JW, Cassady RE. 2002. Community cancer assessment in response to long-time
exposure to perchlorate and trichloroethylene in drinking water. J Occup Environ Med
44:616-621.
Morton LM, Wang SS, Devesa SS, Hartge P, Weisenburger DD, Linet MS. 2006. Lymphoma
incidence patterns by WHO subtype in the United States, 1992-2001. Blood 107:265-276.
Morton LM, Turner JJ, Cerhan JR, Linet MS, Treseler PA, Clarke CA, et al. 2007. Proposed
classification of lymphoid neoplasms for epidemiologic research from the Pathology
Working Group of the International Lymphoma Epidemiology Consortium (InterLymph).
Blood 110:695-708.
Nakajima T, Yamanoshita O, Kamijima M, Kishi R, Ichihara G. 2003. Generalized skin
reactions in relation to trichloroethylene exposure: a review from the viewpoint of drug-
metabolizing enzymes. J Occup Health 45:8-14.
NCI (National Cancer Institute). (1976). Carcinogenesis Bioassay of Trichloroethylene. National
Cancer Institute Carcinogenesis Technical Report Series No. 2. HEW Publication No.
(NM) 76-802.
NIAMS (National Institute of Arthritis and Musculoskeletal Diseases). (2007). The Future
Directions of Lupus Research. Available at http://www.niams.nih.gov/About_Us/
Mi s si onandPurp ose/lupus_plan.pdf
Nietert PJ, Sutherland SE, Silver RM, Pandey JP, Knapp RG, Hoel DG, et al. 1998. Is
occupational organic solvent exposure a risk factor for scleroderma? Arthritis Rheum
41:1111-1118.
6/22/2009 700
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Nordstrom M, Hardell L, Magnuson A, Hagberg H, Rask-Andersen A. 1998. Occupational
exposures, animal exposure and smoking as risk factors for hairy cell leukaemia
evaluated in a case-control study. Br J Cancer 77:2048-2052.
NRC (National Research Council). 2006. Assessing the Human Health Risks of
Trichloroethylene. Key Scientific Issues. National Academy Press, Washington, DC.
NTP. (National Toxicology Program) 1988. Technical Report on the Toxicology and
Carcinogenesis Studies of Trichloroethylene in Four Strains of Rats. NTP TR 273.
Research Triangle Park, NC: National Toxicology Program.
NTP (National Toxicology Program) 1990. Carcinogenesis Studies of Trichloroethylene
(Without Epichlorohydrin) (CAS No. 79-01-6) in F344/N Rats and B6C3F1 Mice
(Gavage Study). NTP TR 243. Research Triangle Park, NC: U.S Department of Health
and Human Services.
Olsson AR, Skogh T, Wingren G. 2000. Occupational determinants for rheumatoid arthritis.
Scand J Work Environ Health 26:243-249.
Olsson AR, Skogh T, Axelson O, Wingren G. 2004. Occupations and exposures in the work
environment as determinants for rheumatoid arthritis. Occup Environ Med 61:233-238.
Park, PA; Gilmour, MI; Selgrade, MK. 1993. Pulmonary defenses to streptococcal infection
following acute exposure to trichloroethylene (TCE) or chloroform (CHC13).
Toxicologist 13:107.
Peden-Adams MM, Eudaly JG, Heesemann LM, Smythe J, Miller J, Gilkeson GS, et al. 2006.
Developmental immunotoxicity of trichloroethylene (TCE): studies in B6C3F1 mice. J
Environ Sci Health A Tox Hazard Subst Environ Eng 41:249-271.
Peden-Adams MM, Eudaly JG, Lee AM, Miller J, Keil DE, Gilkeson GS. 2008. Lifetime
exposure to trichloroethylene (TCE) does not accelerate autoimmune disease in MRL +/+
mice. J Environ Sci Health A Tox Hazard Subst Environ Eng 43:1402-1409.
Perdue MP, Severson RK, Colt JS, Stewart P, De Roos AJ, Cerhan JR, Cozen W, Davis S,
Hartge P, Schenk M, Blair A. In-Press. Degreasing and risk of non-Hodgkin lymphoma.
Occup Environ Med doi: 10.1136/oem.2008.040386.
Persson B, Fredrikson M. 1999. Some risk factors for non-Hodgkin's lymphoma. Int J Occup
Med Environ Health 12:135-142.
6/22/2009
701
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Povey A, Guppy MJ, Wood M, Knight C, Black CM, Silman AJ. 2001. Cytochrome P2
polymorphisms and susceptibility to scleroderma following exposure to organic solvents.
Arthritis Rheum 44:662-665.
Radican L, Blair A, Stewart P, Wartenberg D. 2008. Mortality of aircraft maintenance workers
exposed to trichloroethylene and other hydrocarbons and chemicals: extended follow-up.
J Occup Environ Med 50:1306-1319.
Raaschou-Nielsen O, Hansen J, McLaughlin JK, Kolstad H, Christensen JM, Tarone RE, et al.
2003. Cancer risk among workers at Danish companies using trichloroethylene: a cohort
study. Am J Epidemiol 158:1182-1192.
Richardson DB, Terschuren C, Hoffmann W. 2008. Occupational risk factors for non-Hodgkin's
lymphoma: a population-based case-control study in Northern Germany. Am J Ind Med
51:258-268.
Ritz B. 1999. Cancer mortality among workers exposed to chemicals during uranium processing.
J Occup Environ Med 41:556-566.
Sanders VM, Tucker AN, White KL, Jr., Kauffmann BM, Hallett P, Carchman RA, et al. 1982.
Humoral and cell-mediated immune status in mice exposed to trichloroethylene in the
drinking water. Toxicol Appl Pharmacol 62:358-368.
Seidler A, Mohner M, Berger J, Mester B, Deeg E, Eisner G, et al. 2007. Solvent exposure and
malignant lymphoma: a population-based case-control study in Germany. J Occup Med
Toxicol 2:2.
Seo M, Ikeda K, Okamura T, Kida K, Satoh M, Inagaki N, et al. 2008. Enhancing effect of
chlorinated organic solvents on histamine release and inflammatory mediator production.
Toxicology 243:75-83.
Shu XO, Stewart P, Wen WQ, Han D, Potter JD, Buckley JD, et al. 1999. Parental occupational
exposure to hydrocarbons and risk of acute lymphocytic leukemia in offspring. Cancer
Epidemiol Biomarkers Prev 8:783-791.
Shu XO, Perentesis JP, Wen W, Buckley JD, Boyle E, Ross JA, et al. 2004. Parental exposure to
medications and hydrocarbons and ras mutations in children with acute lymphoblastic
leukemia: a report from the Children's Oncology Group. Cancer Epidemiol Biomarkers
Prev 13:1230-1235.
Siemiatycki J. 1991. Risk Factors for Cancer in the Workplace. Boca Raton: CRC Press.
6/22/2009
702
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Sinks T, Lushniak B, Haussler BJ, Sniezek J, Deng JF, Roper P, et al. 1992. Renal cell cancer
among paperboard printing workers. Epidemiology 3:483-489.
Spielberg SP. 1996. N-acetyltransferases: pharmacogenetics and clinical consequences of
polymorphic drug metabolism. J Pharmacokinet Biopharm 24:509-519.
Sung TI, Wang JD, Chen PC. 2008. Increased risk of cancer in the offspring of female
electronics workers. Reprod Toxicol 25:115-119.
Sverdrup B, Kallberg H, Bengtsson C, Lundberg I, Padyukov L, Alfredsson L, et al. 2005.
Association between occupational exposure to mineral oil and rheumatoid arthritis:
results from the Swedish EIRA case-control study. Arthritis Res Ther 7:R1296-1303.
't Mannetje A, Dryson E, Walls C, McLean D, McKenzie F, Maule M, et al. 2008. High risk
occupations for non-Hodgkin's lymphoma in New Zealand: case-control study. Occup
Environ Med 65:354-363.
Tang XJ, Li LY, Huang JX, Deng YY. 2002. Guinea pig maximization test for trichloroethylene
and its metabolites. Biomed Environ Sci 15:113-118.
Tang X, Que B, Song X, Li S, Yang X, Wang H, et al. 2008. Characterization of liver injury
associated with hypersensitive skin reactions induced by trichloroethylene in the guinea
pig maximization test. J Occup Health 50:114-121.
Tatham L, Tolbert P, Kjeldsberg C. 1997. Occupational risk factors for subgroups of non-
Hodgkin's lymphoma. Epidemiology 8:551-558.
Vartiainen T, Pukkala E, Rienoja T, Strandman T, Kaksonen K. 1993. Population exposure to
tri-and tetrachloroethene and cancer risk: Two cases of drinking water pollution.
Chemosphere 27:1171-1181.
Vineis P, Miligi L, Costantini AS. 2007. Exposure to solvents and risk of non-Hodgkin
lymphoma: clues on putative mechanisms. Cancer Epidemiol Biomarkers Prev 16:381-
384.
Wang G, Ansari GA, Khan MF. 2007a. Involvement of lipid peroxidation-derived aldehyde-
protein adducts in autoimmunity mediated by trichloroethene. J Toxicol Environ Health
A 70:1977-1985.
Wang G, Cai P, Ansari GA, Khan MF. 2007b. Oxidative and nitrosative stress in
trichloroethene-mediated autoimmune response. Toxicology 229:186-193.
6/22/2009
703
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Wang G, Konig R, Ansari GA, Khan MF. 2008. Lipid peroxidation-derived aldehyde-protein
adducts contribute to trichloroethene-mediated autoimmunity via activation of CD4(+) T
cells. Free Radic Biol Med.
Wang R, Zhang Y, Lan Q, Holdford TR, Leaderer B, Zahm SH, Boyle P, et al. 2009.
Occupational exposure to solvents and risk of non-Hodgkin lymphoma in Connecticut
women. Am J Epidemiol 169:176-185.
Wartenberg D, Reyner D, Scott CS. 2000. Trichloroethylene and cancer: epidemiologic
evidence. Environ Health Perspect 108 Suppl 2:161-176.
Weisenburger DD. 1992. Pathological classification of non-Hodgkin's lymphoma for
epidemiological studies. Cancer Res 52:5456s-5464s.
White KL, David DW, Butterworth LF, Klykken PC. 2000. Assessment of autoimmunity-
inducing potential using the brown Norway rat challenge model. Toxicol Lett 112-
113:443-451.
Wilcosky TC, Checkoway H, Marshall EG, Tyroler HA. 1984. Cancer mortality and solvent
exposures in the rubber industry. Am Ind Hyg Assoc J 45:809-811.
Woolhiser, MR; Krieger, SM; Thomas, J; Hotchkiss, JA. (2006) Trichloroethylene (TCE):
Immunotoxicity potential in CD rats following a 4-week vapor inhalation exposure. Dow
Chemical Company, Toxicology & Environmental Research and Consulting, Midland,
MI, Study ID 031020, July 5, 2006, unpublished.
Wright PF, Thomas WD, Stacey NH. 1991. Effects of trichloroethylene on hepatic and splenic
lymphocytotoxic activities in rodents. Toxicology 70:231-242.
Zhao Y, Krishnadasan A, Kennedy N, Morgenstern H, Ritz B. 2005. Estimated effects of
solvents and mineral oils on cancer incidence and mortality in a cohort of aerospace
workers. Am J Ind Med 48:249-258.
6/22/2009 704
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.6 Respiratory tract toxicity and cancer
4.6.1 Epidemiologic Evidence
4.6.1.1 Chronic Effects: Inhalation
Two reports of a study of 1,091 gun-manufacturing workers are found on non-cancer
pulmonary toxicity (Cakmak et al., 2004; Saygun et al., 2007). A subset of these workers
(n = 411) had potential exposure to multiple organic solvents including toluene, acetone, butanol,
xylene, benzene and TCE used to clean gun parts; however, both papers lacked information on
exposure concentration. Mean exposure duration in Cakmak et al. (2004) was 17 years
(SD = 7.9) for nonsmokers and 16 years (SD = 7.1) for smokers. Cakmak et al. (2004) indicated
effects of smoking and exposure to solvents, with smoking having the most important effect on
asthma-related symptoms [smoking, OR = 2.8, 95% CI: 2.0, 3.8; solvent exposure, OR = 1.4,
95% CI: 1.1, 1.9], Similarly, smoking, but not solvent exposure, was shown as a statistically
significantly predictor of lung function decrements. Saygun et al. (2007) reported on a five year
follow-up of 393 of the original 1,091 subjects, 214 of who were exposed to solvents. Of the
393 original subjects, the prevalence of definitive asthma symptoms, a more rigorous definition
than used by Cakmak et al. (2004), was 3.3% among exposed and 1.1% among non-exposed
subjects, p>0.05. Saygun et al. (2007) presents observations on lung function tests for 697
current workers, a group which includes the 393 original study subjects. Smoking, but not
solvent exposure, was a predictor of mean annual forced expiratory volume (FEVi) decrease.
4.6.1.2 Cancer
Cancers of the respiratory tract including the lung, bronchus, and trachea are examined in
23 cohort, community studies and case-control studies of TCE. Twelve of the 23 studies
approached standards of epidemiologic design and analysis identified in the systematic review of
the epidemiologic body of literature on TCE and cancer [see Appendix B] (Siemiatycki, 1991;
Axelson et al., 1994; Greenland et al., 1994; Anttila et al., 1995; Blair et al., 1998; Morgan et al.,
1998; Boice et al., 1999, 2006; Hansen et al., 2001; Raaschou-Nielsen et al., 2003; Zhao et al.,
2005; Radican et al., 2008). Cancers at other sites besides lung, bronchus, and trachea in the
respiratory system are more limitedly reported in these studies. Some information is available on
laryngeal cancer; however, only 8 of the 15 occupational cohort studies providing information on
lung cancer also reported findings for this site. Case-control studies of lung or laryngeal cancers
and occupational title or organic solvent exposure were found in the literature. Two case-control
studies of lung cancer, one population-based and the other nested within a cohort, were of TCE
6/22/2009
705
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
exposure specifically. Lung and laryngeal cancer risk ratios reported in cohort, community and
case-control studies are found in Table 4.6.1.
Lung cancer relative risks were reported in 11 of 12 cohort studies of aircraft
manufacturing, aircraft maintenance, aerospace, and metal workers, with potential exposure to
TCE as a degreasing agent, and in occupational cohort studies employing biological markers of
TCE exposures. All 11 studies had a high likelihood of TCE exposure in individual study
subjects and were judged to have met, to a sufficient degree, the standards of epidemiologic
design and analysis (Axelson et al., 1994; Greenland et al., 1994; Anttila et al., 1995; Blair et al.,
1998; Morgan et al., 1998; Boice et al., 1999, 2006; Hansen et al., 2001; Raaschou-Nielsen et al.,
2003; Zhao et al., 2005; Radican et al., 2008). Lung cancer risks were not reported for Fernald
uranium processing workers with potential TCE exposure (Ritz, 1999), a study of less weight
than the other 11 studies.. The incidence study of Raaschou-Nielsen et al. (2003) was the largest
cohort, with 40,049 subjects identified as potentially exposed to TCE in several industries
(primarily, in the iron/metal and electronic industries), including 14,360 of whom had
presumably higher level exposures to TCE. The study included 632 lung cancer cases and
reported a 40% elevated incidence in TCE exposed males and females combined (95% CI: 1.32,
1.55), with no exposure duration gradient. The 95% confidence intervals in other studies of lung
cancer incidence included a risk ratio of 1.0 (Axelson et al., 1994; Anttila et al., 1995; Blair et
al., 1998; Hansen et al., 2001; Zhao et al., 2005). Lung cancer mortality risks in studies of TCE
exposure to aircraft manufacturing, aircraft maintenance, and aerospace workers included a
relative risk of 1.0 in their 95% confidence intervals (Boice et al., 2006; Zhao et al., 2005;
Morgan et al., 1998; Blair et al., 1998). Boice et al. (1999) observed a 24% decrement (95% CI:
0.60, 0.95) for subjects with routine TCE exposure. Exposure-response analyses using internal
controls (unexposed subjects at the same company) showed a statistically significant decreasing
trend between lung cancer risk and routine or intermittent TCE exposure duration. The routine
or intermittent category is broader and includes more subjects with potential TCE exposure.
The population studied by Costa et al. (1989), Garabrant et al. (1998), ATSDR (2004)
and Chang et al. (2005) are all employees (white- and blue-collar) at a manufacturing facility or
plant with potential TCE exposures. Garabrant et al. (1988) observed a 20% deficit in lung
cancer mortality (95% CI: 0.68, 0.95) in their study of all employees working for 4 or more years
at an aircraft manufacturing company. Confidence intervals (95% CI) in Costa et al. (1989),
Chang et al. (2005) and ATSDR (2004) included a risk of 1.0. TCE exposure was not known for
individual subjects in these studies. A wide potential for TCE exposure is likely ranging from
subjects with little to no TCE exposure potential to those with some TCE exposure potential.
Exposure misclassification bias, typically considered as a negative bias, is likely greater in these
studies compared to studies adopting more sophisticated exposure assessment approaches, which
6/22/2009
706
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
are able to assign quantitative exposure metrics to individual study subjects. All three studies
were of lower likelihood for TCE exposure, in addition to limited statistical power and other
design limitations, and these aspects, in addition to potential exposure misclassification bias were
alternative explanations of observed findings.
One population case-control study examined the relationship between lung cancer and
TCE exposure (Siemiatycki et al., 1991) with risk ratios of 0.9 (95% CI: 0.6, 1.5) for any TCE
exposure and 0.6 (95% CI: 0.3, 1.2) for substantial TCE exposure after adjustment for cigarette
smoking. TCE exposure prevalence in cases in this study was 2.5% for any exposure. Only 1%
had "substantial" (author's term) exposure, limiting the sensitivity of this study. Relative risks
above 2.0 could only be detected with sufficient (80%) statistical power. The finding of no
association of lung cancer with TCE exposure, therefore, is not surprising. One nested case-
control study of rubber workers observed a smoking unadjusted risk of 0.64 (95% CI not
presented in paper) in those who had >1 year cumulative exposure to TCE (Wilcosky et al.,
1984).
Three geographic based studies reported lung cancer incidence or mortality risks for
drinking water contamination with TCE (Isacson et al., 1985; Morgan and Cassidy, 2002;
ATSDR, 2006). Morgan and Cassidy (2002) observed a relative risk of 0.71 (99% CI: 0.61,
0.81) for lung cancer among residents of Redlands County, CA, whose drinking water was
contaminated with TCE and perchlorate. However, ATSDR (2006) reported a 28% increase
(95%) CI: 0.99, 1.62) in lung cancer incidence among residents living in a area in Endicott, NY,
whose drinking water was contaminated with TCE and other solvents. No information on
smoking patterns is available for individual lung cancer cases as identified by NYDOH for other
cancer cases in this study (ATSDR, 2008). Isacson et al. (1985) presented lung cancer age-
adjusted incidence rates for Iowa residents by TCE level in drinking water supplies and did not
observe an exposure-response gradient. Exposure information is inadequate in all three of these
studies, with monitoring data, if available, based on few samples and for current periods only,
and no information on water distribution, consumption patterns, or temporal changes. Thus,
TCE exposure potential to individual subjects was not known with any precision, introducing
misclassification bias, and greatly limiting their ability to inform evaluation of TCE and lung
cancer.
Laryngeal cancer risks are presented in a limited number of cohort studies involving TCE
exposure. No case-control or geographic based studies of TCE exposure were found in the
published literature. All but one of the cohort studies providing information on laryngeal cancer
observed less that 5 incident cases or deaths. Accordingly, these studies are limited for
examining the relationship between TCE exposure and laryngeal cancer. Risk ratios for
laryngeal cancer are found in Table 4.6.2.
6/22/2009
707
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
In summary, studies in humans examining lung and laryngeal cancer and TCE exposure
are inconclusive and do not support either a positive or a negative association between TCE
exposure and lung cancer or laryngeal cancer. Raaschou-Nielsen et al. (2003), with the largest
numbers of lung cancer cases of all studies, was the only one to observe a statistically
significantly elevated lung cancer risk with TCE exposure. Raaschou-Nielsen et al. (2003) also
noted several factors that may have confounded or biased their results in either a positive or
negative direction. This study and other cohort studies, as with almost any occupational study,
were not able to control confounding by exposure to chemicals other than TCE (although no
such chemical was apparent in the reports). Information available for factors related to
socioeconomic status (e.g., diet, smoking, alcohol consumption) was also not available. Such
information may positively confound smoking-related cancers such as lung cancer, particularly
in those studies, which adopted national rates to derive expected numbers of site-specific cancer,
if greater smoking rates were over-represented in blue-collar workers or residents of lower socio-
economic status. The finding of a larger risk among subjects with shortest exposure also argues
against a causal interpretation for the observed association for all subjects (NRC, 2006).
Three studies reported a statistically significant deficit in lung cancer incidence
(Garabrant et al., 1988; Boice et al., 1999; Morgan and Cassidy, 2002). Absence of smoking
information in these studies would introduce a negative bias if the studied population smoked
less than the referent population and may partially explain the lung cancer decrements observed
in these studies. Morgan and Cassidy (2002) noted the relatively high education high income
levels, and high access to health care of subjects in this study compared to the averages for the
county as a whole, likely leading to a lower smoking rate compared to their referent population.
Garabrant et al. (1988) similarly attributed their observations to negative selection bias
introduced when comparison is made to national mortality rates, also known as a "healthy
worker effect." The statistically significant decreasing trend in Boice et al. (1999) with exposure
duration to intermittent or routine exposure may reflect a protective effect between TCE and lung
cancer. The use of internal controls in this analysis reduces bias associated with use of an
external population who may have different smoking patterns than an employed population.
However, the exposure assessment approach in this study is limited due to inclusion of subjects
identified with intermittent TCE exposure (i.e., workers who would be exposed only during
particular shop runs or when assisting other workers during busy periods) (Boice et al., 1999).
The Boice et al. (1999) analysis is based on twice as many lung cancer deaths (i.e., 173 lung
cancer deaths) among subjects with routine or intermittent TCE exposure compared to only
routinely exposed subjects (78 deaths). Subjects identified as intermittently exposed are
considered as having a lower exposure potential than routinely exposed subject and their
inclusion in exposure-response analyses may introduce exposure misclassification bias. Such
6/22/2009
708
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
bias is a possible explanation for the decreasing trend observation, particularly if workers with
lower potential for TCE exposure have longer exposure (employment) durations.
Thus, a qualitative assessment suggests the epidemiological literature on respiratory
cancer and TCE is quite limited and has sufficient power to detect only large relative risks.
These studies can only rule out risks of a magnitude of 2.0 or greater for lung cancer and relative
risks greater than 3.0 or 4.0 for laryngeal cancer for exposures to studied populations. Therefore,
the database is limited in its ability to detect lung cancer associated with TCE exposure,
especially if the magnitude of response is similar to those observed for other endpoints.
6/22/2009
709
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.6.1: Selected Results from Epidemiologic Studies of TCE Exposure and Lung Cancer
No. obs.
Exposure Group Relative Risk (95% CI) events Reference
Cohort Studies - Incidence
Aerospace workers (Rocketdyne)
Any exposure to TCENot reported
Low cum TCE scorel.OO1
Med cum TCE scorel.36 (0.86, 2.14)
High TCE score 1.11 (0.60, 2.06)
p for trend0.60
43
35
14
Zhao et al., 2005
All employees at electronics factory (Taiwan)
1.07 (0.72, 1.52)
30
Chang et al., 2005
Danish blue-collar worker w/TCE exposure
Any exposure, all subjects
Raaschou-Nielsen et al., 2003
1.4 (1.32, 1.55)
Employment duration
Any exposure, malesl.4 (1.28, 1.51)
Any exposure, femalesl.9 (1.48, 2.35)
<1 yearl.7 (1.46, 1.93)
1-4.9 yearsl.3 (1.16, 1.52)
> 5 yearsl.4 (1.23, 1.63)
632
559
73
209
218
205
Biologically-monitored Danish workers
Any TCE exposure, males
Any TCE exposure, females
Cumulative exp (Ikeda)
Mean concentration (Ikeda)
Employment duration
0.8(0.5, 1.3)
0.7 (0.01,3.8)
Not reported
<17 ppm-yr
>17 ppm-yr
Not reported
<4 ppm
4+ ppm
Not reported
< 6.25 yr
>6.25
16
1
Hansen etal., 2001
Aircraft maintenance workers (Hill Air Force Base, Utah)
TCE Subcohort
Males, Cumulative exp
Not reported
01.01
< 5 ppm-yrl.O (0.6, 2.0)
24
Blair etal., 1998
6/22/2009
710
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Females, Cumulative exp
5-25 ppm-yr0.8 (0.4, 1.6)
>25 ppm-yr0.8 (0.4, 1.7)
01.01
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
11
15
Biologically-monitored Finnish workers
All subjects
Mean air-TCE (Ikeda extrapolation)
Anttila et al„ 1995
0.92 (0.59, 1.35)
<6 ppml.02 (0.58, 1.66)
6+ ppm0.83 (0.33, 1.71)
25
16
7
Biologically-monitored Swedish workers
Any TCE exposure, males0.69 (0.31, 1.30)
Any TCE exposure, femalesNot reported
Cohort-Mortality
Computer manufacturing workers (IBM), NY
Males 1.03 (0.71, 1.42)
Females 0.95 (0.20, 2.77)
Aerospace workers (Rocketdyne)
Any TCE (utility or engine flush workers) 1.24 (0.92, 1.63)
Engine Flush - Duration of Exposure
Referentl.01
0 year (Utility workers w/ TCE exp)0.5 (0.22, 1.00)
<4 years0.8 (0.50, 1.26)
> 4 years0.8 (0.46, 1.41)
Axelson et al., 1994
35
3
51
472
7
27
24
Clapp and Hoffman 2008
Boice et al., 2006
Any exposure to TCENot reported
Low cum TCE scorel.001
Med cum TCE scorel.05 (0.76, 1.44)
High TCE score 1.02 (0.68, 1.53)
pfortrend0.91
99
62
33
Zhao et al., 2005
View-Master employees
Males
Females
0.81 (0.42, 1.42)2
0.99 (0.71, 1.35)2
12
41
ATSDR, 2004
US Uranium-processing workers (Fernald)
Any TCE exposureNot reported
Light TCE exposure, >2 years duration4Not reported
Mod TCE exposure, >2 years duration4Not reported
Ritz, 1999
6/22/2009
711
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Aerospace workers (Lockheed)
Routine exposure 0.76 (0.60,0.95) 78
Routine-Intermittent exposure1 Not reported 173
Duration of exposure
Oyearsl.O 288
< lyear0.85 (0.65, 1.13) 66
1-4 years0.98 (0.74, 1.30) 63
> 5 years0.64 (0.46, 0.89) 44
Trend testp<0.05
Boice et al., 1999
Aerospace workers (Hughes)
TCE Subcohort
Morgan etal., 1998
1.10 (0.89,1.34) 97
Low Intensity (<50 ppm)1.49 (1.09, 1.99) 45
High Intensity (>50 ppm)0.90 (0.67, 1.20) 52
TCE Subcohort (Cox Analysis)2
Never exposedl.001 291
Ever exposedl. 14 (0.90, 1.44) 97
Peak
No/Lowl.001 324
Med/Hil.07 (0.82, 1.40) 64
Cumulative
Referentl.001 291
Low 1.47 (1.07, 2.03) 45
High0.96 (0.72, 1.29) 52
Aircraft maintenance workers (Hill Air Force Base, Utah)
TCE Subcohort
Any TCE exposure0.9 (0.6, 1.3)1
Males, Cumulative exp
01.01
< 5 ppm-yrl.O (0.7, 1.6)
5-25 ppm-yr0.9 (0.5, 1.6)
>25 ppm-yrl.l (0.7, 1.8)
Blair etal., 1998
Females, Cumulative exp
01.01
< 5 ppm-yr0.6 (0.1, 2.4)
5-25 ppm-yr0.6 (0.1, 4.7)
>25 ppm-yr0.4 (0.1, 1.8)
109
51
43
23
38
2
2
11
2
TCE Subcohort
Males, Cumulative exp
Any TCE exposure 0.83 (0.63,1.08) 166
0.91 (0.67, 1.24) 155
Radican et al. (2008)
6/22/2009
712
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Females, Cumulative exp
0
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
0
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
1.01
0.96 (0.67, 1.37)
0.71 (0.46, 1.11)
1.00 (0.69, 1.45)
0.53 (0.27, 1.07)
1.01
0.69 (0.27, 1.77)
0.65 (0.16,2.73)
0.39(0.14, 1.11)
66
31
58
11
5
2
4
Cardboard manufacturing workers in Arnsburg, Germany
TCE exposed workers
Unexposed workers
1.38 (0.55,2.86)
1.06 (0.34, 2.47)
Henschler et al., 1995
Deaths reported to GE pension fund (Pittsfield, MA)
1.01 (0.69, 1.47)
139 Greenland et al., 1994
Aircraft manufacturing employees (Italy)
All employees
0.99 (0.73, 1.32)
99
Costa et al., 1989
Aircraft manufacturing plant employees (San Diego, CA)
All subjects
0.80 (0.68, 0.95)
138
Garabrant et al., 1988
Rubber industry workers (Ohio)
0.64 (p>0.05)'
11 Wilcosky et al., 1984
Case-control Studies
Population of Montreal, Canada
Any TCE exposure
Substantial TCE exposure
0.9 (0.6, 1.5)
0.6 (0.3, 1.2)5
21
9
Siemiatycki et al., 1991
Geographic Based Studies
Two study areas in Endicott, NY
1.28 (0.99, 1.62)
68 ATSDR, 2006
Residents of 13 census tracts
in Redland, CA
0.71 (0.61, 0.81)6
356
Morgan and Cassidy, 2002
Iowa residents with TCE in water supply
Males
<0.15 ug/L343.17
>0.15 ug/L345.77
Females
<0.15 ug/L 58.77
1,181
299
289
Isacson et al., 1985
6/22/2009
713
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
>0.15 ug/L 47.87 59
1 Internal referents, workers not exposed to TCE
2 Risk ratio from Cox Proportional Hazard Analysis, stratified by age, sex, and decade (Environmental Health
Strategies, 1997)
3 Odds ratio from nested case-control analysis4 Odds ratio from nested case-control study
5 90% Confidence Interval
6 99% Confidence Interval
7 Average annual age-adjusted incidence (per 100,000)
6/22/2009
714
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.6.2: Selected Results from Epidemiologic Studies of TCE Exposure and Laryngeal
Cancer
Exposure Group
Cohort Studies - Incidence
Aerospace workers with TCE exposure
Relative Risk
(95% CI) No. obs. events Reference
Not reported
Zhao et al., 2005
Danish blue-collar worker w/TCE exposure
Employment duration
Raaschou-Nielsen et al., 2003
Any exposure, malesl.2 (0.87, 1.52)
Any exposure, femalesl.7 (0.33, 4.82)
Not reported
<1 year
1-4.9 years
> 5 years
53
3
Biologically-monitored Danish workers
Hansen etal., 2001
Any TCE exposure, males 1.1 (0.1, 3.9)
Any TCE exposure, females
Cumulative exp (Ikeda)Not reported
<17 ppm-yr
>17 ppm-yr
Mean concentration (Ikeda)Not reported
<4 ppm
4+ ppm
Employment durationNot reported
< 6.25 yr
>6.25
0 (0.1 exp)
Aircraft maintenance workers (Hill Air Force Base, Utah)
TCE Subcohort
Blair etal., 1998
Males, Cumulative exp
Females, Cumulative exp
Any exposureNot reported
Not reported
0
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
Not reported
0
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
6/22/2009
715
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Biologically-monitored Finnish workers Not reported
Mean air-TCE (Ikeda extrapolation from U-
TCA) Not reported
<6 ppm
6+ ppm
Anttila et al., 1995
Biologically-monitored Swedish workers
Any TCE exposure, males
Any TCE exposure, females
Cohort-Mortality
Computer manufacturing workers (IBM), NY
1.39 (0.17,5.00) 2
Not reported
Not reported
Axelsonetal., 1994
Clapp and Hoffman (2008)
Aerospace workers (Rocketdyne)
Any TCE (utility or engine flush workers) 1.45 (0.18, 5.25)
Engine Flush - Duration of ExposureNot reported
Referent
0 year (Utility workers w/ TCE exp)
<4 years
> 4 years
Boice et al., 2006
View-Master employees
Males
Females
Any exposure to TCENot reported
Not reported
Zhao et al., 2005
ATSDR, 2004
All employees at electronic factory (Taiwan)
Males
FemalesO
Chang et al., 2003
0 (0.90 exp)
0 (0.23 exp)
US Uranium-processing workers (Fernald)
Any TCE exposureNot reported
Light TCE exposure, >2 years duration4Not reported
Mod TCE exposure, >2 years duration4Not reported
Ritz, 1999
Aerospace workers (Lockheed)
Routine exposure
Routine-Intermittent exposure
1.10 (0.30,2.82) 4
Not reported
Boice et al., 1999
Aerospace workers (Hughes)
TCE Subcohort
Not reported
Low Intensity (<50 ppm)
Morgan etal., 1998
6/22/2009
716
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
High Intensity (>50 ppm)
Peak
Cumulative
Not reported
No/Low
Med/Hi
Not reported
Referent
Low
High
Aircraft maintenance workers (Hill Air Force Base, Utah)
TCE subcohort
Males, Cumulative exp
Not reported
Not reported
0
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
Females, Cumulative exp
Not reported
0
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
Blair etal., 1998
Cardboard manufacturing workers in Arnsburg, Germany
Not reported
Henschler et al., 1995
Deaths reported to GE pension fund (Pittsfield, MA)
Not examined
Greenland et al., 1994
Aircraft manufacturing employees (Italy)
All employees
0.27 (0.03, 0.98)
Costa etal., 1989
Aircraft manufacturing plant employees (San Diego, CA)
All subjects
Garabrant et al., 1988
0 (7.41 exp)
6/22/2009
717
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.6.2 Laboratory Animal Studies
4.6.2.1 Respiratory Tract Animal Toxicity
Limited studies are available to determine the effects of TCE exposure on the respiratory
tract. Many of these studies in mice have examined acute effects following intraperitoneal
administration at relatively high TCE doses. However, effects on the bronchial epithelium have
been noted in mice and rats with TCE administered via gavage, with doses 1,000 mg/kg-d and
higher reported to cause rales and dyspnea (Narotsky et al., 1995) and pulmonary vasculitis
(NTP, 1990) in rats. Mice appear to be more sensitive than rats to histopathological changes in
the lung via inhalation; pulmonary effects are also seen in rats with gavage exposure. It is
difficult to compare intraperitoneal to oral and inhalation routes of exposure given the risk of
peritonitis and paralytic ileus. Any inflammatory response from this route of administration can
also affect the pulmonary targets of TCE exposure such as the Clara cells.
This section reviews the existing literature on TCE, and the role of the various TCE
metabolites in TCE-induced lung effects. The most prominent toxic effect reported is damage to
Clara cells in mouse lung. The nonciliated, columnar Clara cells comprise the majority of the
bronchiolar and terminal bronchiolar epithelium in mice, and alveolar type I and type II cells
constitute the alveolar epithelium. These cells have been proposed as a progenitor of lung
adenocarcinomas in both humans and mice (Kim et al., 2005). Long-term studies have not
focused on the detection of pulmonary adenoma carcinomas but have shown a consistently
positive response in mice but not rats. However, chronic toxicity data on noncancer effects is
very limited.
4.6.2.1.1 Acute and Short-term effects: Inhalation
Relatively high-dose single and multiple inhalation exposures to TCE result in dilation of
endoplasmic reticulum and vacuolation of nonciliated (Clara) cells throughout the bronchial tree
in mice. A single study in rats reported similar findings. In mice, single exposure experiments
show vacuolation at all dose levels tested with the extent of damage increasing with dose.
Villaschi et al. (1991) reported similar degrees of vacuolation in B6C3F1 mice (3/group) at 24 hr
after the start of exposure across all tested doses (500, 1,000, 2,000, 3,500, and 7,000 ppm, 30
min), with the percentage of the nonciliated cells remaining vacuolated at 48 hr increasing with
dose. Clara cell vacuolation was reported to be resolved 7 days after single 30 min exposure to
TCE. Odum et al. (1992) reported that, when observed 24 h after the start of 6h exposure, the
majority of Clara cells in mice were unaffected at the lowest dose of 20 ppm exposures, while
6/22/2009
718
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
marked vacuolation was observed at 200 ppm (no quantitative measures of damage given and
only 3 animals per group were examined).
In rats, Odum et al. (1992) reported no morphological changes in the female Alpk
APfSD rat epithelium after 6 h exposure (500 ppm or 1,000 ppm) when observed 24 hr after the
start of exposure (n = 3/group). However, Kurasawa reported pronounced dose-related
morphological changes in Clara cells at the highest dose (8,000 ppm) for 2 hr in Wistar rats (n =
10 per group). At 500 and 1,000 ppm, slight dilation of the apical surface was reported, but
morphological measurements (the ratio of the lengths of the apical surface to that of the base line
of apical cytoplasm) were not statistically-significantly different from controls. From 2,000 to
8,000 ppm, a progressively increasing flattening of the apical surface was observed. In addition,
at 2,000 ppm, slight dilation of the smooth endoplasmic reticulum was also observed, with
marked dilation and possible necrosis at 8,000 ppm. Kurasawa (1988) also examined the time-
course of Clara cell changes following a single 8,000-ppm exposure, reporting the greatest
effects at 1 day to 1 week, repair at 2 weeks, and nearly normal morphology at 4 weeks. The
only other respiratory effect that has been reported from one study in rats exposed via inhalation
is a reduction in pulmonary surfactant yield following 30 min exposures at 9,030 ppm for 5 or 15
days (Stewart et al., 1979). Therefore, single inhalation experiments (Villaschi et al., 1991;
Odum et al., 1992; Kurasawa, 1988) suggest that the Clara cell is the target for TCE exposure in
both rats and mice and that mice are more susceptible to these effects. However the database is
limited in its ability to discern quantitative differences in susceptibility or the nature of the dose-
response after a single dose of TCE.
Other experiments examined the effects of several days of TCE inhalation exposure in
mice and potential recovery. While single exposures require 1 to 4 weeks for complete recovery,
after short-term repeated exposure, the bronchial epithelium in mice appears to either adapt to or
become resistant to damage Odum et al. (1992) and Green et al. (1997) observed Clara cells in
mice to be morphologically normal at the end of exposures 6 hr/day for 4 or 5 days. As with
single dose experiments, the extent of recovery in multi-dose exposures may be dose-dependent.
Using a very high dose, Lewis et al. (1984) report vacuolation of bronchial epithelial cells after 4
hr/day, but not lhr/day, (10,000 ppm) for 5 days in mice. In addition, Odum et al. (1992)
reported that the damage to Clara cells that resolved after repeated exposures of 5 days, a sign of
adaptation to TCE exposure, returned when exposure was resumed after 2 days.
In rats, only one inhalation study reported in two published articles (Stewart et al., 1979;
Le Mesurier et al., 1979) using repeated exposures examined pulmonary histopathology.
Interestingly, this study reported vacuolation in Type 1 alveolar cells, but not in Clara cells, after
5 days of exposure to approximately 9,030 ppm for 30 min per day (only dose tested). In
addition, abnormalities were observed in the endothelium (bulging of thin endothelial segments
6/22/2009
719
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
into the microcirculatory lumen) and minor morphological changes in Type 2 alveolar cells.
Although exposures were carried out for 5 consecutive days, histopathology was recorded up to
15 days post exposure, giving cell populations time to recover. Because earlier time points were
not examined, it is not possible to discern whether the lack of reported Clara cell damage in rats
following repeated exposure is due to recovery or lack of toxicity in this particular experiment.
Although recovery of individual damaged cells may occur, cell proliferation, presumed
from labeling index data suggestive of increased DNA synthesis, contributes, at least in part, to
the recovery of the bronchial epithelium in mice. Villaschi et al. (1991) observed a dose-
dependent increase in labeling index as compared to controls in the mouse lung at 48h after a
single TCE exposure (30 min; 500, 1,000, 2,000, 3,500, 7,000 ppm), which decreased to baseline
values at 7 days post-exposure. Morphological analysis of cells was not performed, although the
authors stated the dividing cells had the appearance of Clara cells. Interestingly, Green et al.
(1997) reported no increase in BrdU labeling 24 hr after a single exposure (6 hr 450 ppm), but
did see increased BrdU labeling at the end of multiple exposures (1/d, 5d) while Villaschi et al.
(1991) reported increased [3H]Thymidine labeling 2, 5, and 7 days after single 30 minute
exposures to 500-7,000 ppm. Therefore, the data for single exposures at 450-500 ppm may be
consistent if increased cell proliferation occurred only for a short period of time around 48 hr
post-exposure, and was thereby effectively washed-out by the longer "averaging time" in the
experiments by Green et al. (1997). Also, these contradictory results may be due to differences
in methodology. Green et al. (1997) and Villaschi et al. (1991) reported very different control
labeling indices (6% and 0%, respectively) while reporting similar absolute labeling indices at
450-500 ppm (6.5% and 5.2%, respectively). The different control values may be a result of
substantially-different times over which the label was incorporated: the mice in Green et al.
(1997) were given BrdU via a surgically-implanted osmotic pump over four days prior to
sacrifice, while the mice in Villaschi et al. (1991) were given a single intraperitoneal dose of
[3H]Thymidine 1 hour prior to sacrifice. Stewart et al. (1979) observed no stimulation of
thymidine incorporation after daily exposure to TCE (9,000 ppm) for up to 15 days. This study
did, however, report a non-statistically significant reduction in orotate incorporation, an indicator
of RNA synthesis, after 15 days, although the data was not shown.
At the biochemical level, changes in pulmonary metabolism, particularly with respect to
P450 activity, have been reported following TCE exposure via inhalation or intraperitoneal
administration in mice. Odum et al. (1992) reported reduced enzyme activity in Clara cell
sonicates of ethoxycoumarin (9-deethylase, aldrin epoxidation, and NADPH cytochrome c
reductase after 6 hour exposures to 20-2,000 ppm TCE, although the reduction at 20 ppm was
not statistically significant. No reduction of GST activity as determined by chlorodinitrobenzene
as a substrate was detected. With repeated exposure at 450 ppm, the results were substrate-
6/22/2009
720
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
dependent, with ethoxycoumarin O-deethylase activity remaining reduced, while aldrin
epoxidation and NADPH cytochrome c reductase activity showing some eventual recovery by 2
weeks. The results reported by Odum et al. (1992) for NADPH cytochrome c reductase were
consistent with those of Lewis et al. (1984), who reported similarly reduced NADPH cytochrome
c reductase activity following a much larger dose of 10,000 ppm for 1 and 4 hr/d for 5 days in
mice (strain not specified). TCE exposure has also been associated with a decrease in pulmonary
surfactant. Repeated exposure of female Wistar rats to TCE (9,000 ppm, 30 min/day) for 5 or 15
days resulted in a significant decrease in pulmonary surfactant as compared to unexposed
controls (Le Mesurier et al., 1980).
4.6.2.1.1.1 Acute and Short-term effects: Intraperitoneal Injection and Gavage Exposure
As stated above the intraperitoneal route of administration is not a relevant paradigm for
human exposure. A number of studies have used this route of exposure to study the effects of
acute TCE exposure in mice. In general, similar lung targets are seen following inhalation or
intraperitoneal treatment in mice (Forkert et al., 2006, 1985; Forkert and Birch, 1989; Scott et al.,
1988). Inhalation studies generally reported the Clara cell as the target in mice. No lung
histopathology from intraperitoneal injection studies in rats is available. Forkert et al. (1985) and
Forkert and Birch (1989) reported vacuolation of Clara cells as soon as lh following
intraperitoneal administration of a single dose of 2,000 mg/kg in mice. At 2,500 mg/kg, both
Forkert et al. (1985) and Scott et al. (1988) reported exfoliation of Clara cells and parenchymal
changes, with morphological distortion in alveolar Type II cells and inconsistently observed
minor swelling in Type I cells at 24h post-exposure. Furthermore, at 3,000 mg/kg, Scott et al.
(1988) also reported a significant (85%) decrease in intracellularly stored surfactant
phospholipids at 24h post-exposure. These data indicate that both Clara cells and alveolar type I
and II cells are targets of TCE toxicity at these doses and using this route of administration.
Recently, Forkert et al. (2006) reported Clara cell toxicity that showed increased severity with
increased dose (pyknotic nuclei, exfoliation) at 500-1,000 mg/kg intraperitoneal doses as soon as
4h post-exposure in mice. Even at 500 mg/kg, a few Clara cells were reported with pyknotic
nuclei that were in the process of exfoliation. Damage to alveolar Type II cells was not observed
in this dose range. The study by Scott et al. (1988) examined surfactant phospholipids and
phospholipase A2 activity in male CD-I mice exposed by intraperitoneal injection of TCE (2,500
or 3,000 mg/kg, 24 h). The lower concentration led to damage to and exfoliation of Clara cells
from the epithelial lining into the airway lumen, while only the higher concentration led to
6/22/2009
721
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
changes in surfactant phospholipids. This study demonstrated an increase in total phospholipid
content in the lamellar body fractions in the mouse lung.
The study by Narotsky et al. (1995) exposed F344 timed-pregnant rats to TCE (0, 1,125
and l,500mg/kg bw) by gavage and examined both systemic toxicity and developmental effects
at 14 d postexposure. Rales and dyspnea in the dams were observed in the high dose group, with
two of the animals with dyspnea subsequently dying. The developmental effects observed in this
study are discussed in more detail in Section 4.7.
4.6.2.1.1.2 Subchronic and Chronic Effects
There are a few reports of the subchronic and chronic non-cancer effects of TCE on the
respiratory system from intraperitoneal exposure in mice and from gavage exposure in rats.
Forkert and Forkert (1994) reported pulmonary fibrosis in mice 90 days after intraperitoneal
administration of a single 2,000 mg/kg dose of TCE. The effects were in the lung parenchyma,
not the bronchioles where Clara cell damage has been observed after acute exposure. It is
possible that fibrotic responses in the alveolar region occur irrespective of where acute injury
occurs. Effects upon Clara cells can also impact other areas of the lung via cytokine regulation
(Elizur et al., 2008). Alternatively, the alveolar and/or capillary components of the lung may
have been affected by TCE in a manner that was not morphologically apparent in short-term
experiments. In addition effects from a single or a few short-term exposures may take longer to
manifest. The latter hypothesis is supported by the alveolar damage reported by Odum et al.
(1992) after chloral administration by inhalation, and by the adducts reported in alveolar type II
cells by Forkert et al. (2006) after 500-1,000 mg/kg TCE intraperitoneal administration.
As noted previously, rats have responded to short-term inhalation exposures of TCE with
Clara cell and alveolar type I and II effects. After repeated inhalation exposures over 6 weeks (8
hr/d, 5 d/wk, 730 ppm) and continuous exposures over 90 days (35 ppm), Prendergast et al.
(1967) noted no histopathologic changes in rats, guinea pigs, rabbits, dogs, or monkeys after
TCE exposure, but did describe qualitatively observing some nasal discharge in the rats exposed
for 6 weeks. The study details in Prendergast et al. (1967) are somewhat limited. Exposed
animals are described as 'typically' 15 Long-Evans or Sprague-Dawley rats, 15 Hartley guinea
pigs, 3 squirrel monkeys, 3 New Zealand albino rabbits, and 2 beagle dogs. Controls were
grouped between studies. In a 13-week NTP study in F344/N rats (// = 10/group) exposed to
TCE (0-2,000 mg/kg/d 5d/week) by gavage, pulmonary vasculitis was observed in 6/10 animals
of each sex of the highest dose group (2,000 mg/kg/d), in contrast tol/10 in controls of each sex
(NTP, 1990).
6/22/2009
722
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.6.2.2 Respiratory Tract Cancer
Limited studies have been performed examining lung cancer following TCE exposure.
TCE inhalation exposure was reported to cause statistically-significant increase in pulmonary
tumors (i.e., pulmonary adenocarcinomas) in mice but not in rats and hamsters. Oral
administration of TCE frequently resulted in elevated lung tumor incidences in mice, but not in
any tested species was there a statistically significant increase. This section will describe the
data regarding TCE induction of pulmonary tumors in rodent models. The next sections will
explore the role of metabolism and potential MO As for inhalation carcinogenicity, primarily in
mice.
4.6.2.2.1 Inhalation
There are three published inhalation studies examining the carcinogenicity of TCE at
exposures from 0-600 ppm, two of which reported statistically-significantly increased lung
tumor incidence in mice at the higher concentrations (Fukuda et al., 1983; Maltoni et al., 1986,
1988; Henschler et al., 1980). Rats and hamsters did not show an increase in lung tumors
following exposure.
The inhalation studies by Fukuda et al. (1983), which involved female ICR mice and
Sprague-Dawley rats, observed a threefold increase in lung tumors per mouse in those exposed
to the two higher concentrations (150-450ppm) but reported no increase in lung tumors in the
rats. Maltoni et al. (1986, 1988) reported statistically-significantly increased pulmonary tumors
in male Swiss and female B6C3F1 mice at the highest dose of 600 ppm, but no significant
increases in any of the other species/strains/sexes tested. Henschler et al. (1980) tested NMRI
mice, Wistar rats and Syrian hamsters of both sexes, and reported no observed increase in
pulmonary tumors any of the species tested (see Chapter 4.3 and Appendix E for details of the
conduct of these studies).
4.6.2.2.2 Gavage
None of the six chronic gavage studies, which exposed multiple strains of rats and mice
to 0-3,000 mg/kg TCE for at least 56 weeks, reported a statistically-significant excess in lung
tumors, although non-statistically-significant increases were frequently observed in mice (Van
Duuren et al., 1979; NCI, 1976; Henschler et al., 1984; NTP, 1988, 1990; Maltoni et al., 1986).
The study by Van Duuren et al. (1979) examined TCE along with 14 other halogenated
compounds for carcinogenicity in both sexes of Swiss mice. While no excess tumors were
6/22/2009
723
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
observed, the dose rate of 0.5 mg once per week is equivalent to an average dose rate of
approximately 2.4 mg/kg-day for a mouse weighing 30 g, which is about 400-fold smaller than
that in the other gavage studies. In the NCI (1976) study, the results for Osborne-Mendel rats
were considered inconclusive due to significant early mortality, but female B6C3F1 mice
(though not males) exhibited a non-statistically-significant elevation in pulmonary tumor
incidence. The NCI study (1976) used technical grade TCE which contained two known
carcinogenic compounds as stabilizers (epichlorohydrin and 1,2-epoxybutane), but a later study
by Henschler et al. (1984) in which mice were given TCE that was either pure, industrial, and
stabilized with one or both of these stabilizers found similar pulmonary tumors regardless of the
presence of stabilizers. In this study, female mice (n = 50) had elevated, but again not
statistically-significant, increases in pulmonary tumors. A later gavage study by NTP (1988),
which used TCE stabilized with diisopropylamine, observed no pulmonary tumors, but chemical
toxicity and early mortality rendered this study inadequate for determining carcinogenicity. The
final NTP study (1990) in male and female F344 rats and B6C3F1 mice, using epichlorohydrin-
free TCE, again showed early mortality in male rats. Similar to the other gavage studies, a non-
statistically significant elevation in (malignant) pulmonary tumors was observed in mice, in this
case in both sexes. These animal studies show that while there is a limited increase in lung
tumors following gavage exposure to TCE in mice, the only statistically significant increase in
lung tumors occurs following inhalation exposure in mice.
4.6.3 Role of Metabolism in Pulmonary Toxicity
TCE oxidative metabolism has been demonstrated to play a main role in TCE pulmonary
toxicity in mice. However, data is not available on the role of specific oxidative metabolites in
the lung. The Clara cell is thought to be the cell type responsible for much of the P450
metabolism in the lung. Therefore, damage to this cell type would be expected to also affect
metabolism. More direct measures of P450 and isozyme-specific depression following TCE
exposure have been reported following intraperitoneal administration in mice. Forkert et al.
(1985) reported significant reduction in microsomal aryl hydrocarbon hydroxylase activity as
well as P450 content between 1 and 24 hr after exposure (2,000-3,000mg/kg i.p. TCE).
Maximal depression occurred between 2 and 12 hr, with aryl hydrocarbon hydroxylase activity
(a function of CYP450) less than 50% of controls and P450 content less than 20% of controls.
While there was a trend towards recovery from 12 to 24 hr, depression was still significant at 24
hr. Forkert et al. (2005) reported decreases in immunoreactive CYP2E1, CYP2F2, and CYP2B1
in the 4 h after TCE treatment with 750 mg/kg intraperitoneal injection in mice. The amount and
time of maximal reduction was isozyme dependent (CYP2E1: 30% of controls at 2 hr; CYP2F2:
6/22/2009
724
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
abolished at 30 min; CYP2B1: 43% of controls at 4 hr). Catalytic markers for CYP2E1,
CYP2F2, and CYP2B enzymes showed rapid onset (15 min or less after TCE administration) of
decreased activity, and continued depression through 4 hr. Decrease in CYP2E1 and CYP2F2
activity (measured by PNP hydroxylase activity) was greater than that of CYP2B (measured by
pentoxyresorufin O-dealkylase activity). Forkert et al. (2006) reported similar results in which 4
hr after treatment, immunodetectable CYP2E1 protein was virtually abolished at doses
250-1,000 mg/kg and immunodetectable CYP2F2 protein, while still detectable, was reduced.
PNP hydroxylase activity was also reduced 4 hr after treatment to 37% of controls at the lowest
dose tested of 50 mg/kg, with further decreases to around 8% of control levels at doses of 500
mg/kg and higher. These results correlate with previously described increases in Clara cell
cytotoxicity, as well as dichloroacetyl lysine (DAL) protein adduct formation. DAL adducts
were observed in the bronchiolar epithelium of CD-I mice and most prominent in the cellular
apices of Clara cells (Forkert et al., 2006). This study also examined the effect of TCE in vitro
exposure on the formation of chloral hydrate in lung microsomes from male CD-I mice and
CYP2E1 knock-out mice. The rates of CH formation were the same for lysosomes from both
CD-I and CYP2E1 knockout mice from 0.25 mM to 0.75 mM, but the CH formation peaked
earlier for in the wild-type lysosomes (0.75mM) as compared to CYP2El-null lysosomes (1
mM).
The strongest evidence for the necessary role of TCE oxidation is that pre-treatment of
mice with diallyl sulfone (DASO2), an inhibitor of CYP2E1 and CYP2F2, protected against
TCE-induced pulmonary toxicity. In particular, following an intraperitoneal TCE dose of 750
mg/kg, Clara cells and the bronchiolar epithelium in mice pre-treated with the CYP2E1/CYP2F2
inhibitor appeared normal. In naive mice given the same dose, the epithelium was attenuated
due to exfoliation and there was clear morphological distortion of Clara cells (Forkert et al.,
2005). In addition, the greater susceptibility of mouse lungs relative to rat lungs is consistent
with their larger capacity to oxidize TCE, as measured in vitro in lung microsomal preparations
(Green et al., 1997). Analysis by immunolocalization also found considerably higher levels of
CYP2E1 in the mouse lung, heavily localized in Clara cells, as compared to rat lungs, with no
detectable CYP2E1 in human lung samples (Green et al., 1997). In addition, both Green et al.
(1997) and Forkert et al. (2006) report substantially lower metabolism of TCE in human lung
microsomal preparations than either rats or mice. It is clear that CYP2E1 is not the only P450
enzyme involved in pulmonary metabolism, as lung microsomes from CYP2El-null mice
showed greater or similar rates of CH formation compared to those from wild-type mice. Recent
studies have suggested a role for CYP2F2 in TCE oxidative metabolism, although more work is
needed to make definitive conclusions. In addition, there may be substantial variability in human
lung oxidative metabolism, as Forkert et al. (2006) reported that in microsomal samples from
6/22/2009
725
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
eight individuals, five exhibited no detectable TCE oxidation (< 0.05 pmol/mg protein/20
minutes), while others exhibited levels well above the limit of detection (0.4-0.6 pmol/mg
protein/minute).
In terms of direct pulmonary effects of TCE metabolites, Odum et al. (1992) reported that
mice exposed to 100 ppm via inhalation of chloral for 6 hours resulted in bronchiolar lesions
similar to those seen with TCE, although with a severity equivalent to 1,000 ppm TCE
exposures. In addition, some alveolar necrosis, alveolar oedema, and desquamation of the
epithelium were evident. In the same study, TCOH (100 and 500 ppm) also produced Clara cell
damage, but with lower incidence than TCE, and without alveolar lesions, while TCA treatment
produced no observable pulmonary effects. Therefore, it has been proposed that chloral is the
active metabolite responsible for TCE pulmonary toxicity, and the localization of damage to
Clara cells (rather than to other cell types, as seen with direct exposure to chloral) is due to the
localization of oxidative metabolism in that cell type (Odum et al., 1992; Green et al., 1997;
Green, 2000). However, the recent identification by Forkert et al. (2006) of DAL adducts, also
localized with Clara cell, suggests that TCE oxidation to dichloroacetyl chloride, which is not
believed to be derived from chloral, may also contribute to adverse health effects.
Due to the histological similarities between TCE- and chloral-induced pulmonary
toxicity, consistent with chloral being the active moiety, it has been proposed that the limited or
absent capacity for reduction of chloral (rapidly converted to chloral hydrate (CH) in the
presence of water) to TCOH and glucuronidation of TCOH to TCOG in mouse lungs leads to
"accumulation" of chloral in Clara cells. However, the lack of TCOH glucuronidation capacity
of Clara cells reported by Odum et al. (1992), while possibly an important determinant of TCOH
concentrations, should have no bearing on CH concentrations, which depend on the production
and clearance of CH only. While isolated mouse Clara cells form smaller amounts of TCOH
relative to CH (Odum et al., 1992), the cell-type distribution of the enzymes metabolizing CH is
not clear. Indeed, cytosolic fractions of mouse, rat and human whole lungs show significant
activity for CH conversion to TCOH (Green et al., 1997). In particular, in mouse lung
subcellular fractions, 1 micromole of TCE in a 1.3 mL reactivial was converted to CH at a rate of
1 nmol/min/mg microsomal protein, while 10 nmol CH in a 1.3 mL reactivial was converted to
TCOH at a rate of 0.24 nmol/min/mg cytosolic protein (Green et al., 1997). How this 4-fold
difference in activity would translate in vivo is uncertain given the 100-fold difference in
substrate concentrations, lack of information as to the concentration-dependence of activity, and
uncertain differences between cytosolic and microsomal protein content in the lung. It is unclear
whether local pulmonary metabolism of chloral is the primary clearance process in vivo, as in the
presence of water, chloral rapidly converts to chloral hydrate, which is soluble in water and
hence can rapidly diffuse to surrounding tissue and to the blood, which also has the capacity to
6/22/2009
726
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
metabolize chloral hydrate (Lipscomb et al., 1996). Nonetheless, experiments with isolated
perfused lungs of rats and guinea pigs found rapid appearance of TCOH in blood following TCE
inhalation exposure, with no detectable chloral hydrate or TCOG (Dalbey and Bingham, 1978).
Therefore, it appears likely that chloral in the lung either is rapidly metabolized to TCOH, which
then diffuses to blood, or diffuses to blood as CH and is rapidly metabolized to TCOH by
erythrocytes (Lipscomb et al., 1996).
This hypothesis is further supported by in vivo data. No in vivo data in rats on CH after
TCE administration were located, and Fisher et al. (1998) reported CH in blood of human
volunteers exposed to TCE via inhalation were below detection limits. In mice, however, after
both inhalation and oral gavage exposure to TCE, CH has been reported in whole lung tissue at
concentrations similar to or somewhat greater than that in blood (Abbas and Fisher, 1997;
Greenberg et al., 1999). A peak concentration (1.3 jug/g) of pulmonary CH was reported after
inhalation exposure to 600 ppm - at or above exposures where Clara cell toxicity was reported in
acute studies (Odum et al., 1992; Green et al., 1997). However, this was 5-fold less than the
reported pulmonary CH concentration (6.65 jug/g) after gavage exposures of 1,200 mg/kg.
Specifically, a 600-ppm exposure or 450 ppm exposure reported in the Maltoni et al. and Fukuda
et al. studies results in a greater incidence in lung tumors than the 1,000-1,200 mg/kg/day
exposures in the NTP (1990) and NCI (1976) bioassays. However the peak CH levels measured
in whole lung tissues after inhalation exposure to TCE at 600 ppm were reported to be about 5-
fold lower than that at 1,200 mg/kg by gavage, therefore showing the opposite pattern
(Greenberg et al., 1999; Abbas and Fisher, 1997). No studies of Clara cell toxicity after gavage
exposures were located, but several studies in mice administered TCE via intraperitoneal
injection did show Clara cell toxicity at around a dose of 750mg/kg (Forkert et al., 2006) or
above (e.g., Forkert and Forkert, 1994; Forkert and Birch, 1989). However, as noted previously,
i.p. exposures are subject to an inflammatory response, confounding direct comparisons of dose
via other routes of administration.
Although, whole lung CH concentrations may not precisely reflect the concentrations
within specific cell types, as discussed above, the water solubility of CH suggests rapid
equilibrium between cell types and between tissues and blood. Both Abbas and Fisher (1997)
and Greenberg et al. (1999) were able to fit CH blood and lung levels using a PBPK model that
did not include pulmonary metabolism, suggesting that lung CH levels may be derived largely by
systemic delivery, i.e., from CH formed in the liver. However, a more detailed PBPK model-
based analysis of this hypothesis has not been performed, as CH is not included in the PBPK
model developed by Hack et al. (2006) that was updated in Section 3.5.
Two studies have reported formation of reactive metabolites in pulmonary tissues as
assessed by macromolecular binding after TCE intraperitoneal administration. Forkert and Birch
6/22/2009
727
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(1989) reported temporal correlations between the severity of Clara cell necrosis with increased
levels of covalent binding macromolecules in the lung of TCE or metabolites with a single 2,000
mg/kg dose of [14C] TCE. The amount of bound TCE or metabolites per gram of lung tissue,
DNA, or protein peaked at 4 hr and decreased progressively at 8, 12, and 24 hr. The fraction of
radioactivity in lung tissue macromolecules that was covalently bound reached a plateau of about
20% from 4-24 hr, suggesting that clearance of total and covalently bound TCE or metabolites
was similar. The amount of covalent binding in the liver was 3 to 10-fold higher than in the
lung, although hepatic cytotoxicity was not apparent. This tissue difference could either be due
to greater localization of metabolism in the lung, so that concentrations reactive metabolites in
individual Clara cells are greater than both the lung as a whole and hepatocytes, or because of
greater sensitivity of Clara cells as compared to hepatocytes to reactive metabolites. More
recently, Forkert et al. (2006) examined DAL adducts resulting from metabolism of TCE to
dichloroacetyl chloride as an in vivo marker of production of reactive metabolites. Following
intraperitoneal administration of 500-1,000 mg/kg TCE in CD-I mice, they found localization of
DAL adducts believed to be from oxidative metabolism within Clara cell apices, with dose-
dependent increase in labeling with a polyclonal anti-DAL antibody that correlated with
increased Clara cell damage. Dose-dependent DAL adducts were also found in alveolar type II
cells, although no morphologic changes in those cells were observed Both Clara cell damage (as
discussed above) and DAL labeling were abolished in mice pre-treated with DAS02, an inhibitor
of CYP2E1 and CYP2F2. However, Clara cell damage in treated CYP2El-null mice was more
severe than in CD-I mice. Although DAL labeling was less pronounced in CYP2El-null mice
as compared to CD-I mice, this was due in part to the greater histopathologic damage leading to
attenuation of the epithelium and loss of Clara cells in the null mice. In addition, protein
immunoblotting with anti-DAL, anti-CYP2El and anti-CYP2F2 antibodies suggested that a
reactive TCE metabolite including dichloroacetyl chloride was formed that is capable of binding
to CYP2E1 and CYP2F2 and changing their protein structures. Follow-up studies are needed in
the lung and other target tissues to determine the potential role of the DAL adducts in TCE-
induced toxicity.
Finally, although Green (2000) and others have attributed species differences in
pulmonary toxicity to differences in the capacity for oxidative metabolism in the lung, it should
be noted that the concentration of the active metabolite is determined by both its production and
clearance (Clewell et al., 2000). Therefore, while the maximal pulmonary capacity to produce
oxidative metabolites is clearly greater in the mouse than in rats or humans, there is little
quantitative information as to species differences in clearance, whether by local chemical
transformation/metabolism or by diffusion to blood and subsequent systemic clearance. In
addition, existing in vitro data on pulmonary metabolism are at millimolar TCE concentrations
6/22/2009
728
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
where metabolism is likely to be approaching saturation, so the relative species differences at
lower doses has not been characterized. Studies with recombinant P450 enzymes examined
species differences in the catalytic efficiencies of CYP2E1, CYP2F, and CYP2B1, but the
relative contributions of each isoform to pulmonary oxidation of TCE in vivo remains unknown
(Forkert et al., 2005). Furthermore, systemic delivery of oxidative metabolites to the lung may
contribute, as evidenced by respiratory toxicity reported with i.p. administration. Therefore,
while the differences between mice and rats in metabolic capacity are correlated with their
pulmonary sensitivity, it is not clear that differences in capacity alone are accurate quantitative
predictors of toxic potency. Thus, while it is likely that the human lung is exposed to lower
concentrations of oxidative metabolites, quantitative estimates for differential sensitivity made
with currently available data and dosimetry models are highly uncertain.
In summary, it appears likely that pulmonary toxicity is dependent on in situ oxidative
metabolism, however, the active agent has not been confidently identified. The similarities in
histopathologic changes in Clara cells between TCE and chloral inhalation exposure, combined
with the wider range of cell types affected by direct chloral administration relative to TCE, led
some to hypothesize that chloral is the toxic moiety in both cases, but with that generated in situ
from TCE in Clara cells "accumulating" in those cells (Green, 2000). However, chemical and
toxicokinetic data suggest that such "accumulation" is unlikely for several reasons. These
include the rapid conversion of chloral to chloral hydrate in the presence of water, the water
solubility of CH leading to rapid diffusion to other cell types and blood, the likely rapid
metabolism of chloral hydrate to TCOH either in pulmonary tissue or in blood erythrocytes, and
in vivo data showing lack of correlation across routes of exposure between whole-lung CH
concentrations and pulmonary carcinogenicity and toxicity. However, additional possibilities for
the active moiety exist, such as dichloroacetyl chloride, which is derived through a TCE
oxidation pathway independent of chloral and which appears to result in adducts with lysine
localized in Clara cells.
4.6.4 Mode of Action for Pulmonary Carcinogenicity
A number of effects have been hypothesized to be key events in the pulmonary
carcinogenicity of TCE, including cytotoxicity leading to increased cell proliferation, formation
of DAL protein adducts, and mutagenicity. As stated previously, the target cell for pulmonary
adenocarcinoma formation has not been established. Much of the hazard and MOA information
has focused on Clara cell effects from TCE which is a target in both susceptible and non-
susceptible rodent species for lung tumors. However, the role of Clara cell susceptibility to
TCE-induced lung toxicity or to other potential targets such as lung stem cells that are activated
6/22/2009
729
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
to repopulate both Clara and type II alveolar cells after injury, has not been determined for
pulmonary carcinogenicity. While all of the events described above may be plausibly involved
in the MOA for TCE pulmonary carcinogenicity, none have been directly shown to be necessary
for carcinogenesis.
4.6.4.1 Mutagenicity via Oxidative Metabolism
The hypothesis is that TCE acts by a mutagenic MOA in TCE- induced lung tumors.
According to this hypothesis, the key events leading to TCE-induced lung tumor formation
constitute the following: the oxidative metabolism of TCE producing chloral/chloral hydrate
delivered to pulmonary tissues, causes direct alterations to DNA (e.g., mutation, DNA damage,
and/or micronuclei induction). Mutagenicity is a well-established cause of carcinogenicity.
4.6.4.1.1 Experimental Support for the Hypothesized Mode of Action
Pulmonary toxicity has been proposed to be dependent on in situ oxidative metabolism,
however the active agent has not been confidently identified. The similarities in histopathologic
changes in Clara cells between TCE and chloral inhalation exposure, combined with the wider
range of cell types affected by direct chloral administration relative to TCE, led some to
hypothesize that chloral is the toxic moiety. Chloral that is formed from the metabolism of TCE
is quickly converted to chloral hydrate (CH) upon hydration under physiological conditions. As
discussed in Section 4.1.4, CH clearly induces aneuploidy in multiple test systems, including
bacterial and fungal assays in vitro (Kafer, 1986; Kappas, 1989; Crebelli et al., 1991),
mammalian cells in vitro (Vagnarelli et al., 1990; Sbrana et al., 1993), and mammalian germ-line
cells in vivo (Russo et al., 1984; Miller and Adler, 1992). Conflicting results were observed in in
vitro and in vivo mammalian studies of micronuclei formation (Degrassi and Tanzarella, 1988;
Nesslany and Marzin, 1999; Russo and Levis, 1992a, b; Giller et al., 1995; Beland, 1999), with
positive results in germ-line cells (Nutley et al., 1996; Allen et al., 1994). In addition, it is
mutagenic in the Ames bacterial mutation assay for some strains (Haworth et al., 1983; Ni et al.,
1994; Beland, 1999; Giller et al., 1995). Structurally related chlorinated aldehydes 2-
chloroacetyaldehyde and 2,2,-dichloroacetaldehyde are both alkylating agents, are both positive
in a genotoxic assay (Bignami et al., 1980), and both interact covalently with cellular
macromolecules (Guengerich et al., 1979).
As discussed in the section describing the experimental support for the mutagenic MOA
for liver carcinogenesis (4.4.7.1), it has been argued that CH mutagenicity is unlikely to be the
cause of TCE carcinogenicity because the concentrations required to elicit these responses are
several orders of magnitude higher that achieved in vivo (Moore and Harrington-Brock, 2000).
Similar to the case of the liver, it is not clear how much of a correspondence is to be expected
6/22/2009
730
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
from concentrations in genotoxicity assays in vitro and concentrations in vivo, as reported in vivo
CH concentrations are in whole lung homogenate while in vitro concentrations are in culture
media. None of the available in vivo genotoxicity assays used the inhalation route that elicited
the greatest lung tumor response under chronic exposure conditions, so direct in vivo
comparisons are not possible. Finally, as discussed in Section 4.4.7.1, the use of ip
administration in many other in vivo genotoxicity assays complicates the comparison with
carcinogenicity data.
As discussed above (Section 4.6.3), chemical and toxicokinetic data are not supportive of
CH being the active agent of TCE-induced pulmonary toxicity, and directly contradict the
hypothesis of chloral "accumulation." Nonetheless, CH has been measured in the mouse lung
following inhalation and gavage exposures to TCE (Abbas and Fisher, 1997; Greenberg et al.,
1999), possibly the result of both in situ production and systemic delivery. Therefore, in
principle, CH could cause direct alterations in DNA in pulmonary tissue. However, as discussed
above, the relative amounts of CH measured in whole lung tissue from inhalation and oral
exposures do not appear to correlate with sensitivity to TCE lung tumor induction across
exposure routes. While these data cannot rule out a role for mutagenicity mediated by CH due to
various uncertainties, such as whether whole lung CH concentrations accurately reflect cell-type
specific concentrations and possible confounding due to strain differences between inhalation
and oral chronic bioassays, they do not provide support for this MO A.
Additional possibilities for the active moiety exist, such as dichloroacetyl chloride, which
is derived through a TCE oxidation pathway independent of chloral and which appears to result
in adducts with lysine localized in Clara cells (Forkert et al., 2006). DCA, which has some
genotoxic activity, is, also, presumed to be formed through this pathway (see section 3.3).
Currently, however, there are insufficient data to support a role for these oxidative metabolites in
a mutagenic MOA.
4.6.4.2 Cytotoxicity leading to increased cell proliferation
The hypothesis is that TCE acts by a cytotoxicity MOA in TCE-induced pulmonary
carcinogenesis. According to this hypothesis, the key events leading to TCE-induced lung tumor
formation constitute the following: TCE oxidative metabolism in situ leads to currently unknown
reactive metabolites that cause cytotoxicity, leading to compensatory cellular proliferation and
subsequently increased mutations and clonal expansion of initiated cells.
4.6.4.2.1 Experimental Support for the Hypothesized Mode of Action
Evidence for the hypothesized MOA consists primarily of (i) the demonstration of acute
cytotoxicity and transient cell proliferation following TCE exposure in laboratory mouse studies;
(ii) toxicokinetic data supporting oxidative metabolism being necessary for TCE pulmonary
6/22/2009
731
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
toxicity; (iii) the association of lower pulmonary oxidative metabolism and lower potency for
TCE-induced cytotoxicity with the lack of observed pulmonary carcinogenicity in laboratory
rats. However, there is a lack of experimental support linking TCE acute pulmonary cytotoxicity
to sustained cellular proliferation of chronic exposures or clonal expansion of initiated cells.
As discussed above, a number of acute studies have shown that TCE is particularly
cytotoxic to Clara cells in mice, which has been suggested to be involved in the development of
mouse lung tumors (Buckpitt et al., 1995; Forkert and Forkert, 1994, Kim et al., 2005). In
addition, studies examining cell labeling by either BrdU (Green et al., 1997) or 3H-thymidine
incorporation (Villaschi et al., 1991) suggest increased cellular proliferation in mouse Clara cells
following acute inhalation exposures to TCE. Moreover, in short-term studies, Clara cells appear
to become resistant to cytotoxicity with repeated exposure, but regain their susceptibility after
two days without exposure. This observation led to the hypothesis that the 5 day/wk inhalation
dosing regime (Fukuda et al., 1983; Maltoni et al., 1986, 1988; Henschler et al., 1980) in the
chronic mouse studies leads to periodic cytotoxicity in the mouse lung at the beginning of each
week followed by cellular regeneration, and that the increased rate of cell division leads to
increased incidence of tumors by increasing the overall mutation rate and by increasing the
division rate of already initiated cells (Green, 2000). However, longer-term studies to test this
hypothesis have not been carried out.
As discussed above (Section 4.6.3), there is substantial evidence that pulmonary
oxidative metabolism is necessary for TCE-induced pulmonary toxicity, although the active
moiety remains unknown. In addition, the lower capacity for pulmonary oxidative metabolism
in rats as compared to mice is consistent with studies in rats not reporting pulmonary cytoxicity
until exposures higher than those in the bioassays, and the lack of reported pulmonary
carcinogenicity in rats at similar doses to mice. However, rats also have a lower background rate
of lung tumors (Green, 2000), and so would be less sensitive to carcinogenic effects in that tissue
to the extent that relative risks is the important metric across species. In addition, this MOA
hypothesis requires a number of additional key assumptions for which there are currently no
direct evidence. First, the cycle of cytotoxicity, repair, resistance to toxicity, and loss of
resistance after exposure interruption, has not been documented and under the proposed MOA
should continue under chronic exposure conditions. This cycle has thus far only been observed
in short term (up to 13-day) studies. In addition, although Clara cells have been identified as the
target of toxicity whether they or endogenous stem cells in the lung are the cells responsible for
mouse lung tumors has not been established. There is currently no data as to the cell type of
origin for TCE-induced lung tumors.
6/22/2009
732
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.6.4.3 Additional Hypothesized Modes of Action with Limited Evidence or Inadequate
Experimental Support
4.6.4.3.1 Role of Formation of DAL Protein Adducts
As discussed above, Forkert et al. (2006) recently observed dose-dependent formation of
DAL protein adducts in the Clara cells of mice exposed to TCE via intraperitoneal injection.
While adducts were highly localized in Clara cells, they were also found in alveolar type II cells,
though these cells did not show signs of cytotoxicity in this particular experimental paradigm. In
terms of the MOA for TCE-induced pulmonary carcinogenicity, these adducts may either be
causally important in and of themselves, or they may be markers of a different causal effect. For
instance, it is possible that these adducts are a cause for the observed Clara cell toxicity, and
Forkert et al. (2006) suggested that the lack of toxicity in alveolar type II cells may indicate that
"there may be a threshold in adduct formation and hence bioactivation at which toxicity is
manifested." In this case, they are an additional precursor event in the same causal pathway
proposed above. Alternatively, these adducts may be indicative of effects related to
carcinogenesis but unrelated to cytotoxicity. In this case, the Clara cell need not be the cell type
of origin for mouse lung tumors.
Because of their recent discovery, there is little additional data supporting, refuting, or
clarifying the potential role for DAL protein adducts in the MOA for TCE-induced pulmonary
carcinogenesis. For instance, the presence and localization of such adducts in rats has not been
investigated, and could indicate the extent to which the level of adduct formation is correlated
with existing data on species differences in metabolism, cytotoxicity, and carcinogenicity. In
addition, the formation of these adducts has only been investigated in a single dose study using
ip injection. As stated above, i.p. injection may involve the initiation of a systemic inflammatory
response that can activate lung macrophages or affect Clara cells. Experiments with repeated
exposures over chronic durations and by inhalation or oral of administration would be highly
informative. Finally, the biological effects of these adducts, whether cytotoxicity or something
else, have not been investigated.
4.6.4.4 Conclusions about the hypothesized modes of action
1. Is the hypothesized mode of action sufficiently supported in the test animals?
Mutagenicity: Chloral hydrate is clearly genotoxic, as there are substantial data from multiple in
vitro and in vivo assays supporting its ability induce aneuploidy, with more limited data as to
other genotoxic effects, such as point mutations. Chloral hydrate is also clearly present in
pulmonary tissues of mice following TCE exposures similar to those inducing lung tumors in
chronic bioassays. However, chemical and toxicokinetic data are not supportive of CH being the
6/22/2009
733
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
predominant metabolite for TCE carcinogenicity. Such data include the water solubility of CH
leading to rapid diffusion to other cell types and blood, its likely rapid metabolism to TCOH
either in pulmonary tissue or in blood erythrocytes, and in vivo data showing lack of correlation
across routes of exposure between whole lung CH concentrations and pulmonary
carcinogenicity. Therefore, while a role for mutagenicity via CH in the MOA of TCE-induced
lung tumors cannot be ruled about, available evidence is inadequate to support the conclusion
that direct alterations in DNA caused by CH produced in or delivered to the lung after TCE
exposure constitute a MOA for TCE-induced lung tumors.
Cytotoxicity: The MOA hypothesis for TCE-induced lung tumors involving cytotoxicity is
supported by relatively consistent and specific evidence for cytotoxicity at tumorigenic doses in
mice. However, the majority of cytotoxicity-related key events have been investigated in studies
less than 13 days, and none has been shown to be causally related to TCE-induced lung tumors.
In addition, the cell type (or types) of origin for the observed lung tumors in mice has not been
determined, so the contribution to carcinogenicity of Clara cell toxicity and subsequent
regenerative cell division is not known. Similarly, the relative contribution from recently
discovered dichloroacetyl-lysine protein adducts to the tumor response has not been investigated
and has currently only been studied in i.p. exposure paradigms of short duration. In summary,
while there are no data directly challenging the hypothesized MOA described above, the existing
support for their playing a causal role in TCE-induced lung tumors is largely associative, and
based on acute or short term studies. Therefore, there are inadequate data to support a cytotoxic
MOA based on the TCE-induced cytotoxicity in Clara cells in the lungs of test animals.
Additional hypothesis: Inadequate data are available to develop a MOA hypothesis based on
recently discovered DAL adducts induced by TCE inhalation and ip exposures. It will therefore
not be considered further in the conclusions below.
Overall, therefore, the MOA for TCE-induced lung tumors is considered unknown at this time.
2. Is the hypothesized mode of action relevant to humans?
Mutagenicity: The evidence discussed above demonstrates that CH is mutagenic in microbial as
well as test animal species. There is therefore the presumption that they would be mutagenic in
humans. Therefore, this MOA is considered relevant to humans.
Cytotoxicity: No data from human studies are available on the cytotoxicity of TCE and its
metabolites in the lung, and no causal link between cytotoxicity and pulmonary carcinogenicity
has been demonstrated in animal or human studies. Nonetheless, in terms of human relevance,
no data suggest that the proposed key events are not biologically plausible in humans, therefore
qualitatively, TCE-induced lung tumors are considered relevant to humans. Information about
6/22/2009
734
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
the relative pharmacodynamic sensitivity between rodents and humans is absent, but information
on pharmacokinetic differences in lung oxidative metabolism does exist and will be considered
in dose-response assessment when extrapolating between species (see Section 5.X.X).
3. Which populations or lifestages can be particularly susceptible to the hypothesized mode of
action?
Mutagenicity: The mutagenic MOA is considered relevant to all populations and lifestages.
According to EPA's Cancer Guidelines (U.S. EPA, 2005a) and Supplemental Guidance (U.S.
EPA, 2005b), there may be increased susceptibility to early-life exposures for carcinogens with a
mutagenic mode of action. However, because the weight of evidence is inadequate to support a
mutagenic MOA for TCE pulmonary carcinogenicity, and in the absence of chemical-specific
data to evaluate differences in susceptibility, the age-dependent adjustment factors (ADAFs)
should not be applied, in accordance with the Supplemental Guidance.
Cytotoxicity: No information based is available as to which populations or lifestages may be
particularly susceptible to TCE-induced lung tumors. However, pharmacokinetic differences in
lung oxidative metabolism among humans do exist, and because of the association between lung
oxidative metabolism and toxicity, will be considered in dose-response assessment when
extrapolating within species.
4.6.5 Summary and Conclusions
The studies described here show pulmonary toxicity found mainly in Clara cells in mice
(Green et al., 1997; Villaschi et al., 1991; Odum et al., 1992; Forkert et al., 1985; Forkert and
Birch, 1989) and rats (Kurasawa, 1988). The most convincing albeit limited data regarding this
type of toxicity was demonstrated predominantly in mice exposed via inhalation, although some
toxicity was shown in intraperitoneal injection studies. Increased vacuolation of Clara cells was
often seen within the first 24h of exposure, depending on dose, but with cellular repair occurring
within days or weeks of exposure. Continued exposure led to resistance to TCE-induced Clara
cell toxicity, but damage recurred if exposure was stopped after 5 days and then resumed after 2
days without exposure. However, Clara cell toxicity has only been observed in acute and short-
term studies, and it is unclear whether they persist with sub-chronic or chronic exposure,
particularly in mice, which are the more sensitive species. With respect to pulmonary
carcinogenicity, statistically-significantly increased incidence of lung tumors from chronic
inhalation exposures to TCE was observed female ICR mice (Fukuda et al., 1983), male Swiss
mice, and female B6C3F1 mice (Maltoni et al., 1986), though not in other sex/strain
combinations, nor in rats (Henschler et al., 1980; Maltoni et al., 1986). However, lung toxicity
6/22/2009
735
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
and Clara cell effects have also been observed in rats. Overall, the limited carcinogenesis studies
described above are consistent with TCE causing mild increases in pulmonary tumor incidence
in mice, but not in other species tested such as rats and hamsters.
The epidemiologic studies are quite limited for examining the role of TCE in cancers of
the respiratory system, with no studies found on TCE exposure specifically examining toxicity of
the respiratory tract. The two studies found on organic solvent exposure which included TCE
suggested smoking as a primary factor for observed lung function decreases among exposed
workers. Animal studies have demonstrated toxicity in the respiratory tract, particularly damage
to the Clara cells (non-ciliated bronchial epithelial cells), as well as decreases in pulmonary
surfactant following both inhalation and intraperitoneal exposures, especially in mice. Dose-
related increases in vacuolation of Clara cells have been observed in mice and rats as early as 24
h post-exposure (Odum et al., 1992; Kurasawa, 1988; Forkert et al., 1985, 2006; Forkert and
Birch, 1989; Scott et al., 1988). Mice appear to be more sensitive to these changes, but both
species show a return to normal cellular morphology at four weeks post-exposure (Odum et al.,
1992). Studies in mice have also shown an adaptation or resistance to this damage after only
four to five days of repeated exposures (Odum et al., 1992; Green et al., 1997). The limited
epidemiological literature on lung and laryngeal cancer in TCE-exposed groups is inconclusive
due to study limitations (low power, null associations, confidence intervals on relative risks that
include 1.0). These studies can only rule out risks of a magnitude of 2.0 or greater for lung
cancer and relative risks greater than 3.0 or 4.0 for laryngeal cancer for exposures to studied
populations and thus may not detect a level of response consistent with other endpoints. Animal
studies demonstrated a statistically significant increase in pulmonary tumors in mice following
chronic inhalation exposure to TCE (Fukuda et al., 1983; Maltoni et al., 1988, 1986). These
results were not seen in other species tested (rats, hamsters; Maltoni et al., 1986, 1988; Fukuda et
al., 1983; Henschler et al., 1980). By gavage, elevated, but not statistically significant,
incidences of benign and/or malignant pulmonary tumors have been reported in B6C3F1 mice
(NCI, 1976; Henschler et al., 1984; NTP, 1990). No increased pulmonary tumor incidences have
been reported in rats exposed to TCE by gavage (NCI, 1976; NTP, 1988, 1990), although all the
studies suffered from early mortality in at least one sex of rat.
Although no epidemiologic studies on the role of metabolism of TCE in adverse
pulmonary health effects have been published, animal studies have demonstrated the importance
of the oxidative metabolism of TCE by CYP2E1 and/or CYP2F2 in pulmonary toxicity.
Exposure to diallyl sulfone (DASO2), an inhibitor of both enzymes protects against pulmonary
toxicity in mice following exposure to TCE (Forkert et al., 2005). The increased susceptibility in
mice correlates with the greater capacity to oxidize TCE based on increased levels of CYP2E1 in
mouse lungs relative to lungs of rats and humans (Green et al., 1997; Forkert et al., 2006), but it
6/22/2009
736
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
is not clear that these differences in capacity alone are accurate quantitative predictors of
sensitivity to toxicity. In addition, available evidence argues against the previously proposed
hypothesis (e.g., Green, 2000) that "accumulation" of chloral in Clara cells is responsible for
pulmonary toxicity, since chloral is first converted the water-soluble compounds chloral hydrate
and TCOH that can rapidly diffuse to surrounding tissue and blood. Furthermore, the
observation of DAL protein adducts, likely derived dichloroacetyl chloride and not from chloral,
that were localized in Clara cells suggests an alternative to chloral as the active moiety. While
chloral hydrate has shown substantial genotoxic activity, chemical and toxicokinetic data on CH
as well as the lack of correlation across routes of exposure between in vivo measurements of CH
in lung tissues and reported pulmonary carcinogenicity suggest that evidence is inadequate to
conclude that a mutagenic MOA mediated by CH is operative for TCE-induced lung tumors.
Another MOA for TCE-induced lung tumors has been plausibly hypothesized to involve
cytotoxicity leading to increased cell proliferation, but the available evidence is largely
associative and based on short-term studies, so a determination of whether this MOA is operative
cannot be made. The recently discovered formation of DAL protein adducts in pulmonary
tissues may also play a role in the MOA of TCE-induced lung tumors, but an adequately defined
hypothesis has yet to be developed. Therefore, the MOA for TCE-induced lung tumors is
currently considered unknown, and this endpoint is thus considered relevant to humans.
Moreover, none of the available data suggest that any of the currently hypothesized mechanisms
would be biologically precluded in humans.
6/22/2009
737
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.6.3 Animal Toxicity Studies of Trichloroethylene
Reference
Animals (Sex)
Exposure Route
Dose/Exp Cone
Exposed
Results
450ppm, 6h/day, 5 days with 2 day
Increased vacuolation and proliferation of
break then 5 more days; sacrificed
Clara cells caused by accumulation of
Green et al., 1997
CD-1 mice (F)
Inhalation
18h after 1, 5, 6 or 10 exposures
5/group
chloral.
2,000 mg/kg in corn oil (0.01 mL/g
Forkert and Forkert,
Intraperitoneal
bw); sacrificed 15, 30, 60 and 90
Increased fibrotic lesions, with early signs
1994
CD-1 mice (M)
Injection
days after single exposure
10/group
visible at 15d post exposure.
Increased vacuolation and proliferation of
30 min 500, 1,000, 2,000, 3,500
nonciliated bronchial cells. Injury was
BC3F1 mice
and 7,000 ppm; sacrificed 2 h,
maximal at 24hrs with some repair
Villaschi et al., 1991
(M)
single inhalation
24 h, 2 d, 5 d, 7 d post exposure
3/group
occurring between 24h and 48h.
6h/day; separate repeated study in
Dose-dependent increase in Clara cell
mice: 450 ppm for6h/day, 5d/week
vacuolation in mice after a single
for 2 wks; sacrificed 24h after
exposure, resolved after 5 day repeated
exposure; repeat study sacrificed
exposures but recurred following a 2-day
at 2d, 5d, 6d, 8d, 9d, 12d 13d;
break from exposure. Changes
CD-1 mice (F)
inhalation
mice: 20, 100, 200, 450, 1,000, or
2,000ppm
4/group
accompanied by decrease in CYP450
activity in mice. Exposure to chloral
Odum et al., 1992
Alpk APfSD
rats (F)
inhalation
6h/day; repeat study sacrificed at
2d, 5d, 6d, 8d, 9d, 12d 13d; rats:
500 or 1,000 ppm
4/group
alone demonstrated similar response as
TCE exposure in mice. No changes
were seen in rats.
TCE exposure resulted in highly selective
Ethanol-treated
damage to Clara cells that occurred
(130) and non-
500, 1,000, 2,000, 4,000, and
between 8 and 22h after the highest
Kurasawa, 1988
treated (110)
8,000ppm for2h; sacrificed 22 hrs
exposure with repair by 4 weeks post
(translation)
Wistar rats (M)
Inhalation
after exposure
10/group
exposure.
6/22/2009
738
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Animals (Sex)
Exposure Route
Dose/Exp Cone
Exposed
Results
Forkert et al., 2006
CD-1 mice (M);
Wild-type
(mixed 129/Sv
and C57BL)
and CYP2E1-
null mice (M)
Intraperitoneal
Injection
500, 750 and 1,000 mg/kg in corn
oil; for inhibition studies mice
pretreated with 100mg/kg diallyl
sulfone; for immunoblotting, 250,
500, 750 and 1,000 mg/kg; for PNP
hydroxylation, 50, 100, 250, 500,
750 and 1,000 mg/kg; sacrificed 4h
after exposure
4/group
TCE bioactivation by CYP2E1 and/or 2F2
correlated with bronchiolar cytotoxicity in
mice.
Forkert et al., 1985
CD-1 mice (M)
Intraperitoneal
injection
2,000, 2,500 or 3,000 mg/kg in
mineral oil; sacrificed 24h post-
exposure for dose response; time
course sacrificed 1,2, 12 and 24h
post-exposure.
10/group
Clara cell injury was increased following
exposure at all doses tested; time course
demonstrated a rapid and marked
reduction in pulmonary microsomal
cytochrome P450 content and aryl
hydrocarbon hydroxylase activity.
Alveolar Type II cells were also affected.
Forkert and Birch,
1989
CD-1 mice (M)
Intraperitoneal
injection
2,000 mg/kg in corn oil; sacrificed
1, 2, 4, 8, 12, and 24h post-
exposure
10/group
Necrotic changes seen in Clara cells as
soon as 1h post-exposure; increased
vacuolation was seen by 4h post-
exposure; covalent binding of TCE to
lung macromolecules peaked at 4h and
reached a plateau at 12 and 24h post
exposure.
Stewart et al., 1979;
Le Mesurier et al.,
1980
Wistar Rats (F)
Inhalation (whole
body chamber)
30 min, 48.5g/m3 (9,030 ppm);
sacrificed at 5 and 15 days post-
exposure
5/group
Decreased recovery of pulmonary
surfactant (dose-dependent).
Lewis, 1984
mice
inhalation (Pyrex
bell jars)
10,000ppm, 1 -4 hr daily for 5
consecutive days; sacrificed 24h
after last exposure
~28/group
Increased vacuolation and reduced
activity of pulmonary mixed function
oxidases.
6/22/2009
739
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Animals (Sex)
Exposure Route
Dose/Exp Cone
Exposed
Results
single injection of
Intraperitoneal
2,500-3,000mg/kg, sacrificed 24 h
Clara cells were damaged and exfoliated
Scott et al., 1988
CD-1 mice (M)
Injection
post-exposure
4/group
from the epithelium of the lung.
Male rats: 0, 125, 250, 500, 1,000,
2,000 mg/kg bw (corn oil); female
rats: 0, 62.5, 125, 250, 500 or
Increased pulmonary vasculitis in the
F344 rats
1,000 mg/kg bw (corn oil); Mice: 0,
high dose groups of male and female rats
(M,F)
375, 750, 1,500, 3,000, 6,000
(6/10 group as compared to 1/10 in
B6C3F1 mice
mg/kg bw (corn oil); dosed 5d/w for
controls). No pulmonary effects
NTP, 1990
(M,F)
Gavage
13 weeks
10/group
described in mice at this time point.
Sprague-
Dawley or
Long-Evans
rats; Hartley
Guinea pigs;
New Zealand
albino rabbits;
beagle dogs;
Rats (15);
squirrel
Guinea pigs
No histopathological changes observed,
monkeys (sex
(15); Rabbit
although rats were described to show a
Prendergast et al.,
not given for
730 ppm for 8h/d, 5d/w, 6 weeks or
(3); Dog (2);
nasal discharge in the 6 wk study. No
1967
any species)
Inhalation
35 ppm for 90 d constant
Monkey (3)
quantification was given.
Rales and dyspnea were observed in the
21, 16, 17
TCE high-dose group; two females with
Narotsky et al., 1995
F344 rats (F)
Gavage
0, 1,125, 1,500 mg/kg/d
per group
dyspnea subsequently died.
6/22/2009
740
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.6.4. Animal Carcinogenicity Studies of Trichloroethylene
Dose/Exp Cone
Reference Animals (Sex) Exposure Route (stabilizers, if any)
Inhalation, 7h/day, 5
Fukuda et al., ICR mice (F) days/week, 104 wk, 0, 50, 150, 450 ppm
1983 S-D rats (F) hold until 107 wk (Epichlorohydrin)
S-D rats (M, F)
Swiss mice (M,
F) Inhalation, 7h/day, 5
Maltoni et al., B6C3F1 mice days/week, 104 wk,
1986,1988 (M, F) hold until death 0, 100,300,600 ppm
Wistar rats (M, Inhalation, 6h/day, 5
F) days/week, 78
Syrian weeks, hold until
hamsters (M, 130 wk (mice and
Henschler et al., F) NMRI hamsters) or 156 wkO, 100, 500 ppm
Pulmonary tumor incidences
benign+malignant
malignant only
Mice: 6/49, 5/50, 13/50, 11/46; Mice: 1/49; 3/50; 8/50*; 7/46*
Rats: 0/50, 0/50, 1/47, 1/51 Rats: none
Swiss Mice: 25/180, 26/180,
Swiss Mice: 25/180, 26/180, 36/180, 47/180;
36/180, 47/180; B6C3F1 Mice: 20/180,
B6C3F1 Mice: 20/180, 15/180, 15/180, 19/180, 26/180;
19/180, 26/180; Rats: 0/280, 0/260, 0/260,
Rats: 0/280, 0/260, 0/260, 0/2600/260
1980
mice
(rats)
(Triethanolamine)
Rats: 1/57, 2/60, 1/60;
Hamsters: 0/60, 0/59, 0/60;
Mice: 10/59, 9/59, 3/58
Rats: 1/57, 2/60, 1/60;
Hamsters: 0/60, 0/59, 0/60;
Mice: 6/59, 6/59, 1/58
Henschler et al., Swiss mice (M, Gavage, 5/wk, 72
1984 F) wk hold 104 wk
Van Duuren et Swiss mice (M, Gavage, 1/wk, 89
al., 1979
F)
wk
2.4 g/kg bw (M), 1.8 g/kg
bw (F) all treatments; Male: 18/50, 17/50, 14/50,
(control, triethanolamine, 21/50, 15/50, 18/50;
industrial, epichlorohydrin, Female: 12/50, 20/50, 21/50,
1,2-epoxybutane, both) 17/50, 18/50, 18/50
0, 0.5 mg (unknown)
0/30 for all groups
Male: 8/50, 6/50, 7/50, 5/50,
7/50, 7/50;
Female: 5/50, 11/50, 8/50,
3/50, 7/50, 7/50
0/30 for all groups
6/22/2009
741
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference Animals (Sex) Exposure Route
Osborne- Gavage, 5/wk, 78
Mendel rats (M,wk, hold until 110
F) B6C3f1 wk (rats) or 90 wk
NCI, 1976 mice (M, F) (mice)
NTP, 1988
Dose/Exp Cone
(stabilizers, if any)
Rats: TWA: 0, 549, 1,097
mg/kg
Mice: TWA: M: 0, 1,169,
2,339mg/kg; F: 0, 869,
1,739 mg/kg
(Epoxybutane,
epichlorohydrin)
NTP, 1990
Maltoni et al.
1986
ACI, August,
Marshall,
Osborne-
Mendel rats
F344 rats (M,
F) B6C3F1
mice (M, F)
Gavage, 1/day, 5 0, 500, 1,000 mg/kg
days/week, 103wk (diisopropylamine)
Pulmonary tumor incidences
benign+malignant
malignant only
Gavage, 1/day, 5 Mice: 0, 1,000 mg/kg
days/week, 103 wk Rats: 0, 500, 1,000 mg/kg
Gavage, 1/day, 4-5
days/week, 56 wk;
S-D rats (M, F) hold until death 0, 50 or 250 mg/kg
Rats: M: 1/20, 0/50, 0/50; F:
0/20, 1/47, 0/50
Mice: M: 0/20, 5/50, 2/48; F:
1/20, 4/50, 7/47
ACI M: 1/50, 4/47, 0/46; F: 0/49,
2/47, 2/42
August M: 1/50, 1/50, 0/49; F:
1/50, 1/50, 0/50
Marshall M: 3/49, 2/50, 2/47; F:
3/49, 3/49, 1/46
Osborne-Mendel M: 2/50, 1/50,
1/50; F: 0/50, 3/50, 2/50
Mice: M: 7/49, 6/50; F: 1/48,
4/49
Rats: M: 4/50, 2/50, 3/49; F:
1/50, 1/49, 4/50
M: 0/30, 0/30, 0/30; F: 0/30,
0/30, 0/30
Rats: M: 0/20, 0/50, 0/50; F:
0/20, 1/47, 0/50
Mice: M: 0/20, 0/50, 1/48; F:
0/20, 2/50, 2/47
ACI M: 1/50, 2/47, 0/46; F:
0/49, 1/47, 2/42
August M: 0/50, 1/50, 0/49; F:
1/50, 0/50, 0/50
Marshall M: 3/49, 2/50, 2/47;
F: 3/49, 3/49, 1/46
Osborne-Mendel M: 1/50,
1/50, 0/50; F: 0/50, 3/50, 1/50
Mice: M: 3/49, 1/50; F: 1/48,
0/49
Rats: M: 3/50, 2/50, 3/49; F:
0/50, 0/49, 2/50
M: 0/30, 0/30, 0/30; F: 0/30,
0/30, 0/30
* = statistically-significantly different from controls by Fisher's exact test
(p<0.05)
6/22/2009
742
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Inhalation Studies:
Males, Benign+MalignantLung Tumors
Inhalation Studies:
Females, Benign+Malignant Lung Tumors
TCE exposure concentration (ppm)
TCE exposure concentration (ppm)
Gavage Studies:
Males, Benign+Malignant Lung Tumor
Gavage Studies:
Females, Benign+Malignant Lung Tumor
TCE dose (mg/kg-d)
TCE dose (mg/kg-d)
Figure 4.6.1. Pulmonary tumor incidences reported in chronic rodent bioassays.
6/22/2009
743
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
References
Abbas R, Fisher JW. 1997 A physiologically based pharmacokinetic model for trichloroethylene
and its metabolites, chloral hydrate, trichloroacetate, dichloroacetate, trichloroethanol, and
trichloroethanol glucuronide in B6C3F1 mice. Toxicology and Applied Pharmacology
147:15-30.
Agency for Toxic Substance and Disease Registry (ATSDR). 2004. Feasibility Investigation of
Worker Exposure to Trichloroethylene at the View-Master Factory in Beaverton, Oregon.
Prepared by the Oregon Department of Human Services. Report No. PB2005-102173.
Allen BC, Kavlock RJ, Kimmel CA, Faustman EM. 1994. Dose-response assessment for
developmental toxicity. Fundamental and Applied Toxicology 23: 487-495.
ATSDR(Agency for Toxic Substances and Disease Registry). 2006. Health Consultation.
Cancer and Birth Outcome Analysis, Endicott Area, Town of Union, Broome County, New
York. Health Statistics Review. Atlanta, GA: U.S. Department of Health and Human Services,
Public Health Service, Agency for Toxic Substances and Disease Registry. May 26, 2006.
ATSDR. (2008) Health statistics review follow-up: cancer and birth outcome analysis. Endicott
Area Investigation, Endicott Area, Town of Union, Broome County, New York. U.S.
Department of Health and Human Services, May 2008, Atlanta, GA.
Anttila A, Pukkala E, Sallmen M, Hernberg S, Hemminki K. 1995. Cancer incidence among
Finnish workers exposed to halogenated hydrocarbons. J Occup Environ Med 37:797-806.
Axelson O, Selden A, Andersson K, Hogstedt C. 1994. Updated and expanded Swedish cohort
study on trichloroethylene and cancer risk. J Occup Environ 36:556-562.
Beland FA. 1999. NTP technical report on the toxicity and metabolism studies of chloral
hydrate (CAS No. 302-17-0). Administered by gavage to F344/N rats and B6C3F1 mice. Toxic
Rep Ser. 59:1-66, A1-E7.
Bignami M, Conti G, Conti L, Crebelli R, Misuraca F, Puglia AM, Randazzo R, Sciandrello G,
Carere A. 1980. Mutagenicity of halogenated aliphatic hydrocarbons in Salmonella
typhimurium, Steptomyces coelicolor and Aspergillus nidulans. Chem Biol Interact. 30:9-23.
6/22/2009
744
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Blair A, Hartge P, Stewart PA, McAdams M, Lubin J. 1998. Mortality and cancer incidence of
aircraft maintenance workers exposed to trichloroethylene and other organic solvents and
chemicals: extended follow up. Occup Environ Med 55:161-171.
Boice JD, Marano DE, Fryzek JP, Sadler CJ, McLaughlin JK. 1999. Mortality among aircraft
manufacturing workers. Occup Environ Med 56:581-597.
Boice JD, MaranoDE, Cohen SS, Mumma MR, Blott WJ, Brill AB, Fryzek JP, Henderson BE,
McLaughlin JK. 2006. Mortality among Rocketdyne worker who tested rocket engines, 1948-
1999. J Occup Environ Med 48:1070-1092.
Buckpitt A, Chang A-M, Weir A, Van Winkle L, Duan X, Philpot R, Plopper C. 1995.
Relationship of cytochrome P450 activity to Clara cell cytotoxicity. Molecular Pharmacology
47:74-81.
Cakmak A, Ekici A, Ekici M, Arsian M, Iteginli A, Kurtipek E, Kara T. 2004. Respiratory
findings in gun factory workers exposed to solvents. Respirat Med 98:52-56.
Chang Y-M, Tai C-F, Yang S-C, Chan C-J, S Shin T-S, Lin RS, Liou S-H. 2003. A cohort
mortality study of workers exposed to chlorinated organic solvents in Taiwan. Ann Epidemiol
13:652-660.
Chang Y-M, Tai C-F, Yang S-C, Lin R, Sung F-C, Shin T-S, Liou S-H. 2005. Cancer Incidence
among Workers Potentially Exposed to Chlorinated Solvents in An Electronics Factory. J Occup
Health 47:171-180.
Clapp RW, Hoffman K. 2008. Cancer mortality in IBM Endicott IBM workers, 1969-2001; an
update on a NY production plant. Environ Health 7:13.
Clewell HJ, Gentry PR, Covington TR, Gearhart JM. 2000. Development of a physiologically
based pharmacokinetic model of trichloroethylene and its metabolites for use in risk assessment.
Environ Health Perspect 108(Suppl 2):283-305.
Costa G, Merletti F, Segnan N. 1989. A mortality cohort study in a north Italian aircraft factory.
Br JInd Med 46:738-743
6/22/2009
745
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Crebelli R, Conti G, Conti L, Carere A. 1991 .In vitro studies with nine known or suspected
spindle poisons: results in tests for chromosome malsegregation in Aspergillus nidulans.
Mutagenesis 6:131-136.
Dalbey W, Bingham E. 1978. Metabolism of trichloroethylene by the isolated perfused lung.
Toxicol Appl Pharmacol. 43:267-277.
Degrassi F, Tanzarella, C. 1988. Immunofluorescent staining of kinetochore in micronuclei: a
new assay for the detection of aneuploidy. MutatRes. 203:339-45.
Elizur A, Adair-Kirk TL, Kelley DG, Griffin GL, Demello DE, Senior RM. 2008. Tumor
necrosis factor-alpha from macrophages enhances LPS-induced clara cell expression of
keratinocyte-derived chemokine. Am J Respir Cell Mol Biol. 38:8-15 (Epub).
Environmental Health Strategies. 1997. Final Report. Correspondence from Paul A. Cammer,
Ph.D., Trichloroethylene Issues Group, to Cheryl Siegel Scott, US Environmental Protection
Agency. December 22, 1997.
Fisher JW, Mahle D, Abbas R. 1998. A human physiologically based pharmacokinetic model for
trichloroethylene and its metabolites, trichloroacetic acid and free trichloroethanol. Toxicology
and Applied Pharmacology 152:339-359.
Forkert PG, Birch DW. 1989. Pulmonary toxicity of trichloroethylene in mice. Drug Metabolism
and Disposition 17: 106-113.
Forkert PG, Forkert L. 1994. Trichloroethylene induces pulmonary fibrosis in mice. Can J
Physiol Pharmacol. 72: 205-210.
Forkert PG, Sylvestre PL, Poland JS. 1985. Lung injury induced by trichloroethylene.
Toxicology 35: 143-160.
Forkert PG, Baldwin RM, Millen B, Lash LH, Putt DA, Shultz MA, Collins KS. Pulmonary
bioactivation of trichloroethylene to chloral hydrate: relative contributions of CYP2E1, CYP2F
and CYP2B1. 2005. DrugMetab Dispos. 33: 1429-37.
6/22/2009
746
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Forkert PG, Millen B, Lash LH, Putt DA, Ghanayem BI. 2006. Pulmonary bronchiolar
cytotoxicity and formation of dichloroacetyl lysine protein adducts in mice treated with
trichloroethylene. The Journal of Pharmacology and Experimental Therapeutics. 316:520-529.
Fuduka K, Takemoto K, Tsuruta H. 1983. Inhalation carcinogenicity of trichloroethylene in mice
and rats. Industrial Health 21: 243-254.
Garabrant DH, Held J, Langholz B, Bernstein L. 1988. Mortality of Aircraft Manufacturing
Workers in Southern California. Am J Ind Med 13:683-693.
Garabrant DH, Held J, Langholz B, Bernstein L. 1988. Mortality of aircraft manufacturing
workers in Southern California. Am J Ind Med 13:683-693.
Giller S, Le Curieux F, Gauthier L, Erb F, Marzin D. 1995. Genotoxicity assay of chloral hydrate
and chloropicrine. MutatRes 348:147-152.
Green T. 2000. Pulmonary toxicity and carcinogenicity of trichloroethylene: species differences
and modes of action. Environ Health Perspect. 108(Suppl 2):261-264.
Green T, Mainwaring GW, Foster JR. 1997. Trichloroethylene-induced mouse lung tumors:
Studies of the mode of action and comparisons between species. Fundamental and Applied
Toxicology 37: 125-130.
Greenberg MS, Burton GA, Fisher JW. 1999. Physiologically based pharmacokinetic modeling
of inhaled trichloroethylene and its oxidative metabolites in B6C3F1 mice. Toxicology and
Applied Pharmacology 154:264-178.
Greenland S, Salvan A, Wegman DH, Hallock MF, Smith TH. 1994. A case-control study of
cancer mortality at the transformer-assembly facility. Int Arch Occup Environ Health 66:49-54.
Guengerich FP, Crawford WM, Watanabe PG. 1979. Activation of vinyl chloride to covalently
bound metabolites: roles of 2-chloroethylene and 2-chloroacetaldehyde. Biochemistry 18:
5177-5182.
6/22/2009
747
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Hack CE, Chiu WA, Jay Zhao Q, Clewell HJ.(2006). Bayesian population analysis of a
harmonized physiologically based pharmacokinetic model of trichloroethylene and its
metabolites. Regul Toxicol Pharmacol. 46:63-83.
Hansen J, Raaschou-Nielsen O, Christensen JM, Johansen I, McLaughlin JK, Lipworth L, Blot
WJ, Olsen JH. 2001. Cancer incidence among Danish workers exposed to trichloroethylene. J
Occup Environ Med 43:133-139.
Haworth S, Lawlor T, Mortelmans K, Speck W, Zeiger E. 1983. Salmonella mutagenicity test
results for 250 chemicals. Environ Mutagen 5 Suppl 1: 1-142.
Henschler D, Romen W, Elsasser HM, Reichert D, Eder E, Radwan Z. 1980. Carcinogenicity
study of trichloroethylene by longterm inhalation in three animal species. Arch Toxicol. 43:
237-248.
Henschler D, Elsasser H, Romen W, Eder E. 1984. Carcinogenicity study of trichloroethylene,
with and without epoxide stabilizers, in mice. J Cancer Res Clin Oncol 107:149-156.
Henschler D, Vamvakas S, Lammert M, Dekant W, Kraus B, Thomas B, Ulm K. 1995.
Increased incidence of renal cell tumors in a cohort of cardboard workers exposed to
trichloroethene. Arch Toxicol 69:291-299.
Isacson P, Bean JA, Splinter R, Olson DB, Kohler J. 1985. Drinking water and cancer incidence
in Iowa. III. Association of cancer with indices of contamination. Am J Epidemiol
121:856-869.
Kafer E. 1986. Tests which distinguish induced crossing-over and aneuploidy from secondary
segregation in Aspergillus treated with chloral hydrate or gamma-rays. Mutat Res 164:145-166.
Kappas A. 1989. On the mechanisms of induced aneuploidy in Aspergillus nidulans and
validation of tests for genomic mutations. Prog Clin Biol Res 318:377-384.
Kim CF, Jackson EL, Woolfenden AE, Lawrence S, Babar I, Vogel S, Crowley D, Bronson RT,
Jacks T. 2005. Identification of bronchioalveolar stem cells in normal lung and lung cancer.
Cell. 2005 Jun 17; 121 (6):823-35.
6/22/2009
748
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Kurasawa K. 1988. Selective damage of pulmonary nonciliated bronchiolar epithelial (Clara)
cells by trichloroethylene in rats. Jpn J Ind Health 30:121-133 (abstract only, Japanese).
Le Mesurier SM, Stewart BW, O'Connell PJ, Lykke AWJ. 1979. Pulmonary responses to
atmospheric pollutants. II. Effect of petrol vapour inhalation on secretion of pulmonary
surfactant. Pathology 11:81-87.
Le Mesurier SM, Lykke AWJ, Stewart BW. 1980. Reduced yield of pulmonary surfactant:
Patterns of response following administration of chemicals to rats by inhalation. Toxicology
Letters 5: 89-93.
Lipscomb JC, Mahle DA, Brashear WT, Garrett CM. 1996. A species comparison of chloral
hydrate metabolism in blood and liver. Biochemical and Biophysical Research Communications
227: 340-350.
Maltoni C, Lefemine G, Cotti G. 1986. Experimental research on trichloroethylene
carcinogenesis. In Archives of Research on Industrial Carcinogenesis, eds Maltoni C, Mehlman
MA. Volume V. Princeton Scientific Publishing Co, Inc. Princeton, NJ USA.
Maltoni C, Lefemine G, Cotti G, Perino G. 1988.Long-term carcinogenicity bioassays on
trichloroethylene administered by inhalation to Sprague-Dawley rats and Swiss and B6C3F1
mice. In Living in a Chemical World, eds Maltoni C, Selikoff IJ. Volume 534, Annals of the
New York Academy of Sciences, New York, NY USA.
Marrazzini, A., Betti, C., Bernacchi, F., Barrai, I. & Barale, R. (1994) Micronucleus test and
metaphase analyses in mice exposed to known and suspected spindle poisons. Mutagenesis, 9,
505-515
Miller, B.M., Adler, I.-D. (1992) Aneuploidy induction in mouse spermatocytes. Mutagenesis, 7,
69-76
Moore MM, Harrington-Brock K. 2000. Mutagenicity of trichloroethylene and its metabolites:
implications for the risk assessment of trichloroethylene. Environ Health Perspect 108(Suppl
2):215-223.
6/22/2009
749
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Morgan RW, Kelsh MA, Zhao K, Heringer S. 1998. Mortality of aerospace workers exposed to
trichloroethylene. Epidemiology 9:424-431.
Morgan JW, Cassidy RE. 2002. Community cancer assessment in response to long-time
exposure to perchlorate and trichloroethylene in drinking water. J Occup Environ Med.
44:616-621.
Narotsky, MG; Kavlock, RJ. (1995) A multidisciplinary approach to toxicological screening: II.
Developmental toxicity. Journal of Toxicology and Environmental Health 45:145-171.
National Research Council (NRC). 2006. Assessing the Human Health Risks of
Trichloroethylene: Key Scientific Issues. Washington DC: National Academy Press.
NCI (National Cancer Institute). (1976) Carcinogenesis bioassay of trichloroethylene. Division
of Cancer Cause and Prevention, National Cancer Institute, U.S. Department of Health,
Education, and Welfare, DHEW Publication No. (NIH) 76-802, Technical Report Series No. 2,
218 pages; NCI-CG-TR-2; NTIS PB254122.
http://ntp.niehs.nih.gov/ntp/htdocs/LT rpts/tr002.pdf.
Nesslany, F., Marzin, D. (1999) A micromethod for the in vitro micronucleus assay.
Mutagenesis, 14, 403-410
Ni, Y.-C., Wong, T.-Y., Kadlubar, F.F. & Fu, P.P. (1994) Hepatic metabolism of chloral hydrate
to free radical(s) and induction of lipid peroxidation. Biochem. biophys. Res. Commun., 204,
937-943
NTP (National Toxicology Program). (1988) Toxicology and carcinogenesis studies of
trichloroethylene (CAS no. 79-01-6) in four strains of rats (ACI, August, Marshall, Osborne-
Mendel) (gavage studies). Public Health Service, U.S. Department of Health and Human
Services; NTP TR-273; NIH Publication No. 88-2529. Available from the National Institute of
Environmental Health Sciences, Research Triangle Park, NC, and the National Technical
Information Service, Springfield, VA; PB88-218896.
http://ntp.niehs.nih.gov/ntp/htdocs/LT_rpts/tr273.pdf.
NTP (National Toxicology Program). (1990) Carcinogenesis studies of trichloroethylene
(without epichlorohydrin) in F344/N rats and B6C3F1 mice (gavage studies). Public Health
6/22/2009
750
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Service, U.S. Department of Health and Human Services; NTP TR-243; NIH Publication No. 90-
1779. Available from the National Institute of Environmental Health Sciences, Research
Triangle Park, NC. http://ntp.niehs.nih.gov/ntp/htdocs/LT rpts/tr243.pdf.
Nutley EV, Tcheong AC, Allen JW, Collins BW, Ma M, Lowe XR, Bishop JB, Moore DH,
Wyrobek AJ. 1996. Micronuclei induced in round
Odum J Foster JR, Green T. 1992. A mechanism for the development of clara cell lesions in the
mouse lung after exposure to trichloroethylene. Chem-Biol Interactions 83:135-153.
Prendergast JA, Jones RA, Jenkins LJ, Siegel J. 1967. Effects on experimental animals of long-
term inhalation of trichloroethylene, carbon tetrachloride, 1,1,1-trichloroethane,
dichlorodifluoromethane, and 1,1-dichloroethylene. Toxicology and Applied Pharmacology 10:
270-289.
Raaschou-Nielsen O, Hansen J, McLaughlin JK, Kolstad H, Christensen JM, Tarone RE, Olsen
JH. 2003. Cancer risk among workers at Danish companies using trichloroethylene: a cohort
study. Am J Epidemiol 158:1182-1192.
Radican L, Blair A, Stewart P, Wartenberg D. 2008. Mortality of aircraft maintenance workers
exposed to trichloroethylene and other hydrocarbons and chemicals: extended follow-up. J
Occup Environ Med 50:1306-1319.
Ritz B. 1999. Cancer mortality among workers exposed to chemicals during uranium
processing. J Occup Environ Med. 41:556-566.
Russo A, Pacchierotti F, Metalli P. 1984. Nondisjunction induced in mouse spermatogenesis by
chloral hydrate, a metabolite of trichloroethylene. Environmental Mutagenesis 6: 695-703.
Russo A, Levis AG. 1992a. Detection of aneuploidy in male germ cells of mice by means of
meiotic micronucleus assay. MutatRes. 281: 187-191.
Saygun M, Cakmak A, Ekici A, Pinar T, Bulcun E, Ulu N, Ekici M. 2007. Five annual
observations of respiratory findings in gun factory workers exposed to solvents. J Occup
Environ Med 49:909-912.
6/22/2009
751
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Sbrana I, Di Sibio A, Lomi A, Scarcelli V. 1993. C-mitosis and numerical chromosome
aberration analyses in human lymphocytes: 10 known or suspected spindle poisons. Mutat Res.
281: 57-70.
Scott JE, Forkert PG, Oulton M, Rasmusson M-G, Temple S, Fraser MO, Whitefield S. 1988.
Pulmonary toxicity of trichloroethylene: induction of changes in surfactant phospholipids and
phospholipase A2 activity in the mouse lung. Experimental and Molecular Pathology 49:
141-150.
Siemitycki J. 1991. Risk Factors for Cancer in the Workplace. J Siemiatycki, Ed. Baca Raton:
CRC Press.
Stewart BW, Le Mesurier SM, Lykke AWJ. 1979. Correlation of biochemical and morphological
changes induced by chemical injury to the lung. Chem-Biol Interactions 26: 321-338.
U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. Risk Assessment Forum,
Washington, DC; EPA/630/P-03/001B. Available from: .
U.S. EPA. (2005b) Supplemental guidance for assessing susceptibility from early-life exposure
to carcinogens. Risk Assessment Forum, Washington, DC; EPA/630/R-03/003F. Available from:
.
Vagnarelli P, De Sario A, De Carli L. 1990. Aneuploidy induced by chloral hydrate detected in
human lymphocytes with the Y97 probe. Mutagenesis 5: 591-592.
Van Duuren Bl, Goldschmidt BM, Loewengart G, 1979. Carcinogenicity of halogenated olefinic
and aliphatic hydrocarbons in mice. Journal of the National Cancer Institute 63: 1433-1439.
Villaschi S, Giovanetti A, Lombardi CC, Nicolai G, Garbati M, Andreozzi U. 1991. Damage
and repair of mouse bronchial epithelium following acute inhalation of trichloroethylene.
Experimental Lung Research 15: 601-614.
Wilcosky TC, Checkoway H, Marshall EG, Tyroler HA. 1984. Cancer mortality and solvent
exposure in the rubber industry. Am Ind Hyg Assoc J 45:809-811.
6/22/2009
752
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Zhao Y, Krishnadasan A, Kennedy N, Morgenstern H, Ritz B. 2005. Estimated effects of
solvents and mineral oils on cancer incidence and Mortality in a cohort of aerospace workers.
Am J Ind Med 48:249-258.
1
6/22/2009
753
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.7 Reproductive and developmental toxicity
4.7.1 Reproductive toxicity
An assessment of the human and experimental animal data, taking into consideration the
overall weight of the evidence, demonstrates a concordance of adverse reproductive outcomes
associated with TCE exposures. Effects on male reproductive system integrity and function are
particularly notable and are discussed below. Cancers of the reproductive system in both males
and females have also been identified and are discussed below.
4.7.1.1 Human reproductive outcome data
A number of human studies have been conducted that examined the effects of TCE on
male and female reproduction following occupational and community exposures. These are
described below and summarized in Table 4.7-1. Epidemiological studies of female human
reproduction examined infertility and menstrual cycle disturbances related to TCE exposure.
Other studies of exposure to pregnant women are discussed in the section on human
developmental studies (see Section 4.7.2.1). Epidemiological studies of male human
reproduction examined reproductive behavior, altered sperm morphology, altered endocrine
function, and infertility related to TCE exposure.
4.7.1.1.1 Female and male combined human reproductive effects
Reproductive Behavior. A residential study of individuals living near the Rocky
Mountain Arsenal in Colorado examined the reproductive outcomes in 75 men and 71 women
exposed to TCE in drinking water (ATSDR, 2001). TCE exposure was classified as high (>10.0
ppb), medium (>5.0 ppm to <10.0 ppb), and low (<5.0 ppb). Altered libido for men and women
combined was observed in a dose-response fashion, although the results were non-significant.
The results were not stratified by gender.
4.7.1.1.2 Female human reproductive effects
Infertility. Sallmen et al. (1995) examined maternal occupational exposure to organic
solvents and time-to-pregnancy. Cases of spontaneous abortion and controls from a prior study
of maternal occupational exposure to organic solvents in Finland during 1973-1983 and
pregnancy outcome (Lindbohm et al., 1990) were used to study time-to-pregnancy of 197
couples. Exposure was assessed by questionnaire during the first trimester and confirmed with
6/22/2009
754
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
employment records. Biological measurements of TCA in urine in 64 women who held the same
job during pregnancy and measurement (time of measurement not stated) had a median value of
48.1 |imol/L (mean: 96.2 ± 19.2 |imol/L) (Lindbohm et al., 1990). Nineteen women had low
exposure to TCE (used <1 or 1-4 times/week), and 9 had high exposure to TCE (daily use). In
this follow up study, an additional questionnaire on time-to-pregnancy was answered by the
mothers (Sallmen et al., 1995). The incidence density ratio (IDR) was used in this study to
estimate the ratio of average incidence rate of pregnancies for exposed women compared to non-
exposed women; therefore, a lower IDR indicates infertility. For TCE, a reduced incidence of
fecundability was observed in the high exposure group (IDR: 0.61, 95% CI 0.28-1.33) but not in
the low exposure group (IDR: 1.21, 95% CI: 0.73-2.00). A similar study of paternal
occupational exposure (Sallmen et al., 1998) is discussed in Section 4.1.1.2.
The residential study in Colorado discussed above did not observe an effect on lifetime
infertility infertility in the medium (ORadj = 0.45; 95% CI = 0.02-8.92) or high exposure groups
(ORadj = 0.88; 95% CI = 0.13-6.22) (ATSDR, 2001). Curiously, exposed women had more
pregnancies and live births than controls.
Menstrual Cycle Disturbance. The ATSDR (2001) study discussed above also examined
effects on the menstrual cycle (ATSDR, 2001). Non-significant associations without a dose-
response were seen for abnormal menstrual cycle in women (ORadj 2.23, 95% CI: 0.45-11.18).
Other studies have examined the effect of TCE exposure on the menstrual cycle. One
study examined women working in a factory assembling small electrical parts (Zielinski, 1973,
translated). The mean concentration of TCE in indoor air was reported to be 200 mg/m3.
Eighteen percent of the 140 exposed women suffered from amenorrhea, compared to only 2% of
the 44 non-exposed workers. The other study examined 75 men and women working in dry
cleaning or metal degreasing (Bardodej and Vyskocil, 1956). Exposures ranged from 0.28-3.4
mg/L, and length of exposure ranged from 0.5 to 25 years. This study reported that many
women experienced menstrual cycle disturbances, with a trend for increasing air concentrations
and increasing duration of exposure.
An additional case study of a 20-year-old woman was occupationally exposed to TCE via
inhalation. The exposure was estimated to be as high as 10 mg/mL or several thousand ppm,
based on urine samples 21-25 days after exposure of 3.2 ng/mL of total trichloro-compounds.
The primary effect was neurological, although she also experienced amenorrhea, followed by
irregular menstruation and lack of ovulation as measured by basal body temperature curves
(Sagawa et al., 1973).
4.7.1.1.3 Male human reproductive effects
Reproductive Behavior. One study reported on the effect of TCE exposure on the male
6/22/2009
755
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
reproductive behavior in 75 men working in dry cleaning or metal degreasing (Bardodej and
Vyskocil, 1956). Exposures ranged from 0.28-3.4 mg/L, and length of exposure ranged from
0.5 to 25 years. This study found that men experienced decreased potency or sexual
disturbances; the authors speculated that the effects on men could be due to the CNS effects of
TCE exposure. This study also measured serial neutral 17-ketosteroid determinations but they
were found to be not statistically significant (Bardodej and Vyskocil, 1956).
An occupational study of 30 men working in a money printing shop were exposed to
TCE for <1 year to 5 years (El Ghawabi et al., 1973). Depending on the job description, the
exposures ranged from 38-172 ppm TCE. Ten (33%) men suffered from decreased libido,
compared to three (10%) of unexposed controls. However, these results were not stratified by
exposure level or duration. The authors speculate that decreased libido was likely due to the
common symptoms of fatigue and sleepiness.
A case study described a 42 year-old man exposed to TCE who worked as an aircraft
mechanic for approximately 25 years (Saihan et al., 1978). He suffered from a number of health
complaints including gynaecomastia and impotence, along with neurotoxicity and
immunotoxicity. In addition, he drank alcohol daily which could have increased his response to
TCE.
Altered Sperm Quality. Genotoxic effects on male reproductive function were examined
in a study evaluating occupational TCE exposure in 15 male metal degreasers (Rasmussen et al.,
1988). No measurement of TCE exposure was reported. Sperm count, morphology, and
spermatozoa Y-chromosomal non-disjunction during spermatogenesis were examined, along
with chromosomal aberrations in cultured lymphocytes. A non-significant increase in percentage
of two fluorescent Y-bodies (YFF) in spermatozoa were seen in the exposed group (p> 0. 10),
and no difference was seen in sperm count or morphology compared to controls.
An occupational study of men using TCE for electronics degreasing (Chia et al., 1996,
1997; Goh et al., 1998) examined subjects (n = 85) who were offered a free medical exam if they
had no prior history related to endocrine function, no clinical abnormalities, and normal liver
function tests; no controls were used. These participants provided urine, blood, and sperm
samples. The mean urine TCA level was 22.4 mg/g creatinine (range: 0.8-136.4 mg/g
creatinine). In addition, 12 participants provided personal 8-hour air samples, which resulted in
a mean TCE exposure of 29.6 ppm (range: 9-131 ppm). Sperm samples were divided into two
exposure groups; low for urine TCE less than 25 mg/g creatinine, and high for urine TCA greater
than or equal to 25 mg/g creatinine. A decreased percentage of normal sperm morphology was
observed in the sperm samples in the high exposure group (n = 48) compared to the low
exposure group (n = 37). However, TCE exposure had no effect on semen volume, sperm
density, or motility. There was also an increased prevalence of hyperzoospermia (sperm density
6/22/2009
756
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
of >120 million sperm per mL ejaculate) with increasing urine TCA levels (Chia et al., 1996).
Altered Endocrine Function. Two studies followed up on the study by Chia et al. (1996)
to examine endocrine function (Chia et al., 1997; Goh et al., 1998). The first examined serum
testosterone, follicle-stimulating hormone (FSH), dehydroepiandrosterone sulphate (DHEAS),
and sex-hormone binding globulin (SHBG) (Chia et al., 1997). With increased years of exposure
to TCE, an increase in DHEAS levels were seen, from 255 ng/mL for <3 years to 717.8 ng/mL >
7 years exposure. Also with increased years of exposure to TCE, decreased FSH, SHBG and
testosterone levels were seen. The authors speculated these effects could be due to decreased
liver function related to TCE exposure (Chia et al., 1997).
The second follow-up study of this cohort studied the hormonal effects of chronic low-
dose TCE exposure in these men (Goh et al., 1998). Because urine TCE measures only indicate
short-term exposure, long-term exposure was indicated by years of exposure. Hormone levels
examined include androstenedione, Cortisol, testosterone, aldosterone, sex-hormone binding
globulin (SHBG), and insulin. Results show that a decrease in serum levels of testosterone and
SHBG were significantly correlated with years of exposure to TCE, and an increase in insulin
levels were seen in those exposed for less than 2 years. Androstenedione, Cortisol, and
aldosterone were in normal ranges and did not change with years of exposure to TCE.
Infertility. Sallmen et al. (1998) examined paternal occupational exposure and time-to-
pregnancy among their wives. Cases of spontaneous abortion and controls from a prior study of
pregnancy outcome (Taskinen et al., 1989) were used to study time-to-pregnancy of 282 couples.
Exposure was determined by biological measurements of the father who held the same job
during pregnancy and measurement (time of measurement not stated) and questionnaires
answered by both the mother and father. An additional questionnaire on time-to-pregnancy was
answered by the mother for this study six years after the original study (Sallmen et al., 1998).
The level of exposure was determined by questionnaire and classified as "low/intermediate" if
the chemical was used <1 or 1-4 days/week and biological measures indicated high exposure
(defined as above the reference value for the general population), and "high" if used daily or if
biological measures indicated high exposure. For 13 men highly exposed, mean levels of urine
TCA were 45 |imol/L (SD 42 |imol/L; median 31 |imol/L); for 22 men low/intermediately
exposed, mean levels of urine TCA were 41 |imol/L (SD 88 |imol/L; median 15 |imol/L). The
terminology IDR was replaced by fecundability density ratio (FDR) in order to reflect that
pregnancy is a desired outcome; therefore, a high FDR indicates infertility. No effect was seen
on fertility in the low exposure group (FDR: 0.99, 95% CI 0.63-1.56) or in the intermediate/high
exposure group (FDR: 1.03, 95% CI 0.60-1.76). However, the exposure categories were
grouped by low/intermediate versus high, whereas the outcome categories were grouped by low
versus intermediate/high, making a dose-response association difficult.
6/22/2009
757
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
A small occupational study reported on eight male mechanics exposed to TCE for at least
two years who sought medical treatment for infertility (Forkert et al., 2003). The wives were
determined to have normal fertility. Samples of urine from two of the eight male mechanics
contained TCA and/or TCOH, demonstrating the rapid metabolism in the body. However,
samples of seminal fluid taken from all eight individuals detected TCE and the metabolites
chloral hydrate and TCOH, with two samples detecting DCA and one sample detecting TCA.
Five unexposed controls also diagnosed with infertility did not have any TCE or metabolites in
samples of seminal fluid. There was no control group that did not experience infertility.
Increased levels of TCE and its metabolites in the seminal fluid of exposed workers compared to
lower levels found in their urine samples was explained by cumulative exposure and
mobilization of TCE from adipose tissue, particularly that surrounding the epididymis. In
addition, CYP2E1 was detected in the epididymis, demonstrating that metabolism of TCE can
occur in the male reproductive tract. However, this study could not directly link TCE to the
infertility, as both the exposed and control populations were selected due to their infertility.
The ATSDR (2001) study discussed above on the reproductive effects from TCE in
drinking water of individuals living near the Rocky Mountain Arsenal in Colorado did not
observe infertility or other adverse reproductive effects for the high exposure group compared to
the low exposure group (ORadj = 0.83; 95% CI = 0.11-6.37). Curiously, exposed men had more
pregnancies and live births than controls.
4.7.1.1.4 Summary of human reproductive toxicity
Following exposure to TCE, adverse effects on the female reproductive system observed
include reduced incidence of fecundability (as measured by time-to-pregnancy) and menstrual
cycle disturbances. Adverse effects on the male reproductive system observed include altered
sperm morphology, hyperzoospermia, altered endocrine function, decreased sexual drive and
function, and altered fertility. These are summarized in Table 4.7-1.
Table 4.7.1. Human reproductive effects
Subjects
Exposure
Effect
Reference
Female and Male Combined Effects
Reproductive Behavior
75 men and 71 women
Low: <5.0 ppb
Altered libido a
ATSDR, 2001
living near Rocky
Medium: >5.0—<10.0 ppb
Low: referent
Mountain Arsenal,
High: <10.0 ppb
Med: ORadj = 0.67 (95% CI = 0.18-2.49)
Colorado
Highest: <15 ppb
High: ORadj = 1.65 (95% CI = 0.54-5.01)
Highest: ORadj = 2.46 (95%
CI = 0.59-10.28)
6/22/2009
758
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
Female Effects
Infertility
197 women
occupationally exposed
to solvents in Finland
1973-1983
Urine TCA (|imol/L)b
Median: 48.1
Mean: 96.2 ± 19.2
Reduced incidence of fecundability in the
high exposure group0 as measured by time to
pregnancy
Low: IDR d = 1.21 (95%CI = 0.73-2.00)
High: IDR d = 0.61 (95%CI = 0.28-1.33)
Sallmen et al.,
1995
71 women living near
Rocky Mountain
Arsenal, Colorado
Low: <5.0 ppb
Med: >5.0 to <10.0 ppb
High: <10.0 ppb
No effect on lifetime infertility a
Low: referent
Med: ORadj = 0.45 (95% CI = 0.02-8.92)
High: ORadj = 0.88 (95% CI = 0.13-6.22)
ATSDR, 2001
Menstrual Cvcle Disturbance
71 women living near
Rocky Mountain
Arsenal, Colorado
Low: <5.0 ppb
Med: >5.0 to <10.0 ppb
High: <10.0 ppb
Increase in abnormal menstrual cycle (defined
as <26 days or >30 days)
Low: referent
Med: ORadj = 4.17 (95% CI = 0.31-56.65)
High: ORadj = 2.39 (95% CI = 0.41-13.97)
ATSDR, 2001
184 women working in
a factory assembling
small electrical parts in
Poland
mean indoor air TCE: 200
mg/m3
18% reporting increase in amenorrhea in
exposed group (n = 140), compared to 2%
increase in unexposed group (n = 44)
Zielinski,
1973
32 women working in
dry cleaning or metal
degreasing in
Czechoslovakia d
0.28-3.4 mg/L TCE for
0.5-25 years
31% reporting increase in menstrual
disturbances a
Bardodej and
Vyskocil,
1956
20-year-old woman was
occupationally exposed
to TCE via inhalation
Urine total trichloro-
compounds 3.2 ng/mL
(21-25 days after
exposure)
amenorrhea, followed by irregular
menstruation and lack of ovulation
Sagawa et al.,
1973
Male Effects
Reproductive Behavior
43 men working in dry
cleaning or metal
degreasing in
Czechoslovakia
0.28-3.4 mg/L TCE for
0.5-25 years
30% reporting decreased potency a
Bardodej and
Vyskocil,
1956
6/22/2009
759
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
30 male workers in a
money printing shop in
Egypt
38-172 ppm TCE
Decreased libido reported in 10 men (33%),
compared to 3 men in the control group
(10%)
El Ghawabi et
al., 1973
42 year-old male aircraft
mechanic in UK
TCE exposure reported
but not measured;
exposure for 25 years
Gynaecomastia, impotence
Saihan et al.,
1978
Altered Sperm Oualitv
15 men working as
metal degreasers in
Denmark
TCE exposure reported
but not measured
Non-significant increase in percentage of two
fluorescent Y-bodies (YFF) in spermatozoa;
no effect on sperm count or morphology
Rasmussen et
al., 1988
85 men of Chinese
descent working in an
electronics factory
Mean personal air TCE:
29.6 ppm; Mean urine
TCA: 22.4 mg/g creatinine
Decreased normal sperm morphology and
hyperzoospermia
Chia et al.,
1996
Altered Endocrine Function
85 men of Chinese Mean personal air TCE:
descent working in an 29.6 ppm; Mean urine
electronics factory TCA: 22.4 mg/g creatinine
Increased DHEAS and decreased FSH, SHBG
and testosterone levels; dose-response
observed
Chia et al.,
1997
85 men of Chinese
descent working in an
electronics factory
Mean personal air TCE:
29.6 ppm; Mean urine
TCA: 22.4 mg/g creatinine
Decreased serum levels of testosterone and
SHBG were significantly correlated with
years of exposure to TCE; increased insulin
levels for exposure <2 years
Goh etal.,
1998
Infertilitv
282 men occupationally
exposed to solvents in
Finland 1973-1983
Urine TCA (|imol/L):
High exposure:0
Mean: 45 (SD 42)
Median 31
Low exposure:0
Mean: 41 (SD 88)
Median: 15
No effect on fecundability 0 (as measured by
time to pregnancy)
Low: FDR d= 0.99 (95% CI = 0.63-1.56)
Intermediate/High: FDR° = 1.03 (95%
CI = 0.60-1.76)
Sallmen et al.,
1998
6/22/2009
760
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
8 male mechanics seeking Urine (|imol/):
treatment for infertility in TCA: <0.30-4.22
Infertility could not be associated with TCE
as controls were 5 men also in treatment for
infertility
Forkert et al.,
2003
Canada
TCOH: <0.60-0.89
Seminal fluid (pg/extract):
TCE: 20.4-5,419.0
Chloral: 61.2-1,739.0
TCOH 2.7-25.5
TCA: <100-5,504
DCA: <100-13,342
75 men living near
Rocky Mountain
Arsenal, Colorado
Low: <5.0 ppb
Med: >5.0 to <10.0 ppb
High: <10.0 ppb
No effect on lifetime infertility (not defined) ATSDR, 2001
Low: referent
Med: n/a
High: ORadj = 0.83 (95% CI = 0.11-6.37)
aNot defined by the authors.
b As reported in Lindbohm et al. (1990).
0 Low/intermediate exposure indicated use of TCE <1 or 1-4 days/week, and biological measures indicated high
exposure. High exposure indicated daily use of TCE, or if biological measures indicated high exposure.
d IDR = incidence density ratio; FDR = fecundity density ratio.
e Number inferred from data provided in Tables 2 and 3 in Bardodej and Vyskocil (1956).
4.7.1.2 Animal reproductive toxicity studies
A number of animal studies have been conducted that examined the effects of TCE on
reproductive organs and function following either inhalation or oral exposures. These are
described below and summarized in Tables 4.7.2 and 4.7.3. Other animal studies of offspring
exposed during fetal development are discussed in the section on animal developmental studies
(see Section 4.7.2.2).
4.7.1.2.1 Inhalation exposures
Studies in rodents exposed to TCE via inhalation are described below and summarized in
Table 4.7.2. These studies focused on various aspects of male reproductive organ integrity,
spermatogenesis, or sperm function in rats or mice. In the studies published after the year 2000,
the effects of either 376 or 1,000 ppm TCE were studied following exposure durations ranging
from 1 to 24 weeks, and adverse effects on male reproductive endpoints were observed.
Kumar et al. (2000a) exposed male Wistar rats in whole body inhalation chambers to 376
ppm TCE for 4 hours/day, 5 days/week over several duration scenarios. These were: 2 weeks (to
observe the effect on the epididymal sperm maturation phase), 10 weeks (to observe the effect on
6/22/2009
761
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
the entire spermatogenic cycle), 5 weeks with 2 weeks rest (to observe the effect on primary
spermatocytes differentiation to sperm), 8 weeks with 5 weeks rest (to observe effects on an
intermediate stage of spermatogenesis), and 10 weeks with 8 weeks rest (to observe the effect on
spermatogonial differentiation to sperm). Control rats were exposed to ambient air. Weekly
mating with untreated females was conducted. At the end of the treatment/rest periods, the
animals were sacrificed; testes and cauda epididymes tissues were collected. Alterations in testes
histopathology (smaller, necrotic spermatogenic tubules), increased sperm abnormalities, and
significantly increased pre- and/or post-implantation loss in litters were observed in the groups
with 2 or 10 weeks of exposure, or 5 weeks of exposure with 2 weeks rest. It was hypothesized
that post-meiotic cells of spermatogenesis and epididymal sperm were affected by TCE
exposure, leading to reproductive impairment.
To test the hypothesis that TCE exposure adversely affects sperm function and
fertilization, Xu et al. (2004) conducted a study in which male CD-I mice were exposed by
inhalation to atmospheres containing 1,000 ppm (5.37 mg/L) TCE for 1 to 6 weeks (6 hours/day,
5 days/week). After each TCE exposure, body weights were recorded. Following termination,
the right testis and epididymis of each treated male were weighed, and sperm was collected from
the left epididymis and vas deferens for assessment of the number of total sperm and motile
sperm. Sperm function was evaluated in the following experiments: 1) suspensions of
capacitated vas deferens/cauda epididymal sperm were examined for spontaneous acrosome
reaction, 2) in vitro binding of capacitated sperm to mature eggs from female CF-1 mice
(expressed as the number of sperm bound per egg) was assessed, and 3) in vivo fertilization was
evaluated via mating of male mice to superovulated female CF-1 mice immediately following
inhalation exposure; cumulus masses containing mature eggs were collected from the oviducts of
the females, and the percentage of eggs fertilized was examined. Inhalation exposure to TCE did
not result in altered body weight, testis and epididymis weights, sperm count, or sperm
morphology or motility. Percentages of acrosome-intact sperm populations were similar
between treated and control animals. Nevertheless, for males treated with TCE for 2 or more
weeks decreases were observed in the number of sperm bound to the oocytes in vitro (significant
at 2 and 6 weeks, p < 0.001). In a follow-up assessment, control sperm were incubated for 30-
minutes in buffered solutions of TCE or metabolites (chloral hydrate or trichloroethanol); while
TCE-incubation had no effect on sperm-oocyte binding, decreased binding capacity was noted
for the metabolite-incubated sperm. The ability for sperm from TCE-exposed males to bind to
and fertilize oocytes in vivo was also found to be significantly impaired (p < 0.05).
A study designed to investigate the role of testosterone, and of cholesterol and ascorbic
acid (which are primary precursors of testosterone) in TCE-exposed rats with compromised
reproductive function was conducted by Kumar et al. (2000b). Male Wistar rats 12-13/group)
6/22/2009
762
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
were exposed (whole body) to 376 ppm TCE by inhalation for 4 hours/day, 5 days/week, for
either 12 or 24 weeks and then terminated. Separate ambient-air control groups were conducted
for the 12- and 24-week exposure studies. Epididymal sperm count and motility were evaluated,
and measures of 17-P-hydroxy steroid dehydrogenase (17-P-HSD), testicular total cholesterol
and ascorbic acid, serum testosterone, and glucose 6-p dehydrogenase (G6-PDH) in testicular
homogenate were assayed. In rats exposed to TCE for either 12 or 24 weeks, total epididymal
sperm count and motility, serum testosterone concentration, and specific activities of both 17-P-
HSD and G6-PDH were significantly decreased (p < 0.05), while total cholesterol content was
significantly (p < 0.05) increased. Ascorbic acid levels were not affected.
In another study, Kumar et al. (2001) utilized the same exposure paradigm to examine
cauda epididymal sperm count and motility, testicular histopathology, and testicular marker
enzymes: sorbitol dehydrogenase (SDH), G6-PDH, glutamyl transferase (GT), and
glucuronidase, in Wistar rats (6/group). After 24 weeks of exposure, testes weights and
epididymal sperm count and motility were significantly decreased (p < 0.05). After 12 weeks of
TCE exposure, histopathological examination of the testes revealed a reduced number of
spermatogenic cells in the seminiferous tubules, fewer spermatids as compared to controls, and
the presence of necrotic spermatogenic cells. Testicular atrophy, smaller tubules, hyperplastic
Ley dig cells, and a lack of spermatocytes and spermatids in the tubules were observed after 24
weeks of TCE exposure. After both 12 and 24 weeks of exposure, SDH and G6-PDH were
significantly (p < 0.05) reduced while GT and P-glucuronidase were significantly (p < 0.05)
increased.
In a study by Land et al. (1981), 8-10 week old male mice (C57BlxC3H)Fl (5 or
10/group) were exposed (whole body) by inhalation to a number of anesthetic agents for 5
consecutive days at 4 hours per day and sacrificed 28 days after the first day of exposure.
Chamber concentration levels for the TCE groups were 0.02 and 0.2%. The control group
received ambient air. Epididymal sperm were evaluated for morphological abnormalities. At 0.2
% TCE, the percent abnormal sperm in a sample of 1,000 was significantly (p < 0.01) increased
as compared to control mice; no treatment-related effect on sperm morphology was observed at
0.02 % TCE.
Forkert et al. (2002) exposed male CD-I mice by inhalation to 1,000 ppm TCE (6
hours/day, 5 day/week) for 4 consecutive weeks and observed sloughing of portions of the
epithelium upon histopathological evaluation of testicular and epididymal tissues.
Kan et al. (2007) also demonstrated that damage to the epididymal epithelium and sperm
of CD-I mice (4/group) resulted from exposure to 0 or 1,000 ppm TCE by inhalation for 6
hours/day, 5 days/week, for 1 to 4 weeks. Segments of the epididymis (caput, corpus, and
cauda) were examined by light and electron microscope. As early as 1 week after TCE exposure,
6/22/2009
763
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
degeneration and sloughing of epithelial cells from all three epididymal areas were observed by
light microscopy; these findings became more pronounced by 4 weeks of exposure. Vesiculation
in the cytoplasm, disintegration of basolateral cell membranes, and epithelial cell sloughing were
observed with electron microscopy. Sperm were found in situ in the cytoplasm of degenerated
epididymal cells. A large number of sperm in the lumen of the epididymis were abnormal,
including head and tail abnormalities.
Table 4.7.2. Summary of mammalian in vivo reproductive toxicity studies - inhalation
exposures
Reference
Species/strain/
sex/number
Exposure
level/
Duration
NOAEL;
LOAELa
Effects
Forkert et al.,
2002
Mouse, CD-I,
male, 6/group
0, 1,000 ppm
(5,374
mg/m3)b
6 hr/day,
5 days/wk,
19 days over
4 wks
LOAEL: 1,000
ppm
Urinary TCA and TCEOH increased by 2nd
and 3rd week, respectively. Cytochrome P450
2E1 and /j-nitrophenol hydroxylation in
epididymal epithelium > testicular Leydig
cells. Choral also generated from TCE in
epididymis > testis. Sloughing of epididymal
epithelial cells after 4 wk exposure.
Kan et al., 2007
Mouse, CD-I,
male, 4/group
0, 1,000 ppm
6 hrs/day,
5 days/wk,
1 to 4 wks
LOAEL: 1,000
ppm
Light microscopy findings: degeneration and
sloughing of epididymal epithelial cells as
early as 1 week into exposure; more severe by
4 weeks. Ultrastructional findings:
vesiculation in cytoplasm, disintegration of
basolateral cell membranes, sloughing of
epithelial cells. Sperm found in situ in
cytoplasm of degenerated epididymal cells.
Abnormalities of the head and tail in sperm
located in the epididymal lumen.
Kumar et al.,
2000a
Rat, Wistar,
male,
12-13/group
0, 376 ppm
4 hr/day,
5 days/wk,
2 to 10 wks
exposure, 2
to 8 wks rest
period
LOAEL: 376
ppm
Alterations in testes histopathology (smaller,
necrotic spermatogenic tubules), T sperm
abnormalities, and sig. T pre- and/or post-
implantation loss in litters observed in the
groups with 2 or 10 weeks of exposure, or 5
weeks of exposure with 2 weeks rest.
6/22/2009
764
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
sex/number
Exposure
level/
Duration
NOAEL;
LOAELa
Effects
Kumar et al.,
2000b
Rat, Wistar,
males,
12-13/group
0, 376 ppm
4 hr/day,
5 days/wk,
12 and 24
wks
LOAEL: 376
ppm
Sig. 1 in total epididymal sperm count and
sperm motility, with sig. 1 in serum
testosterone, sig. T in testes cholesterol, sig. 1
of glucose 6-phosphate dehydrogenase and
17-p-hydroxy steroid dehydrogenase at 12
and 24 wks exposure.
Kumar et al.,
2001
Rat, Wistar,
male, 6/group
0, 376 ppm
4 hr/day, 5
days/wk,
12 and 24
wks
LOAEL: 376
ppm
BW gain sig. j. Testis weight, sperm count
and motility sig. effect stronger with
exposure time. After 12 wk, numbers of
spermatogenic cells and spermatids |, some
of the spermatogenic cells appeared necrotic.
After 24 wk testes were atrophied, tubules
were smaller, had Sertoli cells and were
almost devoid of spermatocytes and
spermatids. Ley dig cells were hyperplastic.
SDH, G6PDH sig. j, GGT and p-
glucuronidase sig. t; effects stronger with
exposure time.
Land et al., 1981
Mouse,
C57BlxC3H(Fl),
male, 5 or
10/group
0, 0.02%,
0.2%
4 hrs/day,
5 days, 23
days rest
NOAEL: 0.02%
LOAEL: 0.2%
Sig. t percent morphologically abnormal
epididymal sperm
Xu et al., 2004
Mouse, CD-I,
male, 4 to
27/group
0, 1,000 ppm
(5.37 mg/L)
b
6 hrs/day,
5 days/wk,
1-6 wks
LOAEL: 1,000
ppm
Sig. i in vitro sperm-oocyte binding and in
vivo fertilization
a NOAEL (No Observed Adverse Affect Level) and LOAEL (Lowest Observed Adverse Affect Level) are based
upon reported study findings.
b Dose conversion calculations by study author(s).
6/22/2009
765
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.7.1.2.2 Oral exposures
A variety of studies were conducted to assess various aspects of male and/or female
reproductive capacity in laboratory animal species following oral exposures to TCE. These are
described below and summarized in Table 4.7-3. They include studies that focused on male
reproductive outcomes in rats or rabbits following gavage or drinking water exposures (Zenick et
al., 1984; DuTeaux et al., 2003, 2004b; Veeramachaneni et al., 2001), studies that focused on
female reproductive outcomes in rats following gavage or drinking water exposures (Berger and
Horner, 2003; Cosby and Dukelow, 1992; Manson et al., 1984; Wu and Berger, 2007, 2008), and
studies assessed fertility and reproductive outcome in both sexes following dietary exposures to
CD-I mice or F344 rats (George et al., 1985, 1986).
Studies assessing male reproductive outcomes
Zenick et al. (1984) conducted a study in which sexually experienced Long-Evans
hooded male rats were administered 0, 10, 100, or 1,000 mg/kg-day TCE by gavage in corn oil
for 6 weeks. A 4-week recovery phase was also incorporated into the study design. Endpoints
assessed on Weeks 1 and 5 of treatment included copulatory behavior, ejaculatory plug weights,
and ejaculated or epididymal sperm measures (count, motility, and morphology). Sperm
measures and plug weights were not affected by treatment, nor were Week 6 plasma testosterone
levels found to be altered. TCE effects on copulatory behavior (ejaculation latency, number of
mounts, and number of intromissions) were observed at 1,000 mg/kg-day; these effects were
recovered by 1-4 weeks post-treatment. Although the effects on male sexual behavior in this
study were believed to be unrelated to narcotic effects of TCE, a later study by Nelson and
Zenick (1986) showed that naltrexone (an opioid receptor antagonist, 2.0 mg/kg, IP,
administered 15 minutes prior to testing) could block the effect. Thus, it was hypothesized that
the adverse effects of TCE on male copulatory behavior in the rat at 1,000 ppm may in fact be
mediated by the endogenous opioid system at the CNS level.
In a series of experiments by DuTeaux et al. (2003, 2004b), adult male rats were
administered 0%, 0.2%, or 0.4 % TCE (v/v) (equivalent to 0, 2.73 mg/L, or 5.46 mg/L) in a
solution of 3% ethoxylated castor oil in drinking water for 14 days. These concentrations were
within the range of measurements obtained in formerly contaminated drinking water wells, as
reported by ATSDR (1997). The average ingested doses of TCE (based upon animal body
weight and average daily water consumption of 28 mL) were calculated to be 143 mg/kg-day or
270 mg/kg-day for the low and high dose groups, respectively (DuTeaux et al., 2008). Cauda
epididymal and vas deferens sperm from treated males were incubated in culture medium with
oviductal cumulus masses from untreated females to assess in vitro fertilization capability.
Treatment with TCE resulted in a dose-dependent decrease in the ability of sperm to fertilize
6/22/2009
766
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
oocytes. Terminal body weights and testis/epididymal weights were similar between control and
treated groups. Evaluation of sperm concentration or motility parameters did not reveal any
treatment-related alterations; acrosomal stability and mitochondrial membrane potential were not
affected by treatment. Although no histopathological changes were observed in the testis or in
the caput, corpus, or cauda epididymis, exposure to 0.2% and 0.4% TCE resulted in slight
cellular alterations in the efferent ductule epithelium.
Veeramachaneni et al. (2001) evaluated the effects of drinking water containing
chemicals typical of ground water near hazardous waste sites (including 9.5 or 28.5 ppm TCE)
on male reproduction. In this study, pregnant Dutch-belted rabbits were administered treated
drinking water from gestation day 20; treatment continued through the lactation period and to
weaned offspring (7-9/group) through postnatal week 15. Deionized water was administered
from postnatal weeks 16-61, at which time the animals were terminated. At 57-61 weeks of
age, ejaculatory capability, and seminal, testicular, epididymal, and endocrine characteristics
were evaluated. In both treated groups, long-term effects consisted of decreased copulatory
behavior (interest, erection, and/or ejaculation), significant increases in acrosomal dysgenesis
and nuclear malformations (p < 0.03), and significant decreases in serum concentration of
luteinizing hormone (p < 0.05) and testosterone secretion after human chorionic gonadotropin
administration (p < 0.04). There were no effects on total spermatozoa per ejaculate or on daily
sperm production. The contribution of individual drinking water contaminants to adverse male
reproductive outcome could not be discerned in this study. Additionally, it was not designed to
distinguish between adverse effects that may have resulted from exposures in late gestation (i.e.,
during critical period of male reproductive system development) versus postnatal life.
Studies assessing female reproductive outcomes
In a study that evaluated postnatal growth following gestational exposures, female
B6D2F1 mice (7-12/group) were administered TCE at doses of 0, 1% LD50 (24 mg/kg-day),
and 10%) LD50 (240 mg/kg-day) by gavage in corn oil from gestation days 1-5, 6-10, or 11-15
(day of mating was defined as gestation day 1) (Cosby and Dukelow, 1992). Litters were
examined for pup count, sex, weight, and crown-rump measurement until postnatal day 21.
Some offspring were retained to 6 weeks of age, at which time they were killed and the gonads
were removed, weighed and preserved. No treatment-related effects were observed in the dams
or offspring. In a second series of studies conducted by Cosby and Dukelow and reported in the
same paper, TCE and its metabolites dichloroacetic acid (DCA), trichloroacetic acid (TCA), and
trichloroethanol (TCOH) were added to culture media with capacitated sperm and cumulus
masses from B6D2F1 mice to assess effects on in vitro fertilization. Dose-related decreases in
6/22/2009
767
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
fertilization were observed for DC A, TCA, and TCOH at 100 and 1,000 ppm, but not with TCE.
Synergystic effects were not observed with TCA and TCOH.
A study was conducted by Manson et al. (1984) to determine if subchronic oral exposure
to TCE affected female reproductive performance, and if TCE or its metabolites trichloroacetic
acid or trichloroethanol accumulated in female reproductive organs or neonatal tissues. Female
Long-Evans hooded rats (22-23/group) were administered 0 (corn oil vehicle), 10, 100, or 1,000
mg/kg-day of TCE by gavage for 2 weeks prior to mating, throughout mating, and to gestation
day 21. Delivered pups were examined for gross anomalies, and body weight and survival were
monitored for 31 days. Three maternal animals per group and 8-10 neonates per group (killed
on postnatal days 3 and 31) were analyzed for TCE and metabolite levels in tissues. TCE
exposure resulted in 5 deaths and decreased maternal body weight gain at 1,000 mg/kg-day, but
did not affect estrous cycle length or female fertility at any dose level. There were no evident
developmental anomalies observed at any treatment level; however, at 1,000 mg/kg-day there
was a significant increase in the number of pups (mostly female) born dead, and the cumulative
neonatal survival count through PND 18 was significantly decreased as compared to control.
TCE levels were uniformly high in fat, adrenal glands, and ovaries across treatment groups, and
trichloroacetic acid (TCA) levels were high in uterine tissue. TCE levels in the blood, liver, and
milk contents of the stomach increased in female PND-3 neonates across treatment groups.
These findings suggest that increased metabolite levels did not influence fertility, mating
success, or pregnancy outcome.
In another study that examined the potential effect of TCE on female reproductive
function, Berger and Horner (2003) conducted 2-week exposures of Sprague-Dawley derived
female Simonson rats to tetrachloroethylene, trichloroethylene, several ethers, and 4-
vinylcyclohexene diepoxide in separate groups. The TCE-treated group received 0.45% TCE in
drinking water containing 3% Tween vehicle; control groups were administered either untreated
water, or water containing the 3% Tween vehicle. There were 5-6 females/group, and three
replicates were conducted for each group. At the end of exposure, ovulation was induced, the
rats were killed, and the ovaries were removed. The zona pellucida was removed from dissected
oocytes, which were then placed into culture medium and inseminated with sperm from
untreated males. TCE treatment did not affect female body weight gain, the percentage of
females ovulating, or the number of oocytes per ovulating female. Fertilizability of the oocytes
from treated females was reduced significantly (46% for TCE-treated females versus 56% for
vehicle controls). Oocytes from TCE-treated females had reduced ability to bind sperm plasma
membrane proteins compared with vehicle controls.
In subsequent studies, Wu and Berger (2007, 2008) examined the effect of TCE on
oocyte fertilizibility and ovarian gene expression. TCE was administered to female Simonson
6/22/2009
768
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
rats (number of subjects not reported) in the drinking water at 0 or 0.45% (in 3% Tween vehicle);
daily doses were estimated to be 0.66 g TCE/kg body weight/day. In the oocyte fertilizibility
study (Wu and Berger, 2007), the female rats were treated on days 1-5, 6-10, 11-14, or 1-14 of
the 2-week period preceding ovulation (on day 15). Oocytes were extracted and fertilized in
vitro with sperm from a single male donor rat. With any duration of TCE exposure, fertilization
(as assessed by the presence of decondensed sperm heads) was significantly (p < 0.05) decreased
as compared to controls. After exposure on days 6-10, 11-14, or 1-14, the oocytes from TCE-
treated females had a significantly decreased ability to bind sperm (p < 0.05) in comparison to
oocytes from vehicle controls. Increased protein carbonyls (an indicator of oxidatively modified
proteins) were detected in the granulosa cells of ovaries from females exposed to TCE for 2
weeks. The presence of oxidized protein was confirmed by Western blot analysis. Microsomal
preparations demonstrated the localization of cytochrome P450 2E1 and glutathione s-transferase
(TCE-metabolizing enzymes) in the ovary. Ovarian mRNA transcription for ALCAM and
Cuzdl protein was not found to be altered after 1 or 5 days of exposure (Wu and Berger, 2008),
suggesting that the post-translational modification of proteins within the ovary may partially
explain the observed reductions in oocyte fertilization.
Studies assessing fertility and reproductive outcomes in both sexes
Assessments of reproduction and fertility with continuous breeding were conducted in
NTP studies in CD-I mice (George et al., 1985) and Fischer 344 rats (George et al., 1986). TCE
was administered to the mice and rats at dietary levels of 0, 0.15, 0.30, or 0.60%, based upon the
results of preliminary 14-day dose-range finding toxicity studies. Actual daily intake levels for
the study in mice were calculated from the results of dietary formulation analyses and body
weight/food consumption data at several time points during study conduct; the most conservative
were from the second week of the continuous breeding study: 0, 52.5, 266.3, and 615.0 mg/kg-
day. No intake calculations were presented for the rat study. In these studies, which were
designed as described by Chapin and Sloane (1996), the continuous breeding phase in F0 adults
consisted of a 7-day pre-mating exposure, 98-day cohabitation period, and 28-day segregation
period. In rats, a crossover mating trial (i.e., control males x control females; 0.60% TCE males
x control females; control males x 0.60% TCE females) was conducted to further elucidate
treatment-related adverse reproductive trends observed in the continuous breeding phase. The
last litter of the continuous breeding phase was raised to sexual maturity for an assessment of
fertility and reproduction in control and high-dose groups; for the rats, this included an open field
behavioral assessment of F1 pups. The study protocol included terminal studies in both
generations, including sperm evaluation (count morphology, and motility) in 10 selected males
per dose level, macroscopic pathology, organ weights, and histopathology of selected organs.
6/22/2009
769
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
In the continuous breeding phase of the CD-I mouse study (George et al., 1985), no
clinical signs of toxicity were observed in the parental (FO) animals, and there were no treatment-
related effects on the proportion of breeding pairs able to produce a litter, the number of live
pups per litter, the percent born live, the proportion of pups born live, the sex of pups born live,
absolute live pup weights, or adjusted female pup weights. At the high dose level of 0.60%, a
number of adverse outcomes were observed. In the parental animals, absolute and body-weight-
adjusted male and female liver weight values were significantly increased (p < 0.01), and right
testis and seminal vesicle weights were decreased (p < 0.05), but kidney/adrenal weights were
not affected. Sperm motility was significantly (p < 0.01) decreased by 45% in treated males as
compared to controls. Histopathology examination revealed lesions in the liver (hypertrophy of
the centrilobular liver cells) and kidneys (tubular degeneration and karyomegaly of the
corticomedullary renal tubular epithelium) of F0 males and females. In the pups at 0.60%,
adjusted live birth weights for males and both sexes combined were significantly decreased
(p < 0.01) as compared to control. The last control and high-dose litters of the continuous
breeding assessment were raised to the age of sexual maturity for a further assessment of
reproductive performance. In these F1 pups, body weights (both sexes) were significantly
decreased at postnatal day (PND) 4, and male offspring body weights were significantly
(p < 0.05) less than controls at PND 74 (± 10). It was reported that perinatal mortality (PND
0-21) was increased, with a 61.3% mortality rate for TCE-treated pups versus a 28.3% mortality
rate for control pups. Reproductive performance was not affected by treatment, and postmortem
evaluations of the F1 adult mice revealed significant findings at 0.60% TCE that were consistent
with those seen in the F0 adults and additionally demonstrated renal toxicity, i.e., elevated liver
and kidney/adrenal weights and hepatic and renal histopathological lesions in both sexes,
elevated testis and epididymis weights in males, and decreased sperm motility (18% less than
control).
The F344 rat study continuous breeding phase demonstrated no evidence of treatment-
related effects on the proportion of breeding pairs able to produce a litter, percent of pups born
alive, the sex of pups born alive, or absolute or adjusted pup weights (George et al., 1986).
However, the number of live pups per litter was significantly (p < 0.05) decreased at 0.30% and
0.60% TCE, and a significant (p < 0.01) trend toward a dose-related decrease in the number of
live litters per pair was observed; individual data were reported to indicate a progressive decrease
in the number of breeding pairs in each treatment group producing third, fourth, and fifth litters.
The crossover mating trial conducted in order to pursue this outcome demonstrated that the
proportion of detected matings was significantly depressed (p < 0.05) in the mating pairs with
TCE-treated partners compared to the control pairs. In the F0 adults at 0.60% TCE, postpartum
dam body weights were significantly decreased (p < 0.01 or 0.05) in the continuous breeding
6/22/2009
770
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
phase and the crossover mating trials, and terminal body weights were significantly decreased
(p < 0.01) for both male and female rats. Postmortem findings for F0 adults in the high dose
group included significantly increased absolute and body-weight-adjusted liver and
kidney/adrenal weights in males, increased adjusted liver and kidney/adrenal weights in females,
and significantly increased adjusted left testis/epididymal weights. Sperm assessment did not
identify any effects on motility, concentration or morphology, and histopathological examination
was negative. The last control and high-dose litters of the continuous breeding assessment were
raised to the age of sexual maturity for assessment of open field behavior and reproductive
performance. In these F1 pups at 0.60% TCE, body weights of male and females were
significantly (p < 0.05 or 0.01, respectively) decreased at PND 4 and 14. By PND 21, pup
weights in both sexes were significantly reduced in all treated groups, and this continued until
termination (approximately PND 80). A tendency toward decreased postweaning survival (i.e.,
from PND 21 to PND 81 ± 10) was reported for F1 pups at the 0.15% and 0.60% levels. Open
field testing revealed a significant (p < 0.05) dose-related trend toward an increase in the time
required for male and female F1 weanling pups to cross the first grid in the testing device,
suggesting an effect on the ability to react to a novel environment. Reproductive performance
assessments conducted in this study phase were not affected by treatment. Postpartum F1 dam
body weights were significantly decreased (p < 0.05 or 0.01) in all of the TCE-treated groups as
compared to controls, as were terminal body weights for both adult F1 males and females.
Postmortem evaluations of the F1 adult rats revealed significantly (p < 0.01) decreased left
testis/epididymis weight at 0.60% TCE, and significantly (p < 0.05 or 0.01) increased adjusted
mean liver weight in all treated groups for males and at 0.30 and 0.60% for females. Sperm
assessments for F1 males revealed a significant increase (p < 0.05) in the percent abnormal
sperm in the 0.30% TCE group, but no other adverse effects on sperm motility, concentration, or
morphology were observed. As with the F0 adults, there were no adverse treatment-related
findings revealed at histopathological assessment. The study authors concluded that the
observed effects to TCE exposure in this study were primarily due to generalized toxicity and not
to a specific effect on the reproductive system; however, based upon the overall toxicological
profile for TCE, which demonstrates that the male reproductive system is a target for TCE
exposures, this conclusion is not supported.
Table 4.7.3. Summary of mammalian in vivo reproductive toxicity studies - oral exposures
Reference
Species/strain/
sex/number
Dose level/
exposure
duration
Route/vehicle
NOAEL;
LOAEL a
Effects
Studies assessing male reproductive outcomes
6/22/2009
771
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
sex/number
Dose level/
exposure
duration
Route/vehicle
NOAEL;
LOAEL a
Effects
DuTeaux et al.,
Rat, Sprague-
0, 0.2%, or 0.4%
Drinking
LOEL: 0.2%
TCE metabolite-protein
2003
Dawley, male,
(0, 143, or 270
water; 3%
adducts formed by a
3/group
mg/kg-day)
ethoxylated
castor oil
vehicle
cytochrome P-450-mediated
pathway were detected by
fluorescence
imunohistochemistry in the
epithelia of corpus epididymis
and in efferent ducts.
DuTeaux et al.,
Rat, Sprague-
0, 0.2%, or 0.4%
Drinking
LOAEL:
Dose-dependent 1 in ability of
2004b
Dawley, male,
(0, 143, or 270
water, 3%
0.2%
sperm to fertilize oocytes
3/group, or
mg/kg-day)
ethoxylated
collected from untreated s.
Simonson
castor oil
Oxidative damage to sperm
albino (UC-
14 days
vehicle
membrane in head and mid-
Davis), male,
piece was indicated by dose-
3/group
related T in oxidized proteins
and lipid peroxidation.
Veeramachaneni
Rabbit, Dutch
9.5 or 28.5 ppm
Drinking
LOAEL: 9.5
Decreased copulatory
et al., 2001
belted, females
and offspring;
7-9 offspring/
group
TCE d
GD 20 thru
lactation, then to
offspring thru
postnatal wk 15
water
ppm
behavior; acrosomal
dysgenesis, nuclear
malformations; sig. 1 LH and
testosterone.
Zenick et al.,
Rat, Long-
0, 10, 100, 1,000
Gavage, corn
NOAEL:
At 1,000 mg/kg, BW
1984
Evans, male,
10/group
mg/kg-day
6 wk, 5 days/wk;
4 wks recovery
oil vehicle
100 mg/kg-
day
LOAEL:
1,000 mg/kg-
day
liver/BW ratiost, and
impaired copulatory behavior.
Copulatory performance
returned to normal by 5th wk
of exposure. At wk 6, TCE
and metabolites concentrated
to a significant extent in male
reproductive organs.
Studies assessing female reproductive outcomes
Berger and
Rat, Simonson
0, 0.45%
Drinking
LOAEL:
In vitro fertilization and sperm
Horner, 2003
(S-D derived),
water, 3%
0.45%
penetration of oocytes sig. 1
female,
2 weeks
Tween
with sperm harvested from
(5-6) x 3
vehicle
untreated males.
/group
6/22/2009
772
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
sex/number
Dose level/
exposure
duration
Route/vehicle
NOAEL;
LOAEL a
Effects
Cosby and
Dukelow, 1992
Mouse,
B6D2F1,
female,
7-12/group
0, 24, 240 mg/kg-
day
GD 1-5, 6-10, or
11-15
Gavage, corn
oil vehicle
NOAEL:
240 mg/kg-
day
No treatment-related effects on
in vitro fertilization in dams or
offspring.
Manson et al.,
1984
Rat, Long-
Evans, female,
23-25/group
0, 10, 100, 1,000
mg/kg-day
6 weeks: 2 wk
premating, 1 wk
mating period,
GD 1-21
Gavage, corn
oil vehicle
NOAEL:
100 mg/kg-
day
LOAEL:
1,000 mg/kg-
day
Female fertility and mating
success was not affected. At
1,000 mg/kg/day group, 5/23
females died, gestation BW
gain was sig. i. After
subchronic oral TCE exposure,
TCE was detected in fat,
adrenals, and ovaries; TCA
levels in uterine tissue were
high.
At 1,000 mg/kg-day, neonatal
deaths (female pups) were T
onPNDs 1, 10, and 14. Dose-
related T seen in TCA in
blood, liver and milk in
stomach of $ pups, not ; s.
Wu and Berger,
2007
Rat, Simonson
(S-D derived),
female, (no.
/group not
reported)
0, 0.45%
(0.66 g/kg-day)b
Pre-ovulation
days 1-5, 6-10,
11-14, or 1-14
Drinking
water, 3%
Tween
vehicle
LOAEL:
0.45%
In vitro fertilization and sperm
penetration of oocytes sig. 1
with sperm harvested from
untreated males.
Wu and Berger,
2008
Rat, Simonson
(S-D derived),
female, (no.
/group not
reported)
0, 0.45%
(0.66 g/kg-day)b
1 or 5 days
Drinking
water, 3%
Tween
vehicle
NOEL:
0.45%
Ovarian mRNA expression for
ALCAM and Cudzl protein
were not altered.
Studies assessing fertility and reproductive outcome in both sexes
6/22/2009
773
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
sex/number
Dose level/
exposure
duration
Route/vehicle
NOAEL;
LOAEL a
Effects
George et al.,
1985
Mouse, CD-I,
male and
female, 20
pairs/treatment
group; 40
controls/sex
0,0.15, 0.30 or
0.60% c
microencapsulated
TCE
(TWA dose
estimates: 0, 173,
362, or 737
mg/kg-day)b
Breeders exposed
1 wk premating,
then for 13 wk;
pregnant females
throughout
gestation (i.e., 18
wk total)
Dietary
Parental
systemic
toxicity:
NOAEL:
0.30%
LOAF,!/
0.60%
At 0.60%, inFO: sig.T liver
weights in both sexes; sig. -1
testis and seminal vesicle
weight; histopathology of liver
and kidney in both sexes.
At 0.60%, inFl: sig. -i- BW on
PND 74, and in postpartum F1
dams; sig. T liver, testis, and
epididymis weights in males,
sig. T kidney weights in both
sexes; sig. -1 testis and seminal
vesicle weight; histopathology
of liver and kidney in both
sexes.
Parental
reproductive
function:
LOAF,!/
0.60% c
At 0.60%, in F0 and F1 males:
sig. 1 sperm motility
Offspring
toxicity:
NOAEL:
0.30%
LOAEL:
0.60%
At 0.60%, inFl pups: sig. i
live birth weights, sig. -1 PND
4 pup BW; perinatal mortality
t (PND 0-21)
6/22/2009
774
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
sex/number
Dose level/
exposure
duration
Route/vehicle
NOAEL;
LOAEL a
Effects
George et al.,
1986
Rat, F334,
males and
female,
20
pairs/treatment
group,
40 controls/sex
0,0.15, 0.30 or
0.60% c
microencapsulated
TCE
Breeders exposed
1 wk pre-mating,
then for 13 wk;
pregnant females
throughout
gestation (i.e., 18
wk total)
Dietary
Parental
systemic
toxicity:
LOAEL:
0.15%
At 0.60%, inFO: sig. -i-
postpartum dam BW; sig. 1
term. B W in both sexes; sig. T
liver, and kidney/adrenal
weights in both sexes; sig. T
testis/epididymis weights; in
Fl: sig. -1 testis weight.
At all doses in Fl: sig. -1
postpartum dam BW; sig.l
term. B W in both sexes, sig. T
liver wt. in both sexes.
At 0.30% and 0.60%, in Fl:
sig. T liver wt. in females.
Parental
reproductive
function:
LOAEL:
0.60%c
At 0.60%, sig 1 mating in F0
males and females (in cross-
over mating trials).
Offspring
toxicity:
LOAEL:
0.15%
At 0.60%, sig. ^ Fl BW on
PND 4 and 14.
At all doses, sig. -1 Fl BW on
PND 21 and 80.
At 0.3% and 0.60%, sig. 1 live
Fl pups/litter.
At 0.15% and 0.60%, trend
toward 1 F1 survival from
PND 21 to PND 80.
11 NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level), NOEL (No
Observed Effect Level), and LOEL (Lowest Observed Effect Level) are based upon reported study findings.
b Dose conversion calculations by study author(s).
0 Fertility and reproduction assessment of last litter from continuous breeding phase and cross-over mating
assessment (rats only) were conducted for 0 or 0.60% dose groups only.
d Concurrent exposure to several ground water contaminants; values given are for TCE levels in the mixture.
6/22/2009
775
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.7.1.3 Discussion/synthesis of non-cancer reproductive toxicity findings
The human epidemiological findings and animal study evidence consistently indicate that
TCE exposures can result in adverse reproductive outcomes. Although the epidemiological data
may not always be robust or unequivocal, they demonstrate the potential for a wide range of
exposure-related adverse outcomes on female and male reproduction. In animal studies, there is
some evidence for female-specific reproductive toxicity; but there is strong and compelling
evidence for adverse effects of TCE exposure on male reproductive system and function.
4.7.1.3.1 Female reproductive toxicity
Although few epidemiological studies have examined TCE exposure in relation to female
reproductive function (Table 4.7-4), the available studies provide evidence of decreased fertility,
as measured by time to pregnancy (Sallmen et al., 1995), and effects on menstrual cycle patterns,
including abnormal cycle length (ATSDR, 2001), amenorrhea (Sagawa et al., 1973; Zielinski,
1973), and menstrual "disturbance" (Bardodej and Vyskocil, 1956). In experimental animals, the
effects on female reproduction include evidence of reduced in vitro oocyte fertilizability in rats
(Berger and Horner, 2003; Wu and Berger, 2007). However, in other studies that assessed
reproductive outcome in female rodents (Cosby and Dukelow, 1992; George et al., 1985, 1986;
Manson et al., 1984), there was no evidence of adverse effects of TCE exposure on female
reproductive function. Overall, although the data are suggestive, there are inadequate data to
make conclusions as to whether adverse effects on human female reproduction are caused by
TCE.
Table 4.7.4. Summary of adverse female reproductive outcomes associated with TCE
exposures
Finding
Species
Citation
Menstrual cycle disturbance
Human
ATSDR, 2001 a
Bardodej and Vyskocil, 1956
Sagawa et al., 1973
Zielinski, 1973
Reduced fertility
Human a
Sallmen et al., 1995
Ratb
Berger and Horner, 2003
Wu and Berger, 2007
a Not significant.
b In vitro oocyte fertilizability.
6/22/2009
776
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.7.1.3.2 Male reproductive toxicity
Notably, the results of a number of studies in both humans and experimental animals
have suggested that exposure to TCE can result in targeted male reproductive toxicity (Table 4.7-
5). The adverse effects that have been observed in both male humans and male animal models
include altered sperm count, morphology, or motility (Chia et al., 1996; George et al., 1985;
Kumar et al, 2000a, b, 2001; Land et al., 1981; Rasmussen et al., 1988; Veeramachaneni et al.,
2001); decreased libido or copulatory behavior (Bardodej and Vyskocil, 1956; El Ghawabi et al.,
1973; George et al., 1986; Saihan et al., 1978; Veeramachaneni et al., 2001; Zenick et al., 1984);
alterations in serum hormone levels (Chia et al., 1997; Goh et al., 1998; Kumar et al., 2000b;
Veeramachaneni et al., 2001); and reduced fertility (George et al., 1986). However, other studies
in humans did not see evidence of altered sperm count or morphology (Rasmussen et al., 1988)
or reduced fertility (Forkert et al., 2003; Sallmen et al., 1998), and some animal studies also did
not identify altered sperm measures (Cosby and Dukelow, 1992; Xu et al., 2004; Zenick et al.,
1984; George et al, 1986). Additional adverse effects observed in animals include
histopathological lesions of the testes (George et al., 1986; Kumar et al., 2000a, 2001) or
epidiymides (Forkert et al., 2002; Kan et al., 2007) and altered in vitro sperm-oocyte binding
and/or in vivo fertilization for TCE and/or its metabolites (Xu et al., 2004; DuTeaux et al.,
2004b).
In spite of the preponderance of studies demonstrating effects on sperm parameters, there
is an absence of overwhelming evidence in the database of adverse effects of TCE on overall
fertility in the rodent studies. That is not surprising, however, given the redundancy and
efficiency of rodent reproductive capabilities. Nevertheless, the continuous breeding
reproductive toxicity study in rats (George et al., 1986) did demonstrate a trend towards
reproductive compromise (i.e., a progressive decrease in the number of breeding pairs producing
third, fourth, and fifth litters).
It is noted that in the studies by George et al. (1985, 1986), adverse reproductive
outcomes in male rats and mice were observed at the highest dose level tested (0.060% TCE in
diet) which was also systemically toxic (i.e., demonstrating kidney toxicity and liver enzyme
induction and toxicity, sometimes in conjunction with body weight deficits). Because of this, the
study authors concluded that the observed reproductive toxicity was a secondary effect of
generalized systemic toxicity; however, this conclusion is not supported by the overall
toxicological profile of TCE which provides significant evidence indicating that TCE is a
reproductive toxicant.
Table 4.7.5. Summary of adverse male reproductive outcomes associated with TCE
exposures
Finding Species Citation
6/22/2009
777
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Testicular toxi city/pathology
Rat
George et al., 1986
Kumar et al., 2000a
Kumar et al., 2001
Mouse
Kan et al., 2007
Epididymal toxicity/pathology
Mouse
Forkert et al., 2002
Decreased sperm quantity/quality
Human
Chia et al., 1996
Rasmussen et al., 1988a
Rat
Kumar et al., 2000a, b, 2001
Mouse
George et al., 1985
Land et al., 1981
Rabbit
Veeramachaneni et al., 2001
Altered in vitro sperm-oocyte binding or in vivo
fertilization
Rat
DuTeaux et al., 2004b
Mouse
Cosby and Dukelow, 1992 b
Xu et al., 2004 b
Altered sexual drive or function
Human
El Ghawabi et al., 1973
Saihan et al., 1978 c
Bardodej and Vyskocil, 1956
Rat
George et al., 1986
Zenick et al., 1984
Rabbit
Veeramachaneni et al., 2001
Altered serum testosterone levels
Human
Chia et al., 1997 d
Gohetal., 1998 e
Rat
Kumar et al., 2000b
Rabbit
Veeramachaneni et al., 2001
Reduced fertility
Rat
George et al., 1986
Gynaecomastia
Human
Saihan et al., 1978 c
a Non-significant increase in percentage of two fluorescent Y-bodies (YFF) in spermatozoa; no effect on sperm
count or morphology.
b Observed with metabolite(s) of TCE only.
0 Case study of one individual.
d Also observed altered levels of dihydroepiandrosterone (DHEAS), follicle stimulating hormone (FSH), and sex-
hormone binding globulin (SHBG).
e Also observed altered levels of SHBG.
6/22/2009
778
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.7.1.3.2.1 The role of metabolism in male reproductive toxicity
There has been particular focus on evidence of exposure to male reproductive organs by
TCE and/or its metabolites, as well as the role of TCE metabolites in the observed toxic effects.
In humans, a few studies demonstrating male reproductive toxicity have measured levels
of TCE in the body. Urine TCA was measured in men employed in an electronics factory, and
adverse effects observed included abnormal sperm morphology and hyperzoospermia and altered
serum hormone levels (Chia et al., 1996, 1997; Goh et al., 1998). Urine TCA was also measured
as a marker of exposure to TCE in men occupationally exposed to solvents, although this study
did not report any adverse effects on fertility (Sallmen et al., 1998).
In the study in Long-Evans male rats by Zenick et al. (1984), blood and tissue levels of
TCE, trichloroacetic acid (TCA) and trichloroethanol (TCOH) were measured in three rats/group
following 6 weeks of gavage treatment at 0, 10, 100, and 1,000 mg/kg-day. Additionally the
levels of TCE and metabolites were measured in seminal plugs recovered following copulation at
Week 5. Marked increases in TCE levels were observed only at 1,000 mg/kg-day, in blood,
muscle, adrenals, and seminal plugs. It was reported that dose-related increases in TCA and
TCOH concentrations were observed in the organs evaluated, notably including the reproductive
organs (epididymis, vas deferens, testis, prostate, and seminal vesicle), thus creating a potential
for interference with reproductive function.
This potential was explored further in a study by Forkert et al. (2002), in which male CD-
1 mice were exposed by inhalation to 1,000 ppm TCE (6 hours/day, 5 day/week) for 4
consecutive weeks. Urine was obtained on days 4, 9, 14, and 19 of exposure and analyzed for
concentrations of TCE and TCOH. Microsomal preparations from the liver, testis and
epididymis were used for immunoblotting, determining p-nitrophenol hydroxylase and CYP2E1
activities, and evaluating the microsomal metabolism of TCE.
Subsequent studies conducted by the same laboratory (Forkert et al., 2003) evaluated the
potential of the male reproductive tract to accumulate TCE and its metabolites including chloral,
trichloroethanol (TCOH), trichloroacetic acid (TCA) and dichloroacetic acid (DCA). Human
seminal fluid and urine samples from eight mechanics diagnosed with clinical infertility and
exposed to TCE occupationally were analyzed. Urine samples from two of the eight subjects
contained TCA and/or TCOH, suggesting that TCE exposure and/or metabolism was low during
the time just prior to sample collection. TCE, chloral, and TCOH were detected in seminal fluid
samples from all eight subjects, while TCA was found in one subject, and DCA was found in
two subjects. Additionally, TCE and its metabolites were assessed in the epididymis and testis
of CD-I mice (4/group) exposed by inhalation (6 hours/day, 5 days/week) to 1,000 ppm TCE for
1, 2 and 4 weeks. TCE, chloral and TCOH were found in the epididymis at all timepoints,
although TCOH levels were increased significantly (tripled) at four weeks of exposure. This
6/22/2009
779
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
study showed that the metabolic disposition of TCE in humans is similar to that in mice,
indicating that the murine model is appropriate for investigating the effects of TCE-induced
toxicity in the male reproductive system. These studies provide support for the premise that TCE
is metabolized in the human reproductive tract, mainly in the epididymis, resulting in the
production of metabolites that cause damage to the epididymal epithelium and affect the normal
development of sperm.
Immunohistochemical experiments (Forkert et al., 2002) confirmed the presence of
CYP2E1 in the epididymis and testis of mice; it was found to be localized in the testicular
Leydig cells and the epididymal epithelium. Similar results were obtained with the
immunohistochemical evaluation of human and primate tissue samples. CYP2E1 has been
previously shown by Lipscomb et al. (1998) to be the predominant P450 enzyme catalyzing the
hepatic metabolism of TCE in both animals and rodents. These findings support the role of
CYP2E1 in TCE metabolism in the male reproductive tract of humans, primates, and mice.
4.7.1.3.2.2 Mode of action for male reproductive toxicity
A number of studies have been conducted to attempt to characterize various aspects of
the mode of action for observed male reproductive outcomes.
Studies by Kumar et al. (2000b, 2001) suggest that perturbation of testosterone
biosynthesis may have some role in testicular toxicity and altered sperm measures. Significant
decreases in the activity of G6PDH and accumulation of cholesterol are suggestive of an
alteration in testicular steroid biosynthesis. Increased testicular lipids, including cholesterol,
have been noted for other testicular toxicants such as lead (Saxena et al., 1987),
triethylenemelamine (Johnson et al., 1967), and quinalphos (Ray et al., 1987), in association with
testicular degeneration and impaired spermatogenesis. Since testosterone has been shown to be
essential for the progression of spermatogenesis (O'Donnell et al., 1994), alterations in
testosterone production could be a key event in male reproductive dysfunction following TCE
exposure. Additionally, the observed TCE-related reduction of 17-P-HSD, which is involved in
the conversion of androstenedione to testosterone, has also been associated with male
reproductive insufficiency following exposure to phthalate esters (Srivastava and Srivastava,
1991), quinalphos (Ray et al., 1987), and lead (Saxena et al., 1987). Reductions in SDH, which
are primarily associated with the pachytene spermatocyte maturation of germinal epithelium,
have been shown to be associated with depletion of germ cells (Mills and Means, 1970; Chapin
et al., 1982), and the activity of G6-PDH is greatest in premeiotic germ cells and Leydig cells of
the interstitium (Blackshaw et al., 1970). The increased GT and glucuronidase observed
following TCE exposures appear to be indicative of impaired Sertoli cell function (Hodgen and
6/22/2009
780
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Sherins, 1973; Sherins and Hodgen, 1976). Based upon the conclusions of these studies, Kumar
et al. (2001) hypothesized that the reduced activity of G6-PDH and SDH in testes of TCE-
exposed male rats is indicative of the depletion of germ cells, spermatogenic arrest, and impaired
function of the Sertoli cells and Ley dig cells of the interstitium.
In the series of experiments by DuTeaux et al. (2003, 2004b), protein dichloroacetyl
adducts were found in the corpus epididymis and in the efferent ducts of rats administered TCE;
this effect was also demonstrated following in vitro exposure of reproductive tissues to TCE.
Oxidized proteins were detected on the surface of spermatozoa from TCE-treated rats in a dose-
response pattern; this was confirmed using a Western blotting technique. Soluble (but not
mitochondrial) cysteine-conjugate P-lyase was detected in the epididymis and efferent ducts of
treated rats. Following a single intraperitoneal injection of dichlorovinyl cysteine (DCVC), no
dichloroacetylated protein adducts were detected in the epididymis and efferent ducts. The
presence of CYP2E1 was found in epididymis and efferent ducts, suggesting a role of
cytochrome P450-dependent metabolism in adduct formation. An in vitro assay was used to
demonstrate that epididymal and efferent duct microsomes are capable of metabolizing TCE;
TCE metabolism in the efferent ducts was found to be inhibited by anti-CYP2El antibody. Lipid
peroxidation in sperm, presumably initiated by free radicals, was increased in a significant
(p < 0.005) dose-dependent manner after TCE-exposure.
Overall, it has been suggested (DuTeaux et al., 2004b) that reproductive organ toxicities
observed following TCE exposure are initiated by metabolic bioactivation, leading to subsequent
protein adduct formation. It has been hypothesized that epoxide hydrolases in the rat epididymis
may play a role in the biological activation of metabolites (DuTeaux et al., 2004a).
4.7.1.3.3 Summary of non-cancer reproductive toxicity
The toxicological database for TCE includes a number of studies that demonstrate
adverse effects on the integrity and function of the reproductive system in females and males.
Both the epidemiological and animal toxicology databases provide suggestive, but limited,
evidence of adverse outcomes to female reproductive outcomes. However, much more extensive
evidence exists in support of an association between TCE exposures and male reproductive
toxicity. The available epidemiological data and case reports that associate TCE with adverse
effects on male reproductive function are limited in size and provide little quantitative dose data
(Lamb and Hentz, 2006). However, the animal data provide extensive evidence of TCE-related
male reproductive toxicity. Strengths of the database include the presence of both functional and
structural outcomes, similarities in adverse treatment-related effects observed in multiple species,
and evidence that metabolism of TCE in male reproductive tract tissues is associated with
adverse effects on sperm measures in both humans and animals (suggesting that the murine
6/22/2009
781
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
model is appropriate for extrapolation to human health risk assessment). Additionally some
aspects of a putative MOA (e.g., perturbations in testosterone biosynthesis) appear to have some
commonalities between humans and animals.
4.7.2 Cancers of the reproductive system
The effects of TCE on cancers of the reproductive system have been examined for males
and females in both epidemiological and experimental animal studies. The epidemiological
literature includes data on prostate in males and cancers of the breast and cervix in females. The
experimental animal literature includes data on prostate and testes in male rodents; and uterus,
ovary, mammary gland, vulva, and genital tract in female rodents. The evidence for these
cancers is generally not robust.
4.7.2.1 Human data
The epidemiologic evidence on TCE and cancer of the prostate, breast, and cervix is from
cohort and geographic based studies. Two additional case-control studies of prostate cancer in
males are nested within cohorts (Greenland et al., 1994; Krishnadasan et al., 2007). The nested
case-control studies are identified in the tables below with cohort studies given their source
population for case and control identification. One population-based case-control study
examined on TCE exposure and prostate (Siemiatycki, 1991); however, no population case-
control studies on breast or cervical cancers and TCE exposure were found in the peer-reviewed
literature.
4.7.2.1.1 Prostate Cancer
Fourteen cohort, 2 nested case-control, one population case-control, and 2 geographic
based studies present relative risk estimates for prostate cancer (Wilcosky et al., 1984; Garabrant
et al., 1988; Axelson et al., 1994; Siemiatycki, 1991; Greenland et al., 1994; Anttila et al., 1995;
Blair et al., 1998; Morgan et al., 1998; Boice et al., 1999, 2006; Ritz, 1999; Hansen et al., 2001;
Morgan and Cassady, 2002; Raaschou-Nielsen et al., 2003; Chang et al., 2003, 2005; ATSDR,
2004, 2006; Krishnadasan et al., 2007; Radican et al. 2008). Three small cohort studies (Costa et
al., 1989; Sinks et al., 1992; Henschler et al., 1995), one multiple-site population case-control
(Siemiatycki, 1991) and one geographic based study (Vartiainen et al., 1993) do not report
estimates for prostate cancer in their published papers. Twelve of the 19 studies with prostate
cancer relative risk estimates had high likelihood of TCE exposure in individual study subjects
6/22/2009
782
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
and were judged to have met, to a sufficient degree, the standards of epidemiologic design and
analysis (Siemiatycki, 1991; Axelson et al., 1994; Anttila et al., 1994; Greenland et al., 1994,
Blair et al., 1998; Morgan et al., 1998, 2000; Boice et al., 1999, 2006; Hansen et al., 2001;
Raaschou-Nielsen et al., 2003; Krishnadasan et al., 2007; Radican et al., 2008). Krishnadasan et
al. (2007) in their nested case-control study of prostate cancer observed a 2-fold odds ratio
estimate with high cumulative TCE exposure score (2.4, 95% CI: 1.3, 4.4, 20 year lagged
exposure) and an increasing positive relationship between prostate cancer incidence and TCE
cumulative exposure score (p = 0.02). TCE exposure was positively correlated with several
other occupational exposures, and Krishnadasan et al. (2007) adjusted for possible confounding
from all other chemical exposures as well as age at diagnosis, occupational physical activity, and
socio-economic status in statistical analyses. Relative risk estimates in studies other than
Krishnadasan et al. (2007) were above 1.0 for overall TCE exposure [1.8, 95% CI: 0.8, 4.0
(Siemiatycki, 1991); 1.1, 95% CI: 0.6, 1.8 (Blair et al., 1998) and 1.20, 95% CI: 0.92, 1.76, with
an additional 10-year follow-up (Radican et al., 2008); 1.58, 95% CI: 0.96, 2.62 (Morgan et al.,
1998, 2000; Environmental Health Strategies, 1997); 1.3, 95% CI: 0.52, 2.69 (Boice et al.,
1999); 1.38, 95% CI: 0.73, 2.35 (Anttila et al., 1995)] and prostate cancer risks did not appear to
increase with increasing exposure. Four studies observed relative risk estimates below 1.0 for
overall TCE exposure [0.93, 95% CI: 0.60, 1.37 (Garabrant et al., 1988); 0.6, 95% CI: 0.2, 1.30
(Hansen et al., 2001); 0.9, 95% CI: 0.79, 1.08 (Raaschou-Nielsen et al., 2003); 0.82, 95% CI:
0.36, 1.62 (Boice et al., 2006)], and are not considered inconsistent because alternative
explanations are possible and included observations are based on few subjects, lowering
statistical power, or to poorer exposure assessment approaches that may result in a higher
likelihood of exposure misclassification.
Five other cohort and geographic based studies were given less weight in the analysis
because of their lesser likelihood of TCE exposure and other study design limitations that would
decrease statistical power and study sensitivity (Wilcosky et al., 1984; Morgan and Cassady,
2002; ATSDR 2004, 2006; Chang et al., 2005). Chang et al. (2005) observed a statistically
significant deficit in prostate cancer risk, based on one case, and an insensitive exposure
assessment (0.14, 95% CI: 0.00, 0.76). Relative risks in the other five studies ranged from 0.62
(CI not presented in paper) (Wilcosky et al., 1984) to 1.11 (95% CI: 0.98, 1.25) (Morgan and
Cassady, 2002).
Risk factors for prostate cancer include age, family history of prostate cancer, and
ethnicity as causal with inadequate evidence for a relationship with smoking or alcohol (Wigle et
al., 2008). All studies except Krishnadasan et al. (2007) were not able to adjust for possible
confounding from other chemical exposures in the work environment. None of the studies
including Krishnadasan et al. (2007) accounted for other well-established non-occupational risk
6/22/2009
783
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
factors for prostate cancer such as race, prostate cancer screening and family history. There is
limited evidence that physical activity may provide a protective effect for prostate cancer (Wigle
et al., 2008). Krishnadasan et al. (2008) examined the effect of physical activity in the
Rocketdyne aerospace cohort (Zhao et al., 2005; Krishnadasan et al., 2007). Their finding of a
protective effects with high physical activity [0.55, 95% CI: 0.32, 0.95, p trend = 0.04] after
control for TCE exposure provides additional evidence (Krishnadasan et al., 2008) and suggests
underlying risk may be obscured in studies lacking adjustment for physical activity.
Table 4.7.6. Summary of human studies on TCE exposure and prostate cancer
Studies Exposure Group
Cohort Studies - Incidence
Aerospace workers (Rocketdyne)
Low/Moderate TCE score
High TCE score Med TCE score
p for trend
Relative Risk
(95% CI)
1.3 (0.81, 2.1)
2.1 (1.2, 3.9)1'2
0.02
No. obs.
events
90
45
Reference
Krishnadasan et al., 2007
Low/Moderate TCE score 1.3 (0.81, 2.1)1,3
High TCE score Med TCE score 2.4 (1.3, 4.4)1'3
p for trend 0.01
All employees at electronics factory (Taiwan) 0.14 (0.00, 0.76)4 1 Chang et al., 2005
Danish blue-collar worker w/TCE exposure Raaschou-Nielsen et al., 2003
Any exposure 0.9(0.79,1.08) 163
Biologically-monitored Danish workers
Any TCE exposure, females
0.6 (0,2, 1.3)
Hansen etal., 2001
Aircraft maintenance workers (Hill Air Force Base, Utah)
TCE Subcohort Not reported 158
Cumulative exp
0 1.05
< 5 ppm-yr 1.1(0.7,1.6) 64
5-25 ppm-yr 1.0(0.6,1.6) 38
>25 ppm-yr 1.2(0.8,1.8) 56
Blair etal., 1998
TCE Subcohort
Cumulative exp
1.2 (0.92, 1.76)
116
Radican et al. 2008
6/22/2009
784
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
0 1.05
< 5 ppm-yr 1.03
5-25 ppm-yr 1.33
>25 ppm-yr 1.31
(0.65, 1.62)
41
(0.82,2.15)
42
(0.84, 2.06)
43
Biologically-monitored Finnish workers 1.38 (0.73,2.35) 13
Mean air-TCE (Ikeda extrapolation
<6 ppm 1.43 (0.62,2.82) 8
6+ ppm 0.68 (0.08, 2.44) 2
Anttila et al., 1995
Cardboard manufacturing workers in Arnsburg, Germany
Exposed workers Not reported
Henschleretal., 1995
Biologically-monitored Swedish workers
Cardboard manufacturing workers, Atlanta area, GA
1.25 (0.84, 1.84) 26
Not reported
Axelson et al., 1994
Sinks et al., 1992
Cohort-Mortality
Aerospace workers (Rocketdyne)
Any TCE (utility/eng flush)
0.82 (0.36, 1.62)
Boice et al., 2006
View-Master employees
All employees at electronics factory (Taiwan)
Fernald workers
Any TCE exposure
Light TCE exposure, >2 years duration
Mod TCE exposure, >2 years duration
1.69 (0.68, 3.48)6
Not reported
Not reported
0.91 (0.38, 2.18)5
1.44 (0.19, 11.4)
10
ATSDR, 2004
Chang et al., 2003
Ritz, 1999
Aerospace workers (Lockheed)
Routine Exposure to TCE
Routine-Intermittent1
1.31 (0.52, 2.69)
Not reported
Boice et al., 1999
Aerospace workers (Hughes) Morgan et al., 1998, 2000
TCE Subcohort 1.18 (0.73,1.80) 21
Low Intensity (<50 ppm) 1.03 (0.51,1.84) 7
High Intensity (>50 ppm) 0.47(0.15,1.11) 14
TCE Subcohort (Cox Analysis)
,5
Never exposed 1.00
Ever exposed 1.58 (0.96, 2.62)8
6/22/2009
785
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Peak
Med/Hi 1.39 (0.80, 2.41)8
No/Low 1.005
Cumulative
Referent 1.005
Low 1.72 (0.78, 3.80)8
High 1.53 (0.85, 2.75)8
Aircraft maintenance workers (Hill Air Force Base, Utah) Blair et al., 1998
TCE Subcohort
Cumulative exp
0
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
1.1 (0.6, 1.8)
54
1.05
0.9 (0.5, 1.8)
19
1.0(0.5,2.1)
13
1.3 (0.7, 2.4)
22
Cardboard manufacturing workers in Arnsburg, Germany
TCE exposed workers
Deaths reported to among GE pension fund (Pittsfield, MA
Cardboard manufacturing workers, Atlanta area, GA
Aircraft manufacturing plant employees (Italy)
All subjects
Aircraft manufacturing plant employees (San Diego, CA)
Rubber workers
Any TCE exposure
Not reported Henschleret al., 1995
0.82 (0.46, 1.46)1 58 Greenland et al., 1994
Not reported 0 Sinks et al., 1992
Costa et al., 1989
Not reported
0.93 (0.60, 1.37) 25 Garabrant et al., 1988
Wilcosky et al., 1984
0.62 (not
reported) 3
Case-control Studies
Population of Montreal, Canada
Any TCE exposure
Substantial TCE exposure
Siemiatycki, 1991
1.1 (0.6,2.1)
1.8 (0.8, 4.0)'
11
7
Geographic Based Studies
Residents in two study areas inEndicott, NY 1.05 (0.75, 1.43) 40 ATSDR, 2006
Residents of 13 census tracts inRedlands, CA 1.11 (0.98, 1.25)10 483 Morgan and Cassady, 2002
6/22/2009
786
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Finnish residents
Vartiainenetal., 1993
Residents of Hausjarvi
Residents of Huttula
Not reported
Not reported
1 Odds ratio from nested case-control study
2 Odds ratio, zero lag
3 Odds ratio, 20 year lag
4 Chang et al. (2005) presents standardize incidence ratio (SIR) for a category site of all cancers of male genital
organs
internal referents, workers without TCE exposure
6 Proportional mortality ratio
7 Analysis for >2 years exposure duration and a lagged TCE exposure period of 15 years
8 Risk ratio from Cox Proportional Hazard Analysis, stratified by age and sex, from Environmental Health
Strategies (1997) Final Report to Hughes Corporation (Communication from Paul A. Cammer, President,
Trichloroethylene Issues Group to Cheryl Siegel Scott, U.S. EPA, December 22, 1997).
9 90% Confidence Interval
10 99% Confidence Interval
4.7.2.1.2 Breast Cancer
Thirteen studies of TCE exposure reported findings on breast cancer in males and
females combined (Garabrant et al., 1988; Greenland et al., 1994; Boice et al., 1999), in males
and females, separately (Hansen et al., 2001; Raaschou-Nielsen et al., 2003; ATSDR, 2004;
Clapp and Hoffman, 2008), or in females only (Blair et al., 1998; Morgan et al., 1998; ATSDR,
2006; Change et al., 2005; Sung et al., 2007; Radican et al., 2008). Six studies have high
likelihood of TCE exposure in individual study subjects and met, to a sufficient degree, the
standards of epidemiologic design and analysis (Blair et al., 1998; Morgan et al., 1998; Boice et
al., 1999; Hansen et al., 2001; Raaschou-Nielsen et al., 2003; Radican et al. 2008). Four other
high-quality studies identified in a systematic review with risk estimates for other cancer sites do
not report risk estimates for breast cancer (Siemiatycki, 1991; Axelson et al., 1994; Anttila et al.,
1995; Boice et al., 2006). No case-control studies were found on TCE exposure, although
several studies examine occupational title or organic solvent as a class (Weiderpass et al., 1999;
Band et al., 2000; Rennix et al., 2005; Ji et al., 2008). While association is seen with
occupational title or industry and breast cancer [employment in aircraft and aircraft part industry,
2.48, 95% CI: 1.14, 5.39 (Band et al., 2000); solvent user: 1.48, 95% CI: 1.03, 2.12 (Rennix et
al., 2005)], TCE exposure is not uniquely identified and greatly limits the study's use for
informing TCE exposure and breast cancer examinations.
Relative risk estimates in the five high-quality studies ranged from 0.75 (0.43, 1.22)
[females and males] (Morgan et al., 1998) to 2.0 (0.9, 4.6) [mortality in females] (Blair et al.,
6/22/2009
787
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1998). Blair et al. (1998), additionally, observed stronger risk estimates for breast cancer
mortality among females with low level intermittent [3.1, 95% CI: 1.5, 6.2] and low level
continuous [3.4, 95% CI: 1.4, 8.0] TCE exposures, but not with frequent peaks [1.4, 95% CI: 0.7,
3.2). A similar pattern of risks was also observed by Radican et al. (2008) who studied mortality
in this cohort and adding 10 years of follow-up, although the magnitude of breast cancer risk in
females was lower than that observed in Blair et al. (1998). Risk estimates did not appear to
increase with increasing cumulative exposure in the two studies that included exposure-response
analyses (Blair et al., 1998; Morgan et al., 1998). None of the five high quality studies reported
a statistically significant deficit in breast cancer and confidence intervals on relative risks
estimates included 1.0 [no risk]. Few female subjects in these studies appear to have high TCE
exposure. For example, Blair et al. (1998) identified 8 of the 28 breast cancer deaths and 3 of the
34 breast cancer cases with high cumulative exposure.
Relative risk estimates in five studies of lower likelihood TCE exposure and other design
deficiencies ranged from 0.81 (95% CI: 0.52, 1.48) (Garabrant et al., 1988) to 1.19 (1.03, 1.36)
(Chang et al., 2005). These studies lack a quantitative surrogate for TCE exposure to individual
subjects and instead classify all subjects as "potentially exposed", with resulting large dilution of
actual risk and decreased sensitivity (Garabrant et al., 1988; Morgan and Cassady, 2002; Chang
et al., 2005; AT SDR, 2006; NRC, 2006; Sung et al., 2007).
Four studies reported on male breast cancer separately (Hansen et al., 2001; Raaschou-
Nielsen et al., 2003; ATSDR, 2004; Clapp and Hoffman, 2008) and a total of three cases were
observed. Breast cancer in men is a rare disease and are best studied using a case-control
approach (Weiss et al., 2005). Further assessment of TCE exposure and male breast cancer is
warranted.
Overall, the epidemiologic studies on TCE exposure and breast cancer are quite limited in
statistical power; observations are based on few breast cancer cases in high-quality studies or on
inferior TCE exposure assessment in studies with large numbers of observed cases.
Additionally, adjustment for non-occupational breast cancer risk factors is less likely in cohort
and geographic based studies given their use of employment and public records. Breast cancer
mortality observations in Blair et al. (1998) and further follow-up of this cohort by Radican et al.
(2008) of an elevated risk with overall TCE exposure, particulalry low level intermittent and
continuous TCE exposure, provide evidence of an association with TCE. No other high-quality
study reported a statistically significant association with breast cancer, although few observed
cases leading to lower statistical power or examination of risk for males and females combined
are alternative explanations for the null observations in these studies. Both Chang et al. (2005)
and Sung et al. (2007), two overlapping studies of female electronics workers exposed to TCE,
perchloroethylene, and mixed solvents, reported association with breast cancer incidence, with
6/22/2009
788
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
breast cancer risk in Chang et al. (2005) appearing to increase with employment duration. Both
studies support Blair et al. (1998) and Radican et al. (2008), although the lack of exposure
assessment is an uncertainty. The epidemiologic evidence is limited for examining TCE and
breast cancer, and while these studies do not provide any strong evidence for association with
TCE exposure they in turn do not provide evidence of an absence of association.
Table 4.7.7. Summary of human studies on TCE exposure and breast cancer
Studies Exposure Group
Cohort Studies - Incidence
Aerospace workers (Rocketdyne)
Any TCE exposure
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
Relative Risk
(95% CI)
Not reported
No. obs.
events Reference
Zhao et al., 2005
All employees at electronics factory (Taiwan)
Females
Females
1.09 (0.96, 1.22)1 286
1.19 (1.03,1.36) 215
Sung et al., 2007
Chang et al., 2005
Danish blue-collar worker w/TCE exposure
Any exposure, males
Any exposure, females
0.5 (0.06, 1.90)
1.1 (0.89, 1.24)
2
145
Raaschou-Nielsen et al., 2003
Biologically-monitored Danish workers
Any TCE exposure, males
Any TCE exposure, females
0.9 (0.2, 2.3)
0(0.2
exp)
4
Hansen etal., 2001
Aircraft maintenance workers (Hill Air Force Base, Utah)
TCE Subcohort Not reported
Females, Cumulative exp
0 1.02
< 5 ppm-yr 0.3 (0.1,1.4)
5-25 ppm-yr 0.4 (0.1,2.9)
>25 ppm-yr 0.4 (0.4,1.2)
Blair etal., 1998
34
20
11
3
Biologically-monitored Finnish workers
Not reported
Anttila et al., 1995
Cardboard manufacturing workers in Arnsburg, Germany
Henschleretal., 1995
6/22/2009
789
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Exposed workers
Biologically-monitored Swedish workers
Not reported
Not reported
Axelson et al., 1994
Cardboard manufacturing workers, Atlanta area, GA
Not reported
Sinks et al., 1992
Cohort-Mortality
Aerospace workers (Rocketdyne)
Any TCE (utility/eng flush)
Not reported
Boice et al., 2006
Any exposure to TCE
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
Not reported
Not reported
Not reported
Not reported
Zhao et al., 2005
View-Master employees
Males
Females
0 (0.05
exp)
1.02 (0.67, 1.49) 27
ATSDR, 2004
Fernald workers
Any TCE exposure Not reported
Light TCE exposure, >2 years duration Not reported
Mod TCE exposure, >2 years duration Not reported
Ritz, 1999
Aerospace workers (Lockheed)
Routine Exposure to TCE
Routine-Intermittent1
1.31 (0.52, 2.69)
Not reported
Boice et al., 1999
Aerospace workers (Hughes)
TCE Subcohort
Morgan etal., 1998
0.75 (0.43, 1.22)
Low Intensity (<50 ppm) 1.03 (0.51, 1.84)4
High Intensity (>50 ppm) 0.47 (0.15, l.ll)4
TCE Subcohort (Cox Analysis)
Never exposed 1.00
0.94 (0.51, 1.75)
Ever exposed
Peak
No/Low 1.00
1.14(0.48,2.70)
Med/Hi 4,5
16
11
5
Not
reported
Not
reported
Not
reported
6/22/2009
790
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Cumulative
Referent
1.002
1.20 (0.60, 2.40)
Not
Low
4,5
reported
0.65 (0.25, 1.69)
Not
High
4,5
reported
lance workers (Hill Air Force Base, Utah)
TCE Subcohort (females)
2.0 (0.9, 4.6)
20
Females, Cumulative exp
0
1.02
< 5 ppm-yr
2.4(1.1,5.2)
10
5-25 ppm-yr
1.2(0.3,5.4)
21
>25 ppm-yr
1.4 (0.6, 3.2)
8
Low level intermittent exposure
3.1 (1.5,6.2)
15
Low level continuous exposure
3.4(1.4, 8.0)
8
Frequent peaks
1.4 (0.7, 3.2)
10
TCE Subcohort (females)
1.23 (0.73, 2.06)
26
Females, Cumulative exp
0
1.02
< 5 ppm-yr
1.57 (0.81,3.04)
12
5-25 ppm-yr
1.01 (0.31,3.30)
3
>25 ppm-yr
1.05 (0.53,2.07)
11
Low level intermittent exposure
1.92 (1.08,3.43)
18
Low level continuous exposure
1.71 (0.79, 3.71)
8
Frequent peaks
1.08 (0.57, 2.02)
14
Blair etal., 1998
Radican et al. (2008)
Cardboard manufacturing workers in Arnsburg, Germany
TCE exposed workers Not reported
Henschler et al., 1995
Deaths reported to among GE pension fund (Pittsfield, MA) Not reported
Greenland et al., 1994
Cardboard manufacturing workers, Atlanta area, GA
Not reported
Sinks et al., 1992
Aircraft manufacturing plant employees (Italy)
Not reported
Costa et al., 1989
Aircraft manufacturing plant employees (San Diego, CA)
All subjects, females
0.81 (0.52, 1.48) 16
Garabrant et al., 1988
Case-control Studies
Population of Montreal, Canada
Any TCE exposure
Not reported
Siemiatycki, 1991
6/22/2009
791
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Substantial TCE exposure
Not reported
Geographic Based Studies
Residents in two study areas in Endicott, NY
0.88 (0.65,1.18) 46 ATSDR, 2006
Residents of 13 census tracts in Redlands, CA
1.09 (0.97, 1.21) 536 Morgan and Cassady, 2002
Finnish residents
Vartiainenetal., 1993
Residents of Hausjarvi
Residents of Huttula
Not reported
Not reported
1 15 year lag
2 Internal referents, workers not exposed to TCE
3 Proportional mortality ratio
4 In Garabramt et al. (1998), Morgan et al. (1998) and Boice et al. (1999), breast cancer risk is for males and
females combined (ICD-9, 174, 175)
5 Risk ratio from Cox Proportional Hazard Analysis, stratified by age and sex, from Environmental Health Strategies
(1997) Final Report to Hughes Corporation (Communication from Paul A. Cammer, President, Trichloroethylene
Issues Group to Cheryl Siegel Scott, U.S. EPA, December 22, 1997).
6 The cohort of Costa et al. (1989) is composed of males only
4.7.2.1.3 Cervical Cancer
Ten cohort and 2 geographic based studies present relative risk estimates (Garabrant et
al., 1988; Anttila et al., 1995; Blair et al., 1998; Morgan et al., 1998; Boice et al., 1999; Hansen
et al., 2001; Morgan and Cassady, 2002; Raaschou-Nielsen et al., 2003; ATSDR, 2004, 2006;
Sung et al., 2007; Radican et al., 2008). Seven of these studies had high likelihood of TCE
exposure in individual study subjects and were judged to have met, to a sufficient degree, the
standards of epidemiologic design and analysis (Anttila et al., 1995; Blair et al., 1998; Morgan et
al., 1998; Boice et al., 1999; Hansen et al., 2001; Raaschou-Nielsen et al., 2003; Radican et al.,
2008). Three small cohort studies (Costa et al., 1989; Sinks et al., 1992; Henschler et al., 1995)
as well as three high-quality studies (Axelson et al., 1994; Zhao et al., 2005; Boice et al., 2006)
did not present relative risk estimates for cervical cancer. Additionally, one population case-
control and one geographic study of several site-specific cancers do not present information on
cervical cancer (Siemiatycki, 1991; Vartiainen et al., 1993).
Five high-quality studies observed elevated risk for cervical cancer and overall TCE
exposure [2.42, 95% CI: 1.05, 4.77 (Anttila et al., 1995); 1.8, 95% CI: 0.5, 6.5 (Blair et al., 1998)
that changed little with an additional 10 years follow-up, 1.67, 95% CI; 0.54, 5.22 (Radican et
al., 2008); 3.8, 95% CI: 1.42, 2.37 (Hansen et al., 2001); 1.9, 95% CI: 1.42, 2.37 (Raaschou-
Nielsen et al., 2003)]. The observations of a 3 to 4-fold elevated cervical cancer risk with high
6/22/2009
792
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
mean TCE exposure compared to subjects in the low exposure category [6+ppm: 4.35, 95% CI:
1.41, 10.1 (Anttila et al., 1995); 4+ ppm: 4.3, 95% CI: 0.5, 16 (Hansen et al., 2001)] or with high
cumulative TCE exposure [0.25 ppm-year: 3.0, 95% CI: 0.8, 11.7 (Blair et al., 1998), 2.83, 95%
CI: 0.86, 9.33 (Radican et al., 2008)] provides additional support for association with TCE.
Cervical cancer risk was lowest for subjects in the high exposure duration category (Hansen et
al., 2001; Raaschou-Nielsen et al., 2003); however, duration of employment is a poor exposure
metric given subjects may have differing exposure intensity with similar exposure duration
(NRC, 2006). No deaths due to cervical cancer were observed in two other high-quality studies
(Morgan et al., 1998; Boice et al., 1999), less than 4 deaths were expected, suggesting these
cohorts contained few female subjects with TCE exposure.
Human papilloma virus (HPV) and low socioeconomic status are known risk factors for
cervical cancer (ACS, 2008). Subjects in Raaschou-Nielsen et al. (2003) are blue-collar workers
and low socioeconomic status likely explains observed associations in this and the other high-
quality studies. The use of internal controls in Blair et al. (1998) who are similar in
socioeconomic status as TCE subjects is believed to partly account for possible confounder
related to socio-economic status; however, direct information on individual subjects is lacking.
Five other cohort and geographic based studies were given less weight in the analysis
because of their lesser likelihood of TCE exposure and other study design limitations that would
decrease statistical power and study sensitivity (Garabrant et al., 1988; Morgan and Cassady,
2002; ATSDR, 2004, 2006; Sung et al., 2007). Cervical cancer risk estimates in these studies
ranged between 0.65 (95% CI: 0.38, 1.02) (Morgan and Cassady, 2002) to 1.77 [proportional
mortality ratio] (95% CI: 0.57, 4.12) (ATSDR, 2004). No study reported a statistically
significant deficit in cervical cancer risk.
Table 4.7.8. Summary of human studies on TCE exposure and cervical cancer
Exposure Group
Cohort Studies - Incidence
Aerospace workers (Rocketdyne)
Relative Risk No. obs.
(95% CI) events Reference
Zhao et al., 2005
Any exposure to TCE
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
Not reported
Not reported
All employees at electronics factory (Taiwan)
0.96 (0.86, 1.22)1 337 Sung et al., 2007
6/22/2009
793
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Danish blue-collar worker w/TCE exposure
Any exposure
Exposure Lag Time
20 years
Employment duration
<1 year
1-4.9 years
> 5 years
Raaschou-Nielsen et al., 2003
1.9 (1.42,2.37)
62
1.5 (0.7, 2.9)
9
2.5 (1.7, 3.5)
30
1.6(1.0,2.4)
22
1.3 (0.6, 2.4)
10
Biologically-monitored Danish workers Hansen et al., 2001
Any TCE exposure 3.8(1.0,9.8) 4
Cumulative exp (Ikeda)
Mean concentration (Ikeda)
Employment duration
<17 ppm-yr 2.9 (0.04, 16) 1
>17 ppm-yr 2.6 (0.03, 14) 1
<4 ppm 3.4 (0.4, 12) 2
4+ppm 4.3(0.5,16) 2
< 6.25 yr 3.8(0.1,21) 1
>6.25 2.1 (0.03,12) 1
Aircraft maintenance workers from Hill Air Force Base
TCE subcohort Not reported
Cumulative exposure Not reported
Blair etal., 1998
Biologically-monitored Finnish workers Anttila et al., 1995
All subjects 2.42(1.05,4.77) 8
Mean air-TCE (Ikeda extrapolation)
<6 ppm 1.86 (0.38,5.45) 3
6+ppm 4.35(1.41,10.1) 5
Cardboard manufacturing workers in Arnsburg,
Germany
Exposed workers
Biologically-monitored Swedish workers
Any TCE exposure
Cardboard manufacturing workers, Atlanta area, GA
All subjects
Cohort Studies-Mortality
Aerospace workers (Rocketdyne)
Henschleretal., 1995
Not reported
Axelson et al., 1994
Not reported
Sinks et al., 1992
Not reported
6/22/2009
794
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Any TCE (utility/eng flush)
Any exposure to TCE
View-Master employees
Females
Not reported
Not reported
1.77 (0.57,
4.12)2
US Uranium-processing workers (Fernald)
Any TCE exposure Not reported
Light TCE exposure, >2 years duration Not reported
Mod TCE exposure, >2 years duration Not reported
Aerospace workers (Lockheed)
Routine Exp
Routine -Intermittent1
Aerospace workers (Hughes)
TCE Subcohort
- (0.00, 5.47)
Not reported
(0.00, 1.07)
Low Intensity (<50 ppm)
High Intensity (>50 ppm)
Aircraft maintenance workers (Hill AFB, Utah)
TCE subcohort
Cumulative exposure
1.8 (0.5, 6.5)
TCE sucohort
Cumulative exposure
0(3.5
exp)
0(1.91
exp)
0(1.54
exp)
0 1.01
< 5 ppm-yr 0.9(0.1,8.3) 1
5-25 ppm-yr 0
>25 ppm-yr 3.0(0.8,11.7) 4
1.67 (0.54, 5.22) 6
0 1.01
< 5 ppm-yr 0.76 (0.09, 6.35) 1
5-25 ppm-yr 0
>25 ppm-yr 2.83 (0.86, 9.33) 5
Boice et al., 2006
Zhao et al., 2005
ATSDR, 2004
Ritz, 1999
Boice et al., 1999
Morgan etal., 1998
Blair etal., 1998
Radican et al. (2008)
Cardboard manufacturing workers in Arnsburg,
Germany Henschler et al., 1995
TCE exposed workers Not reported
6/22/2009 795
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Unexposed workers
Not reported
Deaths reported to among GE pension fund (Pittsfield,
MA)
Not examined
Greenland et al., 1994
Cardboard manufacturing workers, Atlanta area, GA Not reported
Sinks et al., 1992
Aircraft manufacturing plant employees (Italy)
Not reported
Costa et al., 1989
Aircraft manufacturing plant employees (San Diego,
CA)
All subjects
0.61 (0.25,
1.26)6
Garabrant et al., 1988
Case-control Studies
Geographic Based Studies
Residents in two study areas in Endicott, NY
1.06 (0.29, 2.71) <6
ATSDR, 2006
Residents of 13 census tracts in Redlands, CA
0.65 (0.38, 1.02) 29
Morgan and Cassady, 2002
Finnish residents
Residents of Hausjarvi
Residents of Huttula
Not reported
Not reported
Vartiainen et al., 1993
1 Standardized incidence ratio for females in Sung et al. (2007) reflects a 15-year lag period
: Proportional mortality ratio
1 Internal referents, workers not exposed to TCE
' Nested case-control analysis
Males only in cohort
' SMR is for cancer of the genital organs (cervix, uterus, endometrium, etc).
4.7.2.2 Animal studies
Histopathology findings have been noted in reproductive organs in various cancer
bioassay studies conducted with TCE. A number of these findings (summarized in Table 4.7-9)
do not demonstrate a treatment-related profile.
Table 4.7.9. Histopathology findings in reproductive organs
Tumor incidence in mice after 18 months inhalation exposure a
Tissue
Finding
Control
100 ppm
500 ppm
Males
No. examined:
30
29
30
6/22/2009
796
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Prostate
Myoma
1
0
0
Testis
Carcinoma
0
0
1
Cyst
0
0
1
Females
No. examined:
29
30
28
Uterus
Adenocarcinoma
1
0
0
Ovary
Adenocarcinoma
1
0
0
Adenoma
3
1
3
Carcinoma
0
2
2
Granulosa cell tumor
4
0
2
Tumor incidence in rats after 18 months inhalation exposure a
Tissue
Finding
Control
100 ppm
500 ppm
Males
No. examined:
29
30
30
Testis
Interstitial cell
tumors
4
0
3
Females
No. examined:
28
30
30
Mammary
Fibroadenoma
2
0
0
Adenocarcinoma
3
2
2
Uterus
Adenocarcinoma
3
1
4
Ovary
Carcinoma
4
0
1
Granulosa cell tumor
1
0
0
Genital
tract
Squamous cell
carcinoma
0
2
0
Tumor incidence in hamsters after 18 months inhalation exposure a
Tissue
Finding
Control
100 ppm
500 ppm
Females
No. examined:
30
29
30
Ovary
Cystadenoma
1
0
0
Tumor incidence in mice after 18 months gavage administration b
Tissue
Finding
Con-
trol
TCE
Pure
TCE
Indus-
trial
TCE+
EPC
TCE
+BO
TCE
+EPC
+BO
Females
No. examined:
50
50
50
50
48
50
Mammary
Carcinoma
1
2
0
0
0
0
Ovary
Granulosa cell tumor
0
1
0
0
0
0
Vulva
Squamous cell
carcinoma
0
0
0
0
1
1
a Henschler et al., 1980.
b Henschler et al., 1984; EPC = epichlorohydrin; BO = 1,2-epoxybutane.
6/22/2009
797
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Cancers of the reproductive system that are associated with TCE exposure and observed
in animal studies are comprised of testicular tumors (interstitial cell and Leydig cell) (U.S. EPA,
2001). A summary of the incidences of testicular tumors observed in male rats is presented in
Table 4.7-10.
4.7.2.3 Mode of action for testicular tumors
The database for TCE does not include an extensive characterization of the mode of
action for Leydig cell tumorigenesis in the rat, although data exist that are suggestive of
hormonal disruption in male rats. A study by Kumar et al. (2000b) found significant decreases in
serum testosterone concentration and in 17-P-hydroxy steroid dehydrogenase (17-P-HSD),
glucose 6-p dehydrogenase (G6-PDH), and total cholesterol and ascorbic acid levels in testicular
homogenate from male rats that had been exposed via inhalation to 376 ppm TCE for 12 or 24
weeks. In a follow-up study, Kumar et al. (2001) also identified decreases in sorbital
dehydrogenase (SDH) in the testes of TCE-treated rats. These changes are markers of disruption
to testosterone biosynthesis. Evidence of testicular atrophy, observed in the 2001 study by
Kumar et al., as well as the multiple in vivo and in vitro studies that observed alterations in
spermatogenesis and/or sperm function, could also be consistent with alterations in testosterone
levels. Therefore, while the available data are suggestive of a MOA involving hormonal
disruption for TCE-induced testicular tumors, the evidence is inadequate to specify and test a
hypothesized sequence of key events.
Leydig cell tumors can be chemically induced by alterations of steroid hormone levels,
through mechanisms such as agonism of estrogen, gonadotropin releasing hormone (GnRH), or
dopamine receptors; antagonism of androgen receptors; and inhibition of 5a-reductase,
testosterone biosynthesis, or aromatase (Cook et al., 1999). For those plausible mechanisms that
involve disruption of the hypothalamic-pituitary-testis (HPT) axis, decreased testosterone or
estradiol levels or recognition is involved, and increased luteinizing hormone (LH) levels are
commonly observed. Although there is evidence to suggest that humans are quantitatively less
sensitive than rats in their proliferative response to LH, evidence of treatment-related Leydig cell
tumors in rats that are induced via HPT disruption is considered to represent a potential risk to
humans (with the possible exception of GnRh or dopamine agonists), since the pathways for
regulation of the HPT axis are similar in rats and humans (Clegg et al., 1997).
6/22/2009
798
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.7.10. Testicular tumors in male rats exposed to TCE, adjusted for reduced survival
a
Interstitial cell tumors after 103 weeks gavage exposure, beginning at 6.5-8 weeks of
age (NTP, 1988,1990)
Administered dose
Untreated
Vehicle
500
1,000
(mg/kg-d)
control
control
Male ACI rats
38/45 (84%)
36/44 (82%)
23/26 (88%)
17/19 (89%)
Male August rats
36/46 (78%)
34/46 (74%)
30/34 (88%)
26/30 (87%)
Male Marshall rats**
16/46 (35%)
17/46 (37%)
21/33 (64%)
32/39 (82%)
Male Osborne-Mendel rats
1/30 (3%)
0/28 (0%)
0/25 (0%)
1/19 (5%)
Male F344/N rats
44/47 (94%)
47/48 (98%)
47/48 (98%)
32/44 (73%)
Leydig cell tumors after 104 weeks inhalation exposure, beginning at 12 weeks of age
(Maltoni et al., 1986)
Administered daily
Control
112.5
337.5
675
concentration (mg/m3) b
Male Sprague-Dawley
6/114(5%)
16/105 (15%)
30/107 (28%)
31/113 (27%)
rats**
** Statistically significant by Cochran-Armitage trend test (p < 0.05).
a ACI rats alive at week 70, August rats at week 65, Marshall rats at week 32, Osborne-Mendel rats at week
97, F344/N rats at week 32, Sprague-Dawley rats at week 81 (except BT304) or week 62 (except BT304 bis).
b Equivalent to 100, 300, 600 ppm (100 ppm = 540 mg/m3), adjusted for 7 hr/d, 5 d/wk exposure.
Sources: NTP (1988) Tables A2, C2, E2, G2; NTP (1990) Table A3; Maltoni et al. (1986) IV/IV Table 21,
IV/V Table 21.
4.7.3 Developmental toxicity
An evaluation of the human and experimental animal data for developmental toxicity,
considering the overall weight and strength of the evidence, suggests a potential for adverse
outcomes associated with pre- and/or postnatal TCE exposures.
4.7.3.1 Human developmental data
Epidemiological developmental studies (summarized in Table 4.7-11) examined the
relationship between TCE exposure and prenatal developmental outcomes including spontaneous
abortion and perinatal death; decreased birth weight, small for gestational age, and postnatal
growth; congenital malformations; and other adverse birth outcomes. Postnatal developmental
6/22/2009
799
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
outcomes examined include developmental neurotoxicity, developmental immunotoxicity, other
developmental outcomes, and childhood cancer.
4.7.3.1.1 Adverse fetal/birth outcomes
Spontaneous Abortion and Perinatal Death. Spontaneous abortion or miscarriage is
defined as non-medically induced premature delivery of a fetus prior to 20 weeks gestation.
Perinatal death is defined as stillbirths and deaths before 7 days after birth. Available data comes
from several studies of occupational exposures in Finland and Santa Clara, California, and by
geographic-based studies in areas with known contamination of water supplies in Woburn, MA;
Tucson Valley, AZ; Rocky Mountain Arsenal, CO; Endicott, NY; and New Jersey.
Occupational Studies
The risks of spontaneous abortion and congenial malformations among offspring of men
occupationally exposed to TCE and other organic solvents were examined by Taskinen et al.
(1989). This nested case-control study was conducted in Finland from 1973-1983. Exposure
was determined by biological measurements of the father and questionnaires answered by both
the mother and father. The level of exposure was classified as "low/rare" if the chemical was
used <1 days/week, "intermediate" if used 1-4 days/week or if TCA urine measurements
indicated intermediate/low exposure, and "high/frequent" if used daily or if TCA urine
measurements indicated clear occupational exposure (defined as above the RfV for the general
population). There was no risk of spontaneous abortion from paternal TCE exposure (OR = 1.0,
95% CI = 0.6-2.0), although there was a significant increase for paternal organic solvent
exposure (OR = 2.7, 95% CI = 1.3-5.6) and a non-significant increase for maternal organic
solvent exposure (OR = 1.4, 95% CI = 0.6-3.0). (Also see section below for results from this
study for congenital malformations).
Another case-control study in Finland examined pregnancy outcomes in 1973-1986
among female laboratory technicians aged 20-34 years (Taskinen et al., 1994). Exposure was
reported via questionnaire, and was classified as "rare" if the chemical was used 1-2 days/week,
and "frequent" if used at least 3 days/week. Cases of spontaneous abortion (n = 206) were
compared with controls who had delivered a baby and did not report prior spontaneous abortions
(n = 329). A non-statistically significant increased risk was seen between spontaneous abortion
and TCE use at least 3 days a week (OR = 1.6, 95% CI = 0.5-4.8).
The association between maternal exposure to organic solvents and spontaneous abortion
was examined in Finland for births 1973-1983 (Lindbohm et al., 1990). Exposure was assessed
by questionnaire and confirmed with employment records, and the level of exposure was either
high, low or none based on the frequency of use and known information about typical levels of
6/22/2009
800
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
exposure for job type. Biological measurements of trichloroacetic acid in urine were also taken
on 64 women, with a median value of 48.1 |imol/L (mean: 96.2 ± 19.2 |imol/L). Three cases and
13 controls were exposed to TCE, with no increased risk seen for spontaneous abortion
(OR = 0.6, 95% CI = 0.2-2.3, p: 0.45).
A case-control study in Santa Clara County, California, examined the association
between solvents and adverse pregnancy outcomes in women >18 years old (Windham et al.,
1991). For pregnancies occurring between June 1986 and February 1987, 361 cases of
spontaneous abortion were compared to 735 women who had a live birth during this time period.
Telephone interviews included detailed questions on occupational solvent exposure, as well as
additional questions on residential solvent use. For TCE exposure, six cases of spontaneous
abortion were compared to four controls of live births; of these ten TCE-exposed individuals,
four reported exposure to tetrachloroethylene, and one reported exposure to paint strippers and
thinners. An increased risk of spontaneous abortions was seen with TCE exposure (OR = 3.1,
95% CI = 0.92-10.4), with a statistically significant increased risk for those exposed > 0.5
hrs/week (OR = 7.7, 95% CI = 1.3-47.4). An increased risk for spontaneous abortion was also
seen for those reporting a more "intense" exposure based primarily on odor, as well as skin
contact or other symptoms (OR = 3.9, p = 0.04). (Also see section below from this study on low
birth weight.)
Geographic-Based Studies
A community in Woburn, MA with contaminated well water experienced an increased
incidence of adverse birth outcomes and childhood leukemia (Lagakos et al., 1986). In 1979, the
wells supplying drinking water were found to be contaminated with 267 ppb TCE, 21 ppb
tetrachloroethylene, 11.8 ppb, and 12 ppb chloroform, and were subsequently closed. Pregnancy
and childhood outcomes were examined from 4,396 pregnancies among residents (Lagakos et
al., 1986). No association between water access and incidence of spontaneous abortion (n = 520)
was observed (p = 0.66). The town's water distribution system was divided into five zones,
which was reorganized in 1970. Prior to 1970, no association was observed between water
access and incidence of perinatal deaths (n = 46 still births and 21 deaths before 7 days)
(p = 0.55). However, after 1970, a statistically significant positive association between access to
contaminated water and perinatal deaths was observed (OR = 10.0, p = 0.003). The authors
could not explain why this discrepancy was observed, but speculated that contaminants were
either not present prior to 1970, or were increased after 1970. (Also see sections below on
decreased birth weight, congenital malformations, and childhood cancer for additional results
from this cohort).
A community in Tucson Valley, Arizona with contaminated well water had a number of
6/22/2009
801
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
reported cases of congenital heart disease. The wells were found to be contaminated with TCE
(range = 6-239 ppb), along with dichloroethylene and chromium (Goldberg et al., 1990). This
study identified 707 children born with congenital heart disease during the years 1969-1987. Of
the study participants, 246 families had parental residential and occupational exposure during
one month prior to conception and during the first trimester of pregnancy, and 461 families had
no exposure before the end of the first trimester. In addition to this control group, two others
were used: (1) those that had contact with the contaminated water area, and (2) those that had
contact with the contaminated water area and matched with cases for education, ethnicity, and
occupation. Among these cases of congenital heart disease, no significant difference was seen
for fetal death (not quantified) for exposed cases compared to unexposed cases. (Also see
section below on congenital malformations for additional results from this cohort.)
A residential study of individuals living near the Rocky Mountain Arsenal in Colorado
examined the outcomes in offspring of 75 men and 71 women exposed to TCE in drinking water
(ATSDR, 2001). TCE exposure was stratified by high (>10.0 ppb), medium (>5.0 ppm to <10.0
ppb), and low (<5.0 ppb). Among women with >5 ppb exposure experiencing miscarriage
(n = 22/57) compared to unexposed women experiencing miscarriage (// = 2/13) an elevated non-
significant association was observed (ORadj = 4.44, 95% CI = 0.76-26.12). For lifetime number
of miscarriages reported by men and women, results were increased but without dose-response
for women (medium: ORadj = 8.56, 95% CI = 0.69-105.99; high: ORadj = 4.16, 95%
CI = 0.61-25.99), but less for men (medium: ORadj = 1.68, 95% CI 0.26-10.77; high:
ORadj = 0.65, 95% CI = 0.12-3.48). Among women with >5 ppb exposure experiencing no live
birth (n = 9/57) compared to unexposed women experiencing no live birth (n = 1/13) an elevated
non-significant association was observed (ORadj = 2.46, 95% CI = 0.24-24.95). (Also see below
for results from this study on birth defects.)
New York State Department of Health (NYS DOH) and ATSDR conducted a study in
Endicott, NY to examine childhood cancer and birth outcomes in an area contaminated by a
number of VOCs, including "thousands of gallons" of TCE (ATSDR, 2006). Soil vapor levels
tested ranged from 0.18-140 mg/m3 in indoor air. A follow-up study by ATSDR (2008)
reported that during the years 1978-1993 only five spontaneous fetal deaths occurring > 20
weeks gestation were reported when 7.5 were expected (SIR = 0.66, 95% CI = 0.22-1.55) (See
sections on low birth weight, congenital malformations, and childhood cancer for additional
results from this cohort).
Women were exposed to contaminated drinking water while pregnant and living in 75
New Jersey towns during the years 1985-1988 (Bove, 1996; Bove et al., 1995). The water
contained multiple trihalomethanes, including an average of 55 ppb TCE, along with
tetrachloroethylene, 1,1,1 trichloroethane, carbon tetrachloride, 1,2-dichloroethane, and benzene.
6/22/2009
802
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
A number of birth outcomes were examined for 81,532 pregnancies, which resulted in 80,938
live births and 594 fetal deaths. No association was seen for exposure to > 10 ppb TCE and fetal
death (ORadj = 1.12). (See below for results from this study on decreased birth weight and
congenital malformations.)
Decreased birth weight, small for gestational age, and postnatal growth. Available data
pertaining to birth weight and other growth-related outcomes come from the case-control study
in Santa Clara, CA (discussed above), and by geographic-based studies as well as geographic
areas with known contamination of water supplies areas in Woburn, MA; Tucson, AZ, Endicott,
NY; Camp Lejeune, NC; and New Jersey.
Occupational Studies
The case-control study of the relationship between solvents and adverse pregnancy
outcomes discussed above (Windham et al., 1991) also examined intrauterine growth restriction
(IUGR). Telephone interviews included detailed questions on occupational solvent exposure, as
well as additional questions on residential solvent use. An increased risk of IUGR was observed
(OR = 12.5), although this was based only on one case that was exposed to both TCE and
tetrachloroethylene (Also see section above on spontaneous abortion).
Geographic-Based Studies
The study of Woburn, MA with contaminated well water discussed above (Lagakos et al.,
1986) examined birth weight. Of 3,462 live births surviving to 7 days, 220 were less than 6
pounds at birth (6.4%). No association was observed between water access and low birth weight
(p = 0.77). (See section on spontaneous abortion for study details, and see sections on
spontaneous abortion, congenital malformations, and childhood cancer for additional results
from this cohort).
An ecological analysis of well water contaminated with TCE in Tucson and birth-weight
was conducted by Rodenbeck et al. (2000). The source of the exposure was a US Air Force plant
and the Tucson International Airport. The wells were taken out of service in 1981 after
concentrations of TCE were measured in the range of <5 |ig/L to 107 |ig/L. The study
population consisted of 1,099 babies born within census tracts between 1979 and 1981, and the
comparison population consisted of 877 babies from nearby unexposed census tracts. There was
a non-significant increased risk for maternal exposure to TCE in drinking water and very-low-
birth-weight (<1,501 g) (OR = 3.3, 95% CI = 0.53-20.6). No increases were observed in the
low-birth-weight (<2,501 g) (OR = 0.9) or full-term (>35 week and < 46 week gestation) low-
birth-weight (OR = 0.81).
6/22/2009
803
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
The study of VOC exposure in Endicott, NY reported data on low birth weight and small
for gestational age (ATSDR, 2006, see section on spontaneous abortion for study details). For
births occurring during the years 1978-2002, low birth weight was slightly but statistically
elevated (OR= 1.26, 95% CI = 1.00-1.59), as was SGA(OR= 1.22, 95% CI = 1.02-1.45), and
full-term low birth weight (OR = 1.41, 95% CI = 1.01-1.95). (Also see sections on spontaneous
abortion, congenital malformations, and childhood cancer for additional results from this cohort).
Well water at the US Marine Corps Base in Camp Lejeune, NC was identified to be
contaminated with TCE, tetrachloroethylene, and 1,2-dichloroethane in April, 1982 and the wells
were closed in December, 1984. ATSDR examined pregnancy outcomes among women living
on the base during the years 1968-1985 (ATSDR, 1998). Compared to unexposed residents10
(n = 5,681), babies exposed to TCE long-term 11 (// = 31) had a lower mean birth weight after
adjustment for gestational age (-139 g, 90% CL = -277, -1), and babies exposed short-term12
(n = 141) had a slightly higher mean birth weight (+70g, 90% CL = -6, 146). For the long-term
group, no effect was seen for very low birth weight (<1,500 grams) or prematurity (>5 ppb,
OR = 1.05). No preterm births were reported in the long-term group and those (n = 8) in the
short-term group did not have an increased risk (0$ = 0.7, 90% CI = 0.3-1.2). A higher
prevalence of SGA was seen for small for gestational age (SGA)13 in the long-term exposed
group (n = 3; OR 1.5, 90% CL = 0.5, 3.8) compared to the short-term exposed group (OR =1.1,
90% CI = 0.2-1.1). When the long-term group was stratified by gender, male offspring were at
more risk for both reduced birth weight (-312 g, 90% CL = -632, -102) and SGA (OR 3.9, 90%
CL = 1.1-11.8). This study is limited due the mixture of chemicals in the water, as well as it
small sample size. ATSDR is currently reanalyzing the findings because of an error in the
exposure assessment related to the start-up date of a water treatment plant (ATSDR, 2007; GAO
2007a, b).
Pregnancy outcomes among women were exposed to contaminated drinking water while
pregnant and living in 75 New Jersey towns during the years 1985-1988 was examined by Bove
et al. (Bove, 1996; Bove et al., 1995). The water contained multiple trihalomethanes, including
an average of 55 ppb TCE, along with tetrachloroethylene, 1,1,1 trichloroethane, carbon
tetrachloride, 1,2-dichloroethane, and benzene. A number of birth outcomes were examined for
81,532 pregnancies, which resulted in 80,938 live births and 594 fetal deaths. A slight decrease
of 17.9 grams in birth weight was seen for exposure > 5 ppb, with a slight increase in risk for
10 Unexposed residents resided at locations not classified for long-term or short-term TCE exposure.
11 Long-term TCE exposed mothers resided at Hospital Point during 1968-1985 for at least one week prior to birth.
12 Short-term TCE exposed mothers resided at Berkeley Manor, Midway Park, Paradise Point, and Wakins Village
at the time of birth and at least 1 week during January 27 to February 7, 1985. In addition, the mother's last
menstrual period occurred on or before January 31, 1985 and the birth occurred after February 2, 1985.
13 SGA defined as singleton births less than the 10th percentile of published sex-specific growth curves.
6/22/2009
804
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
exposure >10 ppb (OR = 1.23), but no effect was seen for very low birth weight or
SGA/prematurity (>5 ppb, OR = 1.05). However, due to the multiple contaminants in the water,
it is difficult to attribute the results solely to TCE exposure. (See below for results from this
study on congenital malformations.)
Congenital Malformations. Three studies focusing on occupational solvent exposure
and congenital malformations from Europe provide data pertaining to TCE. Analyses of risk of
congenital malformations were also included in the studies in the four geographic areas
described above (Woburn, MA; Tucson, AZ, Rocky Mountain Arsenal, CO; Endicott, NY; and
New Jersey), as well as additional sites in Phoenix, AZ; and Milwaukee, WI. Specific categories
of malformations examined include cardiac defects, as well as cleft lip or cleft palate.
Occupational Studies
A study of 1,148 men and 969 women occupationally exposed to TCE in Finland from
1963-1976 to examined congenital malformations of offspring (Tola et al., 1980). Urinary
trichloroacetic acid measurements available for 2,004 employees ranged from <10 mg/L to >
500 mg/L, although 91% of the samples were below 100 mg/L. No congenital malformations
were seen in the offspring of women between the ages of 15-49 years, although 3 were expected
based on the national incidence. Expected number of cases for the cohort could not be estimated
because the number of pregnancies was unknown.
Men from Finland occupationally exposed to organic solvents including TCE did not
observe a risk of congenital malformations from paternal organic solvent exposure based on 17
cases and 35 controls exposed to TCE (OR = 0.6, 95% CI = 0.2-2.0) (Taskinen et al., 1989).
(Also see section above on spontaneous abortion for study details and additional results from this
cohort).
An occupational study of 100 women who gave birth to babies born with oral cleft
defects and 751 control women with normal births were examined for exposure to a number of
agents including TCE during the first trimester of pregnancy (Lorente et al., 2000). All women
were participants in a multi-center European case-referent study whose children were born
between 1989 and 1992. Four women were exposed to TCE, resulting in two cases of cleft lip
(ORadj = 3.21, 95% CI = 0.49-20.9), and two cases of cleft palate (ORadj = 4.47, 95%
CI = 1.02-40.9). Using logistic regression, the increased risk of cleft palate remained high
(OR = 6.7, 95% CI = 0.9-49.7), even when controlling for tobacco and alcohol consumption
(OR = 7.8, 95% CI = 0.8-71.8). However, the number of cases was small, and exposure levels
were not known.
6/22/2009
805
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Geographic-Based Studies
A community in Woburn, MA with contaminated well water experienced an increased
incidence of adverse birth outcomes and childhood leukemia (Lagakos et al., 1986, see section
on spontaneous abortion for study details). Statistically significant positive association between
access to contaminated water and eye/ear birth anomalies (OR = 14.9, p < 0.0001),
CNS/chromosomal/oral cleft anomalies (OR = 4.5, p = 0.01), kidney/urinary tract disorders
(OR = 1.35, p = 0.02) and lung/respiratory tract disorders (OR = 1.16, p = 0.05) were observed.
There were also five cases of cardiovascular anomalies, but there was not a significant
association with TCE (p = 0.91). However, since organogenesis occurs during gestational weeks
3-5 in humans, some of these effects could have been missed if fetal loss occurred. (Also see
sections on spontaneous abortion, perinatal death, decreased birth weight, and childhood cancer
for additional results from this cohort).
A high prevalence of congenital heart disease was found within an area of Tucson Valley,
AZ (Goldberg et al., 1990, see section on spontaneous abortion for study details and additional
results). Of the total 707 case families included, 246 (35%) were exposed to wells providing
drinking water found to be contaminated with TCE (range = 6-239 ppb), along with
dichloroethylene and chromium. Before the wells were closed after the contamination was
discovered in 1981, the OR of congenital heart disease was 3 times higher for those exposed to
contaminated drinking water compared to those not exposed; after the wells were closed, there
was no difference seen. This study observed 18 exposed cases of congenital heart disease when
16.4 would be expected (RR = 1.1). Prevalence of congenital heart disease in offspring after
maternal exposure during the first trimester (6.8 in 1,000 live births) was significantly increased
compared to non-exposed families (2.64 in 1,000 live births) (p < 0.001, 95% CI = 1.14-4.14).
No difference in prevalence was seen if paternal data was included, and there was no difference
in prevalence by ethnicity. In addition, no significant difference was seen for cardiac lesions.
A residential study of individuals living near the Rocky Mountain Arsenal in Colorado
examined the outcomes in offspring of 75 men and 71 women exposed to TCE in drinking water
(ATSDR, 2001). The risk was elevated for the nine birth defects observed (OR = 5.87, 95%
CI = 0.59-58.81), including one nervous system defect, one heart defect, and one incidence of
cerebral palsy. The remaining cases were classified as "other," and the authors speculate these
may be based on inaccurate reports. (See above for study details and results on spontaneous
abortion.)
The study of VOC exposure in Endicott, NY examined a number of birth defects during
the years 1983-2000 (ATSDR, 2006, see section on spontaneous for study details). These
include: total reportable birth defects, structural birth defects, surveillance birth defects, total
cardiac defects, major cardiac defects, cleft lip/cleft palate, neural tube defects, and choanal
6/22/2009
806
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
atresia (blocked nasal cavities). There were 56 expected cases of all birth defects and 61 were
observed resulting in no elevation of risk (rate ratio, RR = 1.08, 95% CI: 0.82-1.42). There were
no cases of cleft lip/cleft palate, neural tube defects, or choanal atresia. Both total cardiac
defects (n = 15; RR = 1.94, 95% CI = 1.21-3.12) and major cardiac defects (n = 6; RR = 2.52,
95% CI = 1.2-5.29) were statistically increased. A follow-up study by ATSDR (2008) reported
that conotruncal heart malformations were particularly elevated (n = 4; RR = 4.83, 95% C =
1.81-12.89). The results remained significantly elevated (aRR = 3.74; 95% CI = 1.21-11.62)
when infants with Down syndrome were excluded from the analysis. (Also see sections on
spontaneous abortion, decreased birth weight, and childhood cancer for additional results from
this cohort).
In the New Jersey study described previously, the prevalence of birth defects reported by
surveillance systems was examined among the women exposed to TCE and other contaminants
in water while pregnant between 1985-1988 (Bove, 1996; Bove et al., 1995). For exposure >10
ppb (n = 1,372), an increased risk, with relatively wide confidence intervals, was seen for all
birth defects (OR = 2.53, 95% CI = 0.77-7.34). An increased risk was also seen for CNS defects
(>10 ppb: OR= 1.68), specifically 56 cases of neural tube defects (<1-5 ppb: 1.58, 95%
CI = 0.61-3.85; >10ppb: OR = 2.53, 95% CI = 0.77-7.34). A slight increase was seen in major
cardiac defects (>10 ppb: OR = 1.24, 50% CI = 0.75-1.94), including ventrical septal defects
(>5ppb: OR = 1.30, 95% CI = 0.88-1.87). An elevated risk was seen for 9 cases of oral clefts
(<5 ppb: OR = 2.24, 95% CI = 1.04-4.66), although no dose-response was seen (>10 ppb,
OR = 1.30). However, due to the multiple contaminants in the water, it is difficult to attribute
the results solely to TCE exposure. (See above for results from this study on fetal death and
decreased birth weight.)
Arizona Department of Heath Services (ADHS) conducted studies of contaminated
drinking water and congenital malformations (<20 years old) in Maricopa County, which
encompasses Phoenix and the surrounding area (ADHS, 1988). TCE contamination was
associated with elevated levels of deaths in children less than 20 years old due to total congenital
anomalies in East Central Phoenix from 1966-1969 (RR = 1.4, 95% CI = 1.1-1.7), from
1970-1981 (RR= 1.5, 95% CI = 1.3-1.7), and from 1982-1986 (RR = 2.0, 95% CI = 1.5-2.5),
as well as in other areas of the county. (See below for results from this study on childhood
leukemia.)
A study was conducted of children born 1997-1999 with congenital heart defects in
Milwaukee, WI (Yauck et al., 2004). TCE emissions data were ascertained from state and US
EPA databases, and distance between maternal residence and the emission source was
determined using a geographic information system (GIS). Exposure was defined as those within
1.32 miles from at least one site. Results showed that an increased risk of congenital heart
6/22/2009
807
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
defects was seen for the offspring of exposed mothers 38 years old or older (OR = 6.2, 95%
CI = 2.6-14.5), although an increased risk was also seen for offspring of unexposed mothers 38
years old or older (OR = 1.9, 95% CI = 1.1-3.5), and no risk was seen for offspring of exposed
mothers younger than 38 years (OR = 0.9, 95% CI = 0.6-1.2). The authors speculate that studies
that did not find a risk only examined younger mothers. The authors also note that statistically-
significant increased risk was seen for mothers with preexisting diabetes, chronic hypertension,
or alcohol use during pregnancy.
An abstract reported that twenty-eight people living in a Michigan town were exposed for
5-10 years to 8-14 ppm TCE in well water (Bernad et al., 1987, abstract). One child was born
with multiple birth defects, with no further details.
Other adverse birth outcomes. TCE was previously used as a general anesthetic during
pregnancy. One study measured the levels of TCE in maternal and newborn blood after use
during 34 vaginal childbirths (Beppu, 1968). TCE was administered through a vaporizer from
two to 98 minutes (mean 34.7 minutes) at volumes from 2 to 8 mL (mean 4.3 mL). Mean blood
TCE concentrations were: 2.80 ±1.14 mg/dl in maternal femoral arteries; 2.36±1.17 mg/dl in
maternal cubital veins; 1.83±1.08 mg/dl in umbilical vein; and 1.91±0.95 mg/dl in the umbilical
arteries. A significant correlation was seen for maternal arterial blood and infants' venous blood,
and the concentration of the fetal blood was lower than that of the mother. Of these newborns,
one had asphyxia and three "sleepy babies" had Apgar scores of 5 to 9; however these results
could not be correlated to length of inhalation and there was no difference in the TCE levels in
the mother or newborn blood compared to those without adverse effects. Discussion included
delayed newborn reflexes (raising the head and buttocks, bending the spine, and sound reflex),
blood pressure, jaundice, and body weight gain; however the results were compared to newborns
exposed to other compounds, not to an unexposed population. This study also examined the
concentration of TCE in one mother at 22 weeks gestation exposed for four minutes, after which
the fetus was "artificially delivered". Maternal blood concentration was 3.0 mg/dl, and 0.9 mg/dl
of TCE was found in the fetal heart, but not in other organs.
Another study of TCE administered during childbirth to the mother as an analgesic
examined perinatal measures, including fetal pH, fetal PC02, fetal base deficit, fetal P02, Apgar
scores, and neonatal capillary blood (Phillips and Macdonald, 1971). The study consisted of 152
women whose fetus was considered to be at risk for hypoxia during labor. Out of this group, 51
received TCE (amount and route of exposure not reported). TCE caused fetal pH to fall more,
base deficit increased more, and P02fell more than the control group by four-fold or more
compared to other analgesics used.
6/22/2009
808
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.7.3.1.2 Postnatal Developmental Outcomes
Developmental neurotoxicity. The studies examining neurotoxic effects from TCE
exposure are discussed in Section 4.2, and the human developmental neurotoxic effects are
reiterated here.
Occupational Studies
An occupational study examined the neurodevelopment of the offspring of 32 women
exposed to various organic solvents during pregnancy (Laslo-Baker et al., 2004; Till et al.,
2001). Three of these women were exposed to TCE; however no levels were measured and the
results for examined outcomes are for total organic solvent exposure, and are not specific to
TCE.
Geographic-Based Studies
A study of three residential cohorts (Woburn, MA, Alpha, OH, and Twin Cities, MN)
examined the neurological effects of TCE exposure in drinking water (White et al., 1997). For
Woburn, MA, 28 individuals ranging from 9-55 years old were assessed, with exposure from a
tanning factor and chemical plant at levels 63-400 ppb for <1 to 12 years; the time between
exposure and neurological examination was about 5 years. In this cohort, six of thirteen children
(46%) had impairments in the verbal naming/language domain. For Alpha, OH, 12 individuals
ranging from 12-68 years old were assessed, with exposure from degreasing used at a
manufacturing operation at levels 3.3-330 ppb for 5-17 years; the time between exposure and
neurological examination was 5-17 years. In this cohort, one of two children (50%) had
impairments in the verbal naming/language domain. For Twin Cities, MN, 14 individuals
ranging from 8-62 years old were assessed, with exposure from an army ammunition plant at
levels 261-2,440 ppb for 0.25-25 years; the time between exposure and neurological
examination was 4-22 years. In this cohort, four of four children (100%) had impairments in the
verbal naming/language, memory, and academic domains and were diagnosed with moderate
encephalopathy; and three of four children {15%) performed poorly on the WRAT-R Reading
and Spelling and WAIS-R Information tests.
A case-control study was conducted to examine the relationship between multiple
environmental agents and autism spectrum disorder (ASD) (Windham et al., 2006). Cases
(n = 284) and controls (n = 657) were born in 1994 in the San Francisco Bay Area. Cases were
diagnosed before age nine. Exposure was determined by geocoding births to census tracts, and
linking to hazardous air pollutants (HAPs) data. An elevated risk was seen for TCE in the upper
3rd quartile (OR = 1.37, 95% CI = 0.96-1.95), and a statistically significant elevated risk was
seen for the upper 4th quartile (OR = 1.47, 95% CI = 1.03-2.08).
6/22/2009
809
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
The Trichloroethylene Subregistry (Burg et al., 1995; Burg and Gist, 1999), including
948 children <18 years old from 13 sites located in 3 states, was examined for any association of
ingestion of drinking water contaminated with TCE and various health effects (Burg et al., 1995;
Burg and Gist, 1999; ATSDR, 2003a). Exposure groups included (1) maximum TCE exposure,
(2) cumulative TCE exposure, (3) cumulative chemical exposure, and (4) duration of exposure.
Exposed children 0-9 years old had statistically increased hearing impairment compared to
controls (RR = 2.13, 99% CI = 1.12-4.07), with children <5 having a 5.2-fold increase over
controls. Exposed children 0-9 years old also had statistically increased speech impairment (RR
2.45, 99% CI = 1.31-4.58). In addition, anemia and other blood disorders were statistically
higher for males 0-9 years old. The authors noted that exposure could have occurred prenatally
or postnatally. There was further analysis on the 116 exposed children and 182 controls who
were under 10 years old at the time that the baseline study was conducted by ATSDR. This
analysis did not find a continued association with speech and hearing impairment in these
children; however, the absence of acoustic reflexes (contraction of the middle ear muscles in
response to sound) remained significant (ATSDR, 2003a). No differences were seen when
stratified by prenatal and postnatal exposure.
Twenty-eight people living in a Michigan town were exposed for 5-10 years to 8-14
ppm TCE in well water (Bernad et al., 1987, abstract). Ten adults and 12 children completed a
questionnaire on neurotoxic endpoints. Nine of the 12 children had poor learning ability,
aggressive behavior, and low attention span.
Developmental immunotoxicity. The studies examining human immunotoxic effects
from TCE exposure are discussed in Section 4.5.1. The studies reporting developmental effects
are reiterated briefly here.
Two studies focused on immunological development in children after maternal exposure
to VOCs (Lehmann et al., 2001, 2002). The first examined premature neonates (1,500-2,500 g)
and neonates at risk of atopy (cord blood IgE >0.9 kU/L; double positive family atopy history) at
36 months of age (Lehmann et al., 2001). Median air level in child's bedroom measured 0.42
|ig/m3. There was no association with allergic sensitization to egg white and milk, or to cytokine
producing peripheral T cells. The second examined healthy, full-term neonates (>2,500 g; >37
weeks gestation) born in Lepzig, Germany (Lehmann et al., 2002). Median air level in the
child's bedroom 3-4 weeks after birth measured 0.6 |ig/m3. A significant reduction of Thl IL-2
producing T cells was observed.
Byers et al. (1988) observed altered immune response in family members of children
diagnosed with leukemia in Woburn, MA (Lagakos et al., 1986, see below for results of this
study). The family members included 13 siblings under 19 years old at the time of exposure;
6/22/2009
810
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
however, an analysis looking at only these children was not done. This study is discussed in
further detail in Section 4.5.1.
Other developmental outcomes. A study demonstrated the adverse effects of TCE used
as an anesthetic in children during operations during 1964 in Poland to repair developmental
defects of the jaw and face (Jasinka, 1965, translation). 55 children ranging from 6 months to 10
years old were anesthetized with at least 10 mL TCE placed into an evaporator. Bradycardia
occurred in 2 children, an accelerated heart rate of 20-25 beats per minute occurred in 7
children, no arrhythmia was observed, and arterial blood pressure remained steady or dropped by
10 mmHG only. Respiratory acceleration was observed in 25 of the children, and was seen more
in infants and younger children.
Childhood Cancer. A number of studies of parental occupational exposure were
conducted in North America and the United Kingdom to determine an association with
childhood cancer. A number of geographic-based studies were conducted in California; New
Jersey; Woburn, MA; Endicott, NY; Phoenix, AZ; and Tucson, AZ. Specific categories of
childhood cancers examined include leukemia, non-Hodgkin's lymphoma, and CNS tumors.
Occupational Studies
Brain tumors in 98 children less than 10 years old at diagnosis from 1972-1977 in Los
Angeles County have been observed in the offspring of fathers (Peters et al., 1981, 1985).
Exposure was determined by questionnaire. Two cases with TCE exposure were reported: one
case of oligodendroglioma in an 8-year-old whose father was a machinist, and astrocytoma in a
7-year-old whose father was an inspector for production scheduling and parts also exposed to
methyl ethyl ketone (Peters et al., 1981). Peters et al. (1985) also briefly mentioned 5 cases and
no controls of paternal exposure to TCE and brain tumors in the offspring (resulting in an
inability to calculate an odds ratio), but without providing any additional data.
A case-control study was conducted to assess an association between parental
occupational exposure and neuroblastoma diagnosed in offspring <19 years old in the US and
Canada from May 1992 to April 1994 (De Roos et al., 2001). Paternal self-reported exposure to
TCE was reported in 22 cases and 12 controls, resulting in an elevated risk of neuroblastoma in
the offspring (OR = 1.4, 95%CI = 0.7-2.9). Maternal exposure to TCE was not reported.
A case-control study of parental occupational exposure and childhood leukemia was
conducted in Los Angeles County (Lowengart et al., 1987). Children (61 boys and 62 girls)
diagnosed less than 10 years old (mean age 4 years) from 1980 to 1984 were included in the
analysis. Paternal occupation exposure to TCE was elevated for one year preconception
6/22/2009
811
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(OR = 2.0, p = 0.16), prenatal (OR = 2.0, p = 0.16), and postnatal (OR = 2.7, p = 0.7). Maternal
exposure to TCE was not reported.
A case-control study children diagnosed with acute lymphoblastic leukemia (ALL)
examined parental occupational exposure to hydrocarbons in the US and Canada (Shu et al.,
1999). Children were under the age of 15 years at diagnosis during the years 1989 to 1993.
Cases were confirmed with a bone marrow sample. 1,842 case-control pairs were given
questionnaires on maternal and paternal exposures, resulting in 15 cases and 9 controls
maternally exposed and 136 cases and 104 controls paternally exposed to TCE. There was an
increased but non-significant risk for maternal exposure to TCE during preconception (OR = 1.8,
95% CI = 0.6-5.2), pregnancy (OR = 1.8, 95% CI = 0.5-6.4), postnatally (OR = 1.4, 95%
CI = 0.5-4.1), or any of these periods (OR = 1.8, 95% CI = 0.8-4.1). However, there was no
increased risk for paternal exposure to TCE.
Occupational exposure in communities in the UK was examined to determine an
association with leukemia and non-Hodgkin's lymphoma diagnosed in the offspring (McKinney
et al., 1991). Paternal occupational exposure was elevated for exposure occurring during
preconception (OR = 2.27, 95% CI = 0.84-6.16), prenatal (OR = 4.40, 95% CI = 1.15-21.01),
and postnatal (OR = 2.66, 95% CI = 0.82-9.19). Risk from maternal preconception exposure
was not elevated (OR = 1.16, 95% CI = 0.13-7.91). However, the number of cases examined in
this study was low, particularly for maternal exposure.
Geographic-Based Studies
A California community exposed to TCE (0.09-97 ppb) in drinking water from
contaminated wells was examined for cancer (Morgan and Cassady, 2002). A specific emphasis
was placed on the examination of 22 cases of childhood cancer diagnosed before 15 years old.
However, the incidence did not exceed those expected for the community for total cancer
(SIR = 0.83, 99% CI = 0.44-1.40), CNS cancer (SIR = 1.05, 99% CI = 0.24-2.70), and leukemia
(SIR = 1.09, 99% CI = 0.38-2.31).
An examination of drinking water was conducted in four New Jersey counties to
determine an association with leukemia and non-Hodgkin's lymphoma (Cohn et al., 1994). A
number of contaminants were reported, including VOCs and trihalomethanes. TCE was found as
high as 67 ppb, and exposure categories were assigned to be >0.1, 0.1-5 and >5 ppb. A
significantly elevated dose-response risk for ALL was observed for girls diagnosed before 20
years old (RR = 3.36, 95% CI = 1.29-8.28), which was increased among girls diagnosed before
5 years old (RR = 4.54, 95% CI = 1.47-10.6). A significantly elevated dose-response risk for
girls was also observed for total leukemia (RR = 1.43, 95% CI = 1.07-1.98).
The Woburn, MA community with contaminated well water experienced an increase in
6/22/2009
812
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
the incidence of childhood leukemia (Costas et al., 2002; Cutler et al., 1986; Lagakos et al.,
1986; MADPH, 1997). An initial study examined twelve cases of childhood leukemia diagnosed
in children less than 15 years old between 1969-1979, when 5.2 cases were expected, and a
higher risk was observed in boys compared to girls; however, no factors were observed to
account for this increase (Cutler et al., 1986). Another study observed statistically significant
positive association between access to contaminated water and 20 cases of childhood cancer
were observed for both cumulative exposure metric (OR = 1.39, p = 0.03), and none versus some
exposure metric (OR = 3.03, p = 0.02) (Lagakos et al., 1986). Massachusetts Department of
Public Health (MADPH, 1997) conducted a case-control study of children less than 20 years old
living in Woburn and diagnosed with leukemia between 1969 and 1989 (n = 21) and observed
that consumption of drinking water increased the risk of leukemia (OR = 3.03, 95%
CI = 0.82-11.28), with the highest risk from exposure during fetal development (OR = 8.33,
95% CI 0.73-94.67). This study found that paternal occupational exposure to TCE was not
related to leukemia in the offspring (MADPH, 1997). In the most recent update, Costas et al.
(2002) reported that between the years 1969 and 1997, 24 cases of childhood leukemia were
observed when 11 were expected. Risk was calculated for cumulative exposure to contaminated
drinking water two years prior to conception (ORadj = 2.61, 95% CI = 0.47-14.97), during
pregnancy (ORadj = 8.33, 95% CI = 0.73-94.67), postnatal (ORadj =1.18, 95% CI = 0.28-5.05),
and any of these time periods (ORadj = 2.39, 95% CI = 0.54-10.59). A dose response was
observed during pregnancy only. Cases were more likely to be male (76%), <9 years old at
diagnosis (62%), breast-fed (OR = 10.17, 95% CI = 1.22-84.50), and exposed during pregnancy
(adjusted OR = 8.33, 95% CI = 0.73-94.67). A dose-response was seen during the pregnancy
exposure period, with the most exposed having an adjusted OR of 14.30 (95%
CI = 0.92-224.52). Other elevated risks observed included maternal alcohol intake during
pregnancy (OR = 1.50, 95% CI = 0.54-4.20), having a paternal grandfather diagnosed with
cancer (OR = 2.01, 95% CI = 0.73-5.58), father employed in a high risk industry (OR = 2.55,
95% CI 0.78-8.30), and public water being the subject's primary beverage (OR = 3.03, 95%
CI = 0.82-11.28). (Also see sections on spontaneous abortion, perinatal death, decreased birth
weight, and congenital malformations for additional results from this cohort).
The study of VOC exposure in Endicott, NY discussed above observed fewer than six
cases of cancer that were diagnosed between 1980 and 2001 in children less than 20 years old,
and did not exceed expected cases or types (ATSDR, 2006). (See section on spontaneous
abortion for study details, and sections on spontaneous abortion, decreased birth weight, and
congenital malformations for additional results from this cohort).
Arizona Department of Heath Services (ADHS) conducted a number of studies of
contaminated drinking water and 189 cases of childhood cancer (<20 years old) (ADHS, 1988,
6/22/2009
813
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1990a, b, c, 1997). In Maricopa County, which encompasses Phoenix and the surrounding area,
TCE contamination (8.9 and 29 ppb in two wells) was associated with elevated levels of
childhood leukemia (n = 67) in west central Phoenix during 1965-1986 (SIR = 1.67, 95%
CI = 1.20-2.27) and 1982-1986 (SIR= 1.91, 95% CI = 1.11-3.12), but did not observe a
significant increase in total childhood cancers, lymphoma, brain/CNS, or other cancers during
these time periods (ADHS, 1990a). (See above for results from this study on congenital
anomalies.) A follow-up study retrospectively asked parents about exposures and found that
residence within 2 miles of wells contaminated with TCE was not a risk factor for childhood
leukemia, but identified a number of other risk factors (ADHS, 1997). A further study of East
Phoenix, reported on TCE contamination found along with 1,1,1-trichloroethane and 25 other
contaminants in well water (levels not reported) and found no increase in incidence of childhood
leukemia (SIR = 0.85, 95% CI = 0.50-1.35) based on 16 cases (ADHS, 1990b). There were also
16 cases of other types of childhood cancer, but were too few to be analyzed separately. In Pima
County, which encompasses Tucson and the surrounding area, TCE was found in drinking wells
(1.1-239 ppb), along with 1,1-DCE, chloroform and chromium and found a non-statistically
elevated risk of leukemia was observed (SIR = 1.50, 95% CI = 0.76-2.70), but no risk was
observed for testicular cancer, lymphoma, or CNS/brain cancer (ADHS, 1990c).
4.7.3.1.3 Summary of human developmental toxicity
Epidemiological developmental studies examined the association between TCE exposure
and a number of prenatal and postnatal developmental outcomes. Prenatal developmental
outcomes examined include spontaneous abortion and perinatal death; decreased birth weight,
small for gestational age, and postnatal growth; congenital malformations; and other adverse
birth outcomes. Postnatal developmental outcomes examined include developmental
neurotoxicity, developmental immunotoxicity, other developmental outcomes, and childhood
cancer related to TCE exposure.
More information on developmental outcomes is expected. A follow-up study of the
Camp Lejeune cohort (ATSDR, 1998) for birth defects and childhood cancers was initiated in
1999 (ATSDR, 2003b) and expected to be completed soon (GAO, 2007a, b). Out of a total of
106 potential cases of either birth defects or childhood cancer, 57 have been confirmed and will
constitute the cases. These will be compared 548 control offspring of mothers who also lived at
Camp Lejeune during their pregnancy from 1968-1985. As part of this study, a drinking water
model was developed to determine a more accurate level and duration of exposure to these
pregnant women (ATSDR, 2007). Additional health studies have been suggested, including
adverse neurological or behavioral effects or pregnancy loss.
6/22/2009
814
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.7.11. Developmental studies in humans
Subjects
Exposure
Effect
Reference
Adverse fetal/birth outcomes
Spontaneous Abortion and Perinatal Death
371 men occupationally exposed Questionnaire:
to solvents in Finland 1973-1983 Low/rare: used <1 day/week;
Intermediate: used 1-4 days/week
or intermediate/low TCA urine
levels;
High/frequent: used daily or high
TCA urine levels
No risk of spontaneous abortion after paternal exposure,
based on 17 cases and 35 controls exposed to TCE
(OR = 1.0, 95% CI = 0.6-2.0)
Taskinen et al., 1989
535 women occupationally
exposed to solvents in Finland
1973-1986
Questionnaire
Rare: used 1-2 days/week;
Frequent: used >3 days/week
Increased risk of spontaneous abortion among frequently-
exposed women, based on 7 cases and 9 controls exposed
to TCE (OR = 1.6, 95% CI = 0.5-4.8)
Taskinen et al., 1994
3,265 women occupationally
exposed to organic solvents in
Finland 1973-1983
Questionnaire
Urine TCA: median: 48.1 |imol/L:
mean 96.2 ± 19.2 ^mol/L
No increased risk of spontaneous abortion based on 3
cases and 13 controls exposed to TCE
OR = 0.6, 95% CI = 0.2-2.3
Lindbohm et al., 1990
361 women occupationally and
residentially exposed to solvents
in Santa Clara County, California
June 1986-February 1987 (735
controls)
Questionnaire
Increased risk of spontaneous abortion based on 6 cases
and 4 controls exposed to TCE14
OR = 3.1, 95% CI = 0.92-10.4
Windham et al., 1991
14 Of those exposed to TCE, 4 were also exposed to tetrachloroethylene and 1 was also exposed to paint strippers and thinners.
6/22/2009 815
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
4,396 pregnancies among
residents of Woburn, MA
1960-1982
TCE: 267 ng/L
Tetrachloroethylene: 21 |ig/L
Chloroform: 12 |ig/L
Increased risk of perinatal death (n = 67) after 1970
(p = 0.55) but not before 1970 (OR = 10, p = 0.003)
No increased risk of spontaneous abortion (n = 520;
p = 0.66)
Lagakos et al., 1986
707 parents of children with
congenital heart disease in Tucson
Valley, AZ 1969-1987
6-239 ppb TCE, along with DCA
and chromium
No increased risk of fetal death (not quantified) based on
246 exposed and 461 unexposed cases
Goldberg et al., 1990
75 men and 71 women living near
Rocky Mountain Arsenal,
Colorado 1981-1986
Low: <5.0 ppb
Medium: >5.0 to <10.0 ppb
High: <10.0 ppb
Increased risk of miscarriage
ORadj = 4.44, 95% CI = 0.76-26.12
Increased risk of no live birth
ORadj = 2.46, 95% CI = 0.24-24.95
ATSDR, 2001
1,440 pregnancies among
residents of Endicott, NY
1978-2002
indoor air from soil vapor: 0.18-140
mg/m3
No increase in spontaneous fetal death
SIR = 0.66, 95% CI = 0.22-1.55
ATSDR, 2006, 2008
81,532 pregnancies among
residents of 75 New Jersey towns
1985-1988 (3 control groups)
55 ppb TCE, along with many other
compounds
No increased risk of fetal death for >10 ppb
OR = 1.12
Bove, 1996; Bove et al.,
1995
Decreased birth weight, small for gestational age. and postnatal growth
361 women occupationally and Questionnaire
residentially exposed to solvents
in Santa Clara County, California
June 1986-February 1987 (735
controls)
Increased risk of intrauterine growth restriction (IUGR)
based on one case exposed to both TCE and
tetrachloroethylene
OR = 12.5
Windham et al., 1991
6/22/2009
816
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
3,462 births in Woburn, MA
1960-1982
267 |ig/L TCE in drinking water,
along with tetrachloroethylene and
chloroform
No increase in low birth weight (p = 0.77)
Lagakos et al., 1986
1,099 singleton births 15 to
residents of 3 census tracts near
Tucson International Airport
1979-1981 (877 controls)
<5-107 (ig/L
No increase in full-term low birth weight (OR = 0.81)
No increase in low birth weight (OR = 0.9)
Increase in very low birth weight
OR = 3.3, 95% CI = 0.53-20.6
Rodenbeck et al., 2000
1,440 births 16 to residents of
Endicott, NY 1978-2002
indoor air from soil vapor: 0.18-140
mg/m3
Small increase in low birth weight
OR = 1.26, 95% CI = 1.00-1.59
Small increase in small for gestational age
OR = 1.22, 95% CI = 1.02-1.45
Increase in full-term low birth weight
OR = 1.41, 95% CI = 1.01-1.95
ATSDR, 2006, 2008
6,289 pregnancies among women
residing at Camp Lejeune, NC
1968-1985 (141 short-term and
31 long-term TCE-exposed, 5,681
unexposed controls)17
Tarrawa Terrace:
TCE: 8 ppb;
1,2-DCE: 12 ppb
PCE: 215 ppb
Hadnot Point:
TCE: 1,400 ppb
1,2-DCE: 407 ppb
Change in mean birth weight
Long-term total: -139 g, 90% CI = -277, -1
Long-term males: -312 g, 90% CI = -540, -85
Short term total: +70g, 90% CI = -6, 146
Increase in small for gestational growth (SGA)
Long-term total: OR 1.5, 90% CI = 0.5, 3.8
Long-term males: OR 3.9, 90% CI= 1.1-11.9
Short term total: OR 1.1, 90% CI = 0.2-1.1
ATSDR, 1998
15 Full term defined as between 35 and 46 weeks gestation, low birth weight as < 2501g, and very low birth weight as < 1501g.
16 Low birth weight defined as <2500, moderately low birth weight (1500g-<2500g), term low birth weight (>=37 weeks gestation and <25000g)
17 Unexposed residents resided at locations not classified for long-term or short-term TCE exposure. Long-term TCE exposed mothers resided at Hospital Point
during 1968-1985 for at least one week prior to birth. Short-term TCE exposed mothers resided at Berkeley Manor, Midway Park, Paradise Point, and Wakins
Village at the time of birth and at least 1 week during January 27 to February 7, 1985. In addition, the mother's last menstrual period occurred on or before
January 31, 1985 and the birth occurred after February 2, 1985.
6/22/2009
817
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
81,532 pregnancies 18among
residents of 75 New Jersey towns
1985-1988
55 ppb TCE, along with many other
compounds
Decreased birth weight at > 5 ppb by 17.9g
No increase in prematurity at >10 ppb: OR = 1.02
Increase in low birth weight, term
>10 ppb: OR = 1.23, 50% CI = 1.09-1.39
No risk for very low birth weight
Bove, 1996; Bove et al.,
1995
Congenital Malformations
1,148 men and 969 women
occupationally exposed to TCE in
Finland 1963-1976
Urinary TCA:
< 10 to > 500 mg/L
No congenital malformations reported
Tola et al., 1980
371 men occupationally exposed
to solvents in Finland 1973-1983
Low/rare: used <1 day/week;
Intermediate: used 1-4 days/week or
if biological measures indicated
high exposure;
High/frequent: used daily or if
biological measures indicated high
exposure
No increase in congenital malformations based on 17
cases and 35 controls exposed to TCE
OR = 0.6, 95% CI = 0.2-2.0
Taskinen et al., 1989
100 babies with oral cleft defects
born to women occupationally
exposed in Europe 1989-1992
Questionnaire
Increase in cleft lip based on 2 of 4 TCE-exposed women
ORadj = 3.21, 95% CI = 0.49-20.9
Increase in cleft palate based on 2 of 4 TCE-exposed
women
ORadj = 4.47, 95% CI = 1.02-40.9
Lorente et al., 2000
18 Low birth weight defined as <2500 g, very low birth weight as <1500 g.
6/22/2009
818
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
4,396 pregnancies among
residents of Woburn, MA
1960-1982
TCE: 267 ng/L
Tetrachloroethylene: 21 |ig/L
Chloroform: 12 |ig/L
Increase in eye/ear birth anomalies: OR = 14.9,
p< 0.0001
Increase in CNS/chromosomal/oral cleft anomalies:
OR = 4.5, p = 0.01
Increase in kidney/urinary tract disorders:
OR = 1.35, p = 0.02
Small increase in lung/respiratory tract disorders:
OR = 1.16, p = 0.05
No increase in cardiovascular anomalies (n = 5): p = 0.91
Lagakos et al., 1986
707 children with congenital heart
disease in Tucson Valley, AZ
1969-1987 (246 exposed, 461
unexposed)
Wells contaminated with TCE
(range: 6-239 ppb), along with DCA
and chromium
Increase in congenital heart disease
<1981: OR-3 (p < 0.005)
>1981: OR~l
Increased prevalence after maternal exposure during first
trimester (p < 0.001, 95% CI = 1.14-4.14)
Goldberg et al., 1990
75 men, 71 women living near
Rocky Mountain Arsenal,
Colorado 1981-1986
Low: <5.0 ppb
Medium: >5.0 to <10.0 ppb
High: <10.0 ppb
Increase in total birth defects (n = 9)
OR = 5.87, 95% CI = 0.59-58.81
ATSDR, 2001
6/22/2009
819
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
Births to residents of Endicott, NY
indoor air from soil vapor: 0.18-140
No increase in total birth defects:
ATSDR, 2006, 2008
1983-2000 19
mg/m3
RR = 1.08, 95% CI = 0.82-1.42
Increase in total cardiac defects:
RR = 1.94, 95% CI = 1.21-3.12
Increase in major cardiac defects:
RR = 2.52, 95% CI = 1.2-5.29
Increase in conotruncal heart defects:
RR = 4.83, 95% CI = 1.81-12.89
81,532 pregnancies among
55 ppb TCE, along with many other
No increase in total birth defects: >10 ppb: OR = 1.12
Bove, 1996; Bove et al.,
residents of 75 New Jersey towns
compounds
Increase in total CNS defects at high dose
1995
1985-1988
>1-5 ppb: OR = 0.93, 90% CI = 0.47-1.77
>10 ppb: OR = 1.68, 90% CI = 0.76-3.52
Increase in neural tube defects
>1-5 ppb: OR = 1.58, 90% CI = 0.69-3.40
>10 ppb: OR = 2.53, 90% CI = 0.91-6.37
Increase in oral clefts:
>5 ppb: OR = 2.24, 95% CI = 1.16-4.20
Increase in major cardiac defects:
>10 ppb: OR = 1.24, 50% CI = 0.75-1.94
Increase in ventrical septal defects
>5ppb: OR = 1.30, 95% CI = 0.88-1.87
1,623 children < 20 years old
8.9 and 29 ppb TCE in drinking
Increase in deaths due to congenital anomalies in East
ADHS, 1988
dying from congenital anomalies
water
Central Phoenix
in Maricopa County, AZ
1966-1969: RR = 1.4, 95% CI = 1.1-1.7
1966-1986
1970-1981: RR = 1.5, 95% CI = 1.3-1.7
1982-1986: RR = 2.0, 95% CI = 1.5-2.5
191440 births reported for years 1978-2002, but number not reported for years 1983-2000.
6/22/2009
820
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
4,025 infants born with congenital
heart defects in Milwaukee, WI
1997-1999
maternal residence within 1.32 miles
from at least one TCE emissions
source
Increase in congenital heart defects for mothers >38 years
old
Exposed: OR = 6.2, 95% CI = 2.6-14.5
Unexposed: OR = 1.9, 95% CI = 1.1-3.5
No increase in congenital heart defects for exposed
mothers < 38 years old: OR = 0.9, 95% CI = 0.6-1.2
Yauck et al., 2004
12 children exposed to TCE in
well water in Michigan
5-10 years to 8-14 ppm
1 born with multiple birth defects
Bernadetal., 1987,
abstract
Other adverse birth outcomes
34 live births for which inhalation
of TCE for anesthesia was used in
Japan 1962-1697
2-8 mL (mean 4.3 mL) for 2-98
min (mean: 34.7 min)
1 case of asphyxia; 3 "sleepy babies" with Apgar scores
of 5-9. Delayed appearance of newborn reflexes
Beppu, 1968
51 UK women whose fetus was
considered to be at risk for
hypoxia during labor administered
TCE as an analgesic (50 controls)
amount and route of exposure not
reported
TCE caused fetal pH to fall more, base deficit increased
more, and P02 fell more than the control group by 4-fold
or more compared to other analgesics used
Phillips and Macdonald,
1971
Postnatal Developmental Outcomes
Developmental neurotoxicity
6/22/2009
821
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
54 individuals from 3 residential
Woburn, MA
Woburn, MA
White etal., 1997
cohorts in the US exposed to TCE
63-400 ppb for <1-12 yrs
Verbal naming/language impairment in 6/13 children
in drinking water
Alpha, OH
(46%)
3.3-330 ppb for 5-17 yrs
Alpha, OH
Twin Cities, MN
Verbal naming/language impairment in 1/2 children
261-2,440 ppb for 0.25-25 yrs
(50%)
Twin Cities, MN
Verbal naming/language impairment in 4/4 children
(100%)
Memory impairment in 4/4 children (100%)
Academic impairment in 4/4 children (100%)
Moderate encephalopathy in 4/4 children (100%)
Poor performance on reading/spelling test in 3/4
children (75%)
Poor performance on information test in 3/4 children
(75%)
284 cases of autism spectrum
births geocoded to census tracts, and
Increase in autism spectrum disorder (ASD)
Windham et al., 2006
disorder (ASD) diagnosed <9
linked to hazardous air pollutants
upper 3rd quartile: OR = 1.37, 95% CI = 0.96-1.95
years old and 657 controls born in
(HAPs) data
upper 4th quartile: OR = 1.47, 95% CI = 1.03-2.08
the San Francisco Bay Area 1994
948 children (<18 years) in the
0.4 to >5,000 ppb TCE
Increase in speech impairment:
ATSDR, 2003a; Burg et
Trichloroethylene Subregistry
0-9 years old: RR = 2.45, 99% CI = 1.31-4.58
al., 1995; Burg and Gist,
10-17 years old: RR = 1.14, 99% CI = 0.46-2.85
1999
Increase in hearing impairment:
0-9 years old: RR = 2.13, 99% CI = 1.12-4.07
10-17 years old: RR = 1.12, 99% CI = 0.52-2.24
6/22/2009
822
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
12 children exposed to TCE in
well water in Michigan
5-10 years to 8-14 ppm
9 of 12 children (75%) had poor learning ability,
aggressive behavior, and low attention span
Bernad et al., 1987,
abstract
Developmental immunotoxicitv
200 children aged 36 months old
born prematurely20 and at risk of
atopy21 inLepzig, Germany
1995-1996
Median air level in child's bedroom:
0.42 |ig/m3
No association with allergic sensitization to egg white and
milk, or to cytokine producing peripheral T cells
Lehmann et al., 2001
85 healthy 22full-term neonates
born in Lepzig, Germany
1997-1999
Median air level in child's bedroom
3-4 weeks afterbirth: 0.6 |ig/m3
Significant reduction of Thl IL-2 producing T cells
Lehmann et al., 2002
Other developmental outcomes
55 children (6 months to 10 years
old) were anesthetized for
operations to repair developmental
defects of the jaw and face in
Poland 1964
>10 mL TCE
Reports of bradycardia, accelerated heart rate, and
respiratory acceleration observed; no arrhythmia was
observed
Jasinka, 1965, translation
Childhood Cancer
98 children (< 10 years old)
diagnosed with brain tumors in
Los Angeles County 1972-1977
Questionnaire of parental
occupational exposures
Two cases were reported for TCE exposure, one with
methyl ethyl ketone
Peters et al., 1981
20 Premature defined as 1500-2500 g at birth.
21 Risk of atopy defined as cord blood IgE >0.9 kU/L; double positive family atopy history.
22 Healthy birth defined as > 2500 g and > 37 weeks gestation.
6/22/2009
823
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
22 children (<19 years old)
diagnosed with neuroblastoma in
US and Canada 1992-1994 (12
controls)
Questionnaire of parental
occupational exposures
Increase in neuroblastoma after paternal exposure
OR = 14, 95%CI = 0.7-2.9
Maternal exposure not reported.
De Roos et al., 2001
61 boys and 62 girls (<10 years
old) diagnosed with leukemia and
123 controls in Los Angeles
County 1980-1984
Questionnaire of parents for
occupational exposure
Increase in leukemia after paternal exposure
Preconception (1 year): OR = 2.0, p = 0.16
Prenatal: OR = 2.0, p = 0.16
Postnatal: OR = 2.7, p = 0.7
Maternal exposure not reported.
Lowengart et al., 1987
1,842 children (<15 years old)
diagnosed with ALL in US and
Canada 1989-1993 (1986
controls)
Questionnaire of parents for
occupational exposure
Increase in ALL after maternal exposure
Preconception: OR = 1.8, 95% CI = 0.6-5.2
Pregnancy: OR = 1.8, 95% CI = 0.5-6.4
Postnatal: OR = 1.4, 95% CI = 0.5-4.1
Anytime: OR = 1.8, 95% CI = 0.8-4.1
No increase in ALL after paternal exposure
Anytime: OR = 1.1, 95% CI = 0.8-1.5
Shu et al., 1999
109 children (<15 years old) born
in UK 1974-1988 (218 controls)
Questionnaire of parents for
occupational exposure
Increase in leukemia and NHL after paternal exposure
Preconception: OR = 2.27, 95% CI = 0.84-6.16
Prenatal: OR = 4.40, 95% CI = 1.15-21.01
Postnatal: OR = 2.66, 95% CI = 0.82-9.19
No increase in leukemia and NHL after maternal exposure
Preconception: OR = 1.16, 95% CI = 0.13-7.91
McKinney et al., 1991
6/22/2009
824
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
22 children (<15 years old)
diagnosed with childhood cancer
in California 1988-1998
0.09-97 ppb TCE in drinking water
No increase in total cancer: SIR = 0.83, 99%
CI = 0.44-1.40
No increase in CNS cancer: SIR = 1.05, 99%
CI = 0.24-2.70
No increase in leukemia: SIR = 1.09, 99% CI = 0.38-2.31
Morgan and Cassady,
2002
1,190 children (<20 years old)
diagnosed with leukemia in 4
counties in New Jersey
1979-1987
0-67 ppb TCE in drinking water
Increase in ALL in girls with >5 ppb exposure
<20 years old: RR = 3.36, 95% CI = 1.29-8.28
<5 years old: RR = 4.54, 95% CI = 1.47-10.6
Cohnetal., 1994
24 children (<15 years old)
diagnosed with leukemia in
Woburn, MA 1969-1997
267 |ig/L TCE in drinking water,
along with tetrachloroethylene,
arsenic, and chloroform
Increase in childhood leukemia
Preconception: ORadj = 2.61, 95% CI = 0.47-14.97
Pregnancy: ORadj = 8.33, 95% CI = 0.73-94.67
Postnatal: ORadj = 1.18, 95% CI = 0.28-5.05
Ever: ORadj = 2.39, 95% CI = 0.54-10.59
Costas et al., 2002;
Cutler etal., 1986;
Lagakos et al., 1986;
MADPH, 199723
347 children (<20 years old)
diagnosed with cancer in Endicott,
NY 1980-2001
indoor air from soil vapor: 0.18-140
mg/m3
No increase in cancer (<6 cases, similar to expected)
ATSDR, 2006, 2008
189 children (<20 years old)
diagnosed with cancer in
Maricopa County, AZ 1965-1990
8.9 and 29 ppb TCE in drinking
water
Increase in leukemia:
1965-1986: SIR = 1.67, 95% CI = 1.20-2.27
1982-1986: SIR= 1.91, 95% CI = 1.11-3.12
No increase in total childhood cancers, lymphoma,
brain/CNS, or other cancers
ADHS, 1988, 1990a,
199724
23 Only results from Costas et al. (2002) are reported in the table.
24 Only results from ADHS, 1990a are reported in the table.
6/22/2009
825
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Subjects
Exposure
Effect
Reference
16 children (< 20 years old)
TCE, TCA, and other contaminants
No increase in leukemia: SIR = 0.85, 95% CI = 0.50-1.35
ADHS, 1990b
diagnosed with cancer in East
in drinking water
Phoenix, AZ 1965-1986
37 children (< 20 years old)
1.1-239 ppb TCE, along with 1,1-
Increase in leukemia (n = 11):
ADHS, 1990c
diagnosed with cancer in Pima
DCE, chloroform and chromium in
SIR = 1.50, 95% CI = 0.76-2.70
County, AZ 1970-1986
drinking water
No increase in testicular cancer (n = 6):
SIR = 0.78, 95% CI = 0.32-1.59
No increase in lymphoma (n = 2):
SIR = 0.63, 95% CI = 0.13-1.80
No increase in CNS/brain cancer (n = 3):
SIR = 0.84, 95% CI = 0.23-2.16
Increase in other cancer (n = 15):
SIR = 1.40, 95% CI = 0.79-2.30
6/22/2009
826
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.7.3.2 Animal developmental toxicology studies
A number of animal studies have been conducted to assess the potential for
developmental toxicity of TCE. These include studies conducted in rodents by prenatal
inhalation or oral exposures (summarized in Tables 4.7-12 and 4.7-14), as well as assessments in
non-mammalian species (e.g., avian, amphibian, and invertebrate species) exposed to TCE
during development. Studies have been conducted that provide information on the potential for
effects on specific organ systems, including the developing nervous, immune, and pulmonary
systems. Additionally, a number of research efforts have focused on further characterization of
the mode of action for cardiac malformations that have been reported to be associated with TCE
exposure.
4.7.3.2.1 Mammalian studies
Studies that have examined the effects of TCE on mammalian development following
either inhalation or oral exposures are described below and summarized in Tables 4.7-12 and
4.7-14, respectively.
4.7.3.2.1.1 Inhalation exposures
Dorfmueller et al. (1979) conducted a study in which TCE was administered by
inhalation exposure to groups of approximately 30 female Long-Evans hooded rats at a
concentration of 1,800 ± 200 ppm before mating only, during gestation only, or throughout the
pre-mating and gestation periods. Half of the dams were killed at the end of gestation and half
were allowed to deliver. There were no effects on body weight change or relative liver weight in
the dams. The number of corpora lutea, implantation sites, live fetuses, fetal body weight,
resorptions, and sex ratio were not affected by treatment. In the group exposed only during
gestation, a significant increase in four specific sternebral, vertebral, and rib findings, and a
significant increase in displaced right ovary were observed upon fetal skeletal and soft tissue
evaluation. Mixed function oxidase enzymes (ethoxycoumarin and ethoxyresorbin) which are
indicative of cytochrome P-450 and P-448 activities, respectively, were measured in the livers of
dams and fetuses, but no treatment-related findings were identified. Postnatal growth was
significantly (p < 0.05) decreased in the group with gestation-only exposures. Postnatal
behavioral studies, consisting of an automated assessment of ambulatory response in a novel
environment on postnatal days 10, 20 and 100, did not identify any effect on general motor
activity of offspring following in utero exposure to TCE.
6/22/2009
827
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
In a study by Schwetz et al. (1975), pregnant Sprague-Dawley rats and Swiss Webster
mice (30-40 dams/group) were exposed to TCE via inhalation at a concentration of 300 ppm for
7 hours/day on gestation days 6-15. The only adverse finding reported was a statistically
significant 4-5% decrease in maternal rat body weight. There were no treatment related effects
on pre- and post-implantation loss, litter size, fetal body weight, crown-rump length, or external,
soft tissue, or skeletal findings.
Hardin et al. (1981) summarized the results of inhalation developmental toxicology
studies conducted in pregnant Sprague-Dawley rats and New Zealand white rabbits for a number
of industrial chemicals, including TCE. Exposure concentrations of 0 or 500 ppm TCE were
administered for 6-7 hours per day, on gestations days 1-19 (rats) or 1-24 (rabbits), and
cesarean sections were conducted on gestation days 21 or 30, respectively. There were no
adverse findings in maternal animals. No statistically significant increase in the incidence of
malformations was reported for either species; however the presence of hydrocephaly in two
fetuses of two TCE-treated rabbit litters was interpreted as a possible indicator of teratogenic
potential.
Healy et al. (1982) did not identify any treatment-related fetal malformations following
inhalation exposure of pregnant inbred Wistar rats to 0 or 100 ppm (535 mg/m3) on GD 8-21. In
this study, significant differences between control and treated litters were observed as an
increased incidence of total litter loss (p < 0.05), decreased mean fetal weight (p < 0.05), and
increased incidence of minor ossification variations (p = 0.003) (absent or bipartite centers of
ossification).
Carney et al. (2006) investigated the effects of whole-body inhalation exposures to
pregnant Sprague-Dawley rats at nominal (and actual) chamber concentrations of 0, 50, 150, or
600 ppm TCE for 6 hours/day, 7 days/week on gestation days 6-20. This study was conducted
under Good Laboratory Practice (GLP) regulations according to current EPA and OECD
regulatory testing guidelines (i.e., OPPTS 870.3700 and OECD GD 414). Maternal toxicity
consisted of a statistically significant decrease (22%) in body weight gain during the first 3 days
of exposure to 600 ppm TCE, establishing a no-observed-effect concentration (NOEC) of 150
ppm for dams. No significant difference between control and TCE-treated groups was noted for
pregnancy rates, number of corpora lutea, implantations, viable fetuses per litter, percent pre-
and post-implantation loss, resorption rates, fetal sex ratios, or gravid uterine weights. External,
soft tissue, and skeletal evaluation of fetal specimens did not identify any treatment-related
effects. No cardiac malformations were identified in treated fetuses. The fetal NOEC for this
study was established at 600 ppm.
Westergren et al. (1984) examined brain specific gravity of NMRI mice pups following
developmental exposures to TCE. Male and female mice were separately exposed 24 hours/day
6/22/2009
828
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(except for limited periods of animal husbandry activities) to 0 or 150 ppm TCE for 30 days and
mated during exposure for 7 days. Exposure of the females was continued throughout gestation,
until the first litter was born. Offspring (6-12/group; litter origin not provided in report) were
terminated by decapitation on PND 1, 10, 21-22, or 30. The specific gravity of the brain frontal
cortex, cortex, occipital cortex, and cerebellum were measured. The cortex specific gravity was
significantly decreased at PND 1 (p < 0.001) and 10 (p < 0.01) in pups from exposed mice.
There were also significant differences (p < 0.05) in the occipital cortex and cerebellum at PND
20-22. This was considered suggestive of delayed maturation. No significant differences
between control and treated pups were observed at one month of age.
Table 4.7.12. Summary of mammalian in vivo developmental toxicity studies - inhalation
exposures
Reference
Species/strain/
sex/number
Exposure
level/
duration
NOAEL; LOAEL a
Effects
Carney et
al., 2006
Rat, Sprague-
Dawley,
females, 27
dams/group
0, 50, 150,
600 ppm
(600 ppm =
3.2 mg/L) b
6 hr/day;
GD 6-20
Mat. NOAEL: 150 ppm
Mat. LOAEL: 600 ppm
1BW gain (22% less than control) on
GD 6-9 at 600 ppm.
Dev. NOAEL: 600 ppm
No evidence of developmental toxicity,
including heart defects.
Dorfmueller
et al., 1979
Rat, Long-
Evans,
females, 30
dams/group
0, 1,800 +
200 ppm
(9,674 +
1,075
mg/m3)b
2 weeks, 6
hr/d, 5 d/wk;
prior to
mating
and/or on
GD 0-20
Mat. NOAEL: 1,800 + 200
ppm
No maternal abnormalities.
Dev. LOAEL: 1,800+200
ppm
Sig. T skeletal and soft tissue anomalies
in fetuses from dams exposed during
pregnancy only. No sig. treatment
effects on behavior of offspring 10, 20,
or 100 d postpartum. BW gains sig. 1
in pups from dams with pregestational
exposure.
Hardin et
al., 1981
Rat, Sprague-
Dawley,
female,
nominal
30/group
0, 500 ppm
6-7 hrs/day;
GD 1-19
Mat. NOAEL: 500 ppm
No maternal toxicity
Dev. NOAEL: 500 ppm
No embryonic or fetal toxicity
Rabbit, New
0, 500 ppm
Mat. NOAEL: 500 ppm
No maternal toxicity
6/22/2009
829
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
sex/number
Exposure
level/
duration
NOAEL; LOAEL a
Effects
Zealand white,
female,
nominal
20/group
6-7 hrs/day;
GD 1-24
Dev. LOAEL: 500 ppm
Hydrocephaly observed in 2 fetuses of 2
litters, considered equivocal evidence of
teratogenic potential.
Healy et al.,
1982
Rat, Wistar,
females, 31-32
dams/group
0, 100 ppm
4 hr/day;
GD 8-21
Mat. NOAEL: 100 ppm
No maternal abnormalities.
Dev. LOAEL: 100 ppm
Litters with total resorptions sig. T. Sig.
1 fetal weight, and T bipartite or
absent skeletal ossification centers.
Schwetz
et al.,
1975
Rat, Sprague-
Dawley,
female,
20-35/group
Mouse, Swiss-
Webster,
females, 30-40
dams/group
0, 300 ppm
7 hr/day;
GD 6-15
Mat. LOAEL: 300 ppm
4-5% 1 maternal BW
Dev. NOAEL: 300 ppm
No embryonic or fetal toxicity;
not teratogenic
Westergren
etal., 1984
Mouse, NMRI,
male and
female, 6-12
offspring/group
0, 150 ppm
24 hr/day;
30 days
(during 7
days of
mating and
until GD 22)
Dev. LOAEL: 150 ppm0
Specific gravity of brains sig. -i-
at PND 0, 10, and 20-22.
Similar effects at PND 20-22 in
occipital cortex and cerebellum.
No effects at 1 month of age.
a NOAEL (No Observed Adverse Affect Level) and LOAEL (Lowest Observed Adverse Affect Level) are based
upon reported study findings. Mat. = Maternal; Dev. = Developmental.
b Dose conversions provided by study author(s).
0 Parental observations not reported.
4.7.3.2.1.2 Oral exposures
A screening study conducted by Narotsky and Kavlock (1995) assessed the
developmental toxicity potential of a number of pesticides and solvents, including TCE. In this
study, Fischer 344 rats were administered TCE by gavage at 0, 1,125, and 1,500 mg/kg-day on
gestation days 6-19, and litters were examined on postnatal days 1, 3, and 6. TCE-related
increased incidences of full-litter resorptions, decreased litter sizes, and decreased mean pup
birth weights were observed at both treatment levels. Additionally, TCE treatment was reported
6/22/2009
830
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
to be associated with increased incidences of eye abnormalities (microphthalmia or
anophthalmia). Increased incidences of fetal loss and percent pups with eye abnormalities were
confirmed by Narotsky et al. (1995) in a preliminary dose-setting study that treated Fischer 344
rats with TCE by gavage doses of 475, 633, 844, or 1,125 mg/kg-day on gestation days 6-15,
and then in a 5 x 5 x 5 mixtures study that used TCE doses of 0, 10.1, 32, 101, and 320 mg/kg-
day on GD 6-15. In both studies, dams were allowed to deliver, and pups were examined
postnatally. The incidence of ocular defects observed across all TCE treatment levels tested is
presented in Table 4.7-13,
Table 4.7.13. Ocular defects observed (Narotsky et al., 1995)
Incidence
Dose TCE (mg/kg-day)
(no. affected pups/
Percent pups
total no. pups)a
with eye defects
0
1/197
0.51
10.1
0/71
0.00
32
0/85
0.00
101
3/68
4.41
320
3/82
3.66
475
6/100
6.00
633
6/100
6.00
844
7/58
12.07
1,125
12/44
27.27
a Reported in Barton and Das (1996)
Other developmental findings in this study included increased full litter resorption at 475, 844,
and 1,125 mg/kg-day; increased postnatal mortality at 425 mg/kg-day. Pup body weights were
decreased (not significantly) on PND 1 and 6 at 1,125 mg/kg-day. In both the Narotsky and
Kavlock (1995) and Narotsky et al. (1995) studies, significantly decreased maternal body weight
gain was observed at the same treatment levels at which full litter resorption was noted.
Additionally, in Narotsky et al. (1995) maternal observations included delayed parturition at 475,
844, and 1,125 mg/kg-day, ataxia at 633 mg/kg-day, and mortality at 1,125 mg/kg-day.
Cosby and Dukelow (1992) administered TCE in corn oil by gavage to female B6D2F1
mice (28-62/group) on gestation days 1-5, 6-10, or 11-15 (where mating = GDI). Dose levels
were 0, 1/100 and 1/10 of the oral LD50 (i.e., 0, 24.02, and 240.2 mg/kg body weight). Dams
were allowed to deliver; litters were evaluated for pup count sex, weight, and crown-rump length
until weaning (PND 21). Some litters were retained until 6 weeks of age at which time gonads
6/22/2009
831
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(from a minimum of 2 litters/group) were removed, weighed, and examined. No treatment-
related reproductive or developmental abnormalities were observed.
A single dose of TCE was administered by gavage to pregnant CD-I mice (9-19/group)
at doses of 0, 0.1, or 1.0 (J,g/kg in distilled water, or 0, 48.3, or 483 mg/kg in olive oil, 24 hours
after pre-mating human chorionic gonadotropin (hCG) injection (Coberly et al., 1992). At 53
hours after the hCG-injection, the dams were terminated, and the embryos were flushed from
excised oviducts. Chimera embryos were constructed, cultured, and examined. Calculated
proliferation ratios did not identify any differences between control and treated blastomeres. A
lack of treatment-related adverse outcome was also noted when the TCE was administered by i.p.
injection to pregnant mice (16-39/group) at 24 and 48 hours post-hCG at doses of 0, 0.01, 0.02,
or 10 (J,g/kg body weight.
In a study intended to confirm or refute the cardiac teratogenicity of TCE that had been
previously observed in chick embryos, Dawson et al. (1990) continuously infused the gravid
uterine horns of Sprague-Dawley rats with solutions of 0, 15 or 1,500 ppm TCE (or 1.5 or 150
ppm dichloroethylene) on gestation Days 7-22. At terminal cesarean section on gestation day
22, the uterine contents were examined, and fetal hearts were removed and prepared for further
dissection and examination under a light microscope. Cardiac malformations were observed in
3% of control fetuses, 9% of the 15 ppm TCE fetuses (p = 0.18), and 14% of the 1,500 ppm TCE
fetuses, (p = 0.03). There was a >60% increase in the percent of defects with a 100-fold increase
in dose. No individual malformation or combination of abnormalities was found to be
selectively induced by treatment.
To further examine these TCE-induced cardiac malformations in rats, Dawson et al.
(1993) administered 0, 1.5 or 1,100 ppm TCE in drinking water to female Sprague-Dawley rats.
Experimental treatment regimens were: 1) a period of approximately 2 months prior to
pregnancy plus the full duration of pregnancy, 2) the full duration of pregnancy only, or 3) an
average of 3 months before pregnancy only. The average total daily doses of TCE consumed for
each exposure group at both dose levels were:
Group 1
Group 2
Group 3
1.5 ppm
23.5 |j,l
0.78 |^1
3.97 |^1
1,100 ppm
1,206 |^1
261 |^1
1,185 |^1
The study also evaluated 0, 0.15, or 110 ppm dichloroethylene in drinking water, with treatment
administered 1) two months prior to pregnancy plus the full duration of pregnancy, or 2) an
average of 2 months before pregnancy only. At terminal cesarean section, uterine contents were
6/22/2009
832
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
examined, fetuses were evaluated for external defects, and the heart of each fetus was removed
for gross histologic examination under a dissecting microscope, conducted without knowledge of
treatment group. There were no differences between TCE-treated and control group relative to
percentage of live births, implants, and resorptions. The percentage of cardiac defects in TCE-
treated groups ranged from 8.2% to 13.0%, and was statistically significant as compared to the
control incidence of 3%. The dose-response was relatively flat, even in spite of the extensive
difference between the treatment levels. There was a broad representation of various types of
cardiac abnormalities identified, notably including multiple transposition, great artery, septal,
and valve defects (Table 4.7-14). No particular combination of defects or syndrome
predominated. Exposure before pregnancy did not appear to be a significant factor in the
incidence of cardiac defects.
Table 4.7.14. Types of congenital cardiac defects observed in TCE-exposed fetuses
(Dawson et al., 1993, Table 3)
Cardiac Abnormalities Control
d-transposition (right chest) 2
1-transposition (left chest
Great artery defects
Atrial septal defects 1
Mitral valve defects
Tricuspid valve defects
Ventricular septal defects
Subaortic 1
Membranous
Muscular 2
Endocardial cushion defect 1
Pulmonary valve defects
Aortic valve defects
Situs inversus
Total abnormalities 7
Total abnormal hearts 7
Premating
1,100 ppm 1.5 ppm
TCE Concentrations
Premating/Gestation
1,100 ppm 1.5 ppm
9
9
1
19
5
1
4
2
4
2
2
1
41
40
2
2
5
Gestation Only
1,100 ppm 1.5 ppm
1
1
7 4
4
1
1
2
23
23
15
11
10
9
In an attempt to determine a threshold for cardiac anomalies following TCE exposures,
Johnson et al. (2003, 2005) compiled and reanalyzed data from five studies conducted from
1989-1995. In these studies, TCE was administered in drinking water to Sprague-Dawley rats
throughout gestation (i.e., a total of 22 days) at levels of 2.5 ppb (0.0025 ppm), 250 ppb (0.25
ppm), 1.5 ppm, or 1,100 ppm. The dams were terminated on the last day of pregnancy and
6/22/2009
833
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
fetuses were evaluated for abnormalities of the heart and great vessels. The control data from the
five studies were combined prior to statistical comparison to the individual treated groups, which
were conducted separately. The study author reported that significant increases in the percentage
of abnormal hearts and the percentage of litters with abnormal hearts were observed in a
generally dose-responsive manner at 250 ppb and greater (Table 4.7-15).
Table 4.7.15. Types of heart malformations per 100 fetuses (Johnson et al., 2003, Table 2, p
290)
TCE dose
group
Type of defect/100 fetuses
Control
1,100 ppm
1.5 ppm
250 ppb
2.5 ppt
Abnormal looping
0.33
1
Coronary artery/sinus
1.82
Aortic hypoplasia
0.55
Pulmonary artery hypoplasia
0.55
Atrial septal defect
1.16
6.67
2.21
0.91
Mitral valve defect
0.17
0.91
Tricuspid valve defect
0.91
Ventricular septal defect
Perimembranous (subaortic)
0.33
2.86
1.66
Muscular
0.33
0.95
0.55
Atriventricular septal defect
0.17
0.95
Pulmonary valve defect
Aortic valve defects
1.9
0.91
Fetuses with abnormal hearts (n)
13
11
9
5
0
Total fetuses (n)
606
105
181
110
144
Litters with fetuses with
9/55
6/9
5/13
4/9
0/12
abnormal hearts/litter (n)
Litter with fetuses with abnormal
16.4
66.7
38.5
44.4
0.0
hearts/no. litters (%)
In a study by Fisher et al. (2001), pregnant Sprague-Dawley rats were administered daily
gavage doses on GD 6-15 of TCE (500 mg/kg-day), TCA (300 mg/kg-day), or DCA (300
mg/kg-day). Cesarean delivery of fetuses was conducted on GD 21. Water and soybean oil
negative control groups, and a retinoic acid positive control group were also conducted
simultaneously. Maternal body weight gain was not significantly different from control for any
of the treated groups. No significant differences were observed for number of implantations,
resorptions, or litter size. Mean fetal body weight was reduced by treatment with TCA and
DCA. The incidence of heart malformations was not significantly increased in treated groups as
compared to controls. The fetal rate of cardiac malformations ranged from 3 to 5% across the
6/22/2009
834
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE, TCA, and DCA dose groups and from 6.5% to 2.9% for the soybean and water control
dose groups, respectively. It was suggested that the apparent differences between the results of
this study and the Dawson et al. (1993) study may be related to factors such as differences in
purity of test substances or in the rat strains, or differences in experimental design (e.g., oral
gavage versus drinking water, exposure only during the period of organogenesis versus during
the entire gestation period, or the use of a staining procedure). The rats from this study were also
examined for eye malformations to follow-up on the findings of Narotsky (1995). As reported in
Warren et al. (2006), gross evaluation of the fetuses as well as computerized morphometry
conducted on preserved and sectioned heads revealed no ocular anomalies in the groups treated
with TCE. This technique allowed for quantification of the lens area, global area, medial canthus
distance, and interlocular distance. DCA treatment was associated with statistically significant
reductions in the lens area, globe area, and interlocular distance. All four measures were reduced
in the TCA-treated group, but not significantly. The sensitivity of the assay was demonstrated
successfully with the use of a positive control group that was dosed on GD 6-15 with a known
ocular teratogen, retinoic acid (15 mg/kg-day).
Johnson et al. (1998a, b) conducted a series of studies to determine whether specific
metabolites of TCE or dichloroethylene were responsible for the cardiac malformations observed
in rats following administration during the period of organogenesis. Several metabolites of the
two chemicals were administered in drinking water to Sprague-Dawley rats from GD 1-22.
These included carboxy methylcystine, dichloroacetaldehyde, dichlorovinyl cystine,
monochloroacetic acid, trichloroacetic acid, trichloroacetaldehyde, and trichloroethanol.
Dichloroacetic acid, a primary common metabolite of TCE and dichloroethylene, was not
included in these studies. The level of each metabolite administered in the water was based upon
the dosage equivalent expected if 1,100 ppm (the limit of solubility) TCE broke down
completely into that metabolite. Cesarean sections were performed on GD 22, uterine contents
were examined, and fetuses were processed and evaluated for heart defects according to the
procedures used by Dawson et al. (1993). No treatment-related maternal toxicity was observed
for any metabolite group. Adverse fetal outcomes were limited to significantly increased
incidences of fetuses with abnormal hearts (Table 4.7-16). Significant increases in fetuses with
cardiac defects (on a per-fetus and per-litter basis) were observed for only one of the metabolites
evaluated, i.e., trichloroacetic acid (2,730 ppm, equivalent to a dose of 291 mg/kg-day).
Notably, significant increases in fetuses with cardiac malformations were also observed with 1.5
or 1,100 ppm TCE (0.218 or 129 mg/kg-day), or with 0.15 or 110 ppm DCE (0.015 or 10.64
mg/kg-day), but in each case only with prepregnancy-plus-pregnancy treatment regimens. The
cardiac abnormalities observed were diverse and did not segregate to any particular anomaly or
grouping. Dose related increases in response were observed for the overall number of fetuses
6/22/2009
835
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
with any cardiac malformation for both TCE and DCE; however no dose-related increase
occurred for any specific cardiac anomaly (Johnson et al., 1998b).
Table 4.7.16. Congenital cardiac malformations (Johnson et al., 1998b, Table 2, p. 997)
Treatment Group
Normal
TCE
TCE
TCE
DCE
DCE
TCAA
MCAA
TCEth
TCAld
DCAld
CMC
DCVC
water
p+p
p+p
P
p+p
p+p
P
P
P
P
P
P
P
Heart abnormalities
1,100
1.5
1,100
110
0.15
2,730
1,570
1,249
1,232
174
473
50
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
Abnormal looping
2
-
2
-
-
-
-
-
-
-
-
-
-
Aortic hypoplasia
-
1
1
-
1
-
1
-
1
-
1
-
1
Pulmonary artery
-
-
1
-
-
-
2
1
-
-
2
-
-
hypoplasia
Atrial septal defects
7
19
5
7
11
7
3
3
-
2
-
-
1
Mitral valve defects,
1
5
8
-
4
3
1
-
1
2
-
-
1
hypoplasia or ectasia
Tricuspid valve defects,
-
1
1
-
1
-
-
-
1
-
-
-
-
hypoplasia or ectasia
Ventricular septal
defects
Perimembranous a
2
6
2
1
4
1
4
-
-
3
-
1
-
Muscular
2
4
-
4
2
1
1
-
1
-
-
2
2
Atrioventricual septal
1
_
_
1
1
_
_
_
_
_
_
_
_
defects
Pulmonary valve defects
-
2
1
-
1
-
1
3
1
1
-
-
-
Aortic valve defects
-
2
2
2
2
3
-
-
1
-
-
1
-
Situs inversus
-
1
-
-
-
-
-
-
-
-
-
-
-
Total
Abnormal hearts
15
41
23
15
25
15
13
7
6
8
3
4
5
Fetuses with abnormal
13
40*
22*
11*
24*
14*
12*
6
5
8
3
4
5
hearts
Fetuses
605
434
255
105
184
121
114
132
121
248
101
85
140
a Subaortic
p+p = pregnancy and preprenancy; p = pregnancy
* Per-fetus statistical significance (Fisher exact test).
The TCE metabolites trichloroacetic acid (TCA) and dichloroacetic acid (DCA) were
also studied by Smith et al. (1989, 1992). Doses of 0, 330, 800, 1,200, or 1,800 mg/kg TCA
were administered daily by oral gavage to Long-Evan hooded rats on gestation days 6-15.
Similarly, DCA was administered daily by gavage to Long-Evans rats on GD 6-15 in two
separate studies, at 0, 900, 1,400, 1,900, or 2,400 mg/kg-day and 0, 14, 140, or 400 mg/kg-day.
6/22/2009
836
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Embryo lethality and statistically or biologically significant incidences of orbital anomalies
(combined soft tissue and skeletal findings) were observed for TCA at >800 mg/kg-day, and for
DCA at >900 mg/kg-day. Fetal growth (body weight and crown-rump length) was affected at
>330 mg/kg-day for TCE and at >400 mg/kg-day for DCA. For TCA, the most common cardiac
malformations observed were levocardia at >330 mg/kg-day and interventricular septal defect at
>800 mg/kg-day. For DCA, levocardia was observed at >900 mg/kg-day, interventricular septal
defect was observed at >1,400 mg/kg-day, and a defect between the ascending aorta and right
ventricle was observed in all treated groups (i.e., >14 mg/kg-day, although the authors appeared
to discount the single fetal finding at the lowest dose tested). Thus, NOAELs were not
definitively established for either metabolite, although it appears that TCA was generally more
potent than DCA in inducing cardiac abnormalities.
These findings were followed up by a series of studies on DCA reported by Epstein et al.
(1992), which were designed to determine the most sensitive period of development and further
characterize the heart defects. In these studies, Long-Evans hooded rats were dosed by oral
gavage with a single dose of 2,400 mg/kg-day on selected days of gestation (6-8, 9-11, or
12-15); with a single dose of 2,400 mg/kg on days 10, 11, 12, or 13; or with a single dose of
3,500 mg/kg on days 9, 10, 11, 12, or 13. The heart defects observed in these studies were
diagnosed as high interventricular septal defects rather than membranous type interventricular
septal defects. The authors hypothesized that high intraventricular septal defects are a specific
type of defect produced by a failure of proliferating interventricular septal tissue to fuse with the
right tubercle of the atrioventricular cushion tissue. This study identified gestation days 9
through 12 as a particularly sensitive period for eliciting high interventricular septal defects. It
was postulated that DCA interferes with the closure of the tertiary interventricular foramen,
allowing the aorta to retain its embryonic connection with the right ventricle. Further, it was
suggested that the selectivity of DCA in inducing cardiac malformations may be due to the
disruption of a discrete cell population.
TCE and its metabolites dichloroethylene (DCE) and trichloroacetic acid (TCAA) were
administered in drinking water to pregnant Sprague-Dawley rats from gestation days 0-11
(Collier et al., 2003). Treatment levels were: 0, 110, or 1,100 ppm (i.e., 0, 830, or 8,300 (^gM)
TCE; 0, 11, or 110 ppm (i.e., 0, 110, or 1,100 (j,gM) DCE; 0, 2.75, or 27.3 mg/mL (i.e., 0, 10, or
100 mM) TCAA. Embryos (including hearts) were harvested between embryonic days 10.5-11,
since this is the stage at which the developmental processes of myoblast differentiation, cardiac
looping, atrioventricular valve formation, and trabeculation would typically be occurring. A
PCR based subtraction scheme was used to identify genes that were differentially regulated with
TCE or metabolite exposure. Numerous differentially regulated gene sequences were identified.
Up-regulated transcripts included genes associated with stress response (Hsp 70) and
6/22/2009
837
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
homeostasis (several ribosomal proteins). Down-regulated transcripts included extracellular
matrix components (GPI-pl37 and vimentin) and Ca2+ responsive proteins (Serca-2 Ca2+-
ATPase and P-catenin). Serca-2 Ca2+ and GPI-pl37 were identified as two possible markers for
fetal TCE exposure. Differential regulation of expression of these markers by TCE was
confirmed by dot blot analysis and semi-quantitative RT-PCR with decreased expression seen at
levels of TCE exposure between 100 and 250 ppb (0.76 and 1.9 |iM),
Developmental neurotoxicity and developmental immunotoxicity: Several studies were
conducted that included assessments of the effects of TCE oral exposure on the developing
nervous system (Fredriksson et al., 1993; Isaacson and Taylor, 1989; Noland-Gerbec et al., 1986;
George et al., 1986; Dorfmueller et al., 1979; Blossom et al., 2008) or immune system (Peden-
Adams et al., 2006, 2008; Blossom and Doss, 2007; Blossom et al., 2008). These studies,
summarized below, are addressed in additional detail in Section 4.2. (nervous system) and
Section 4.5.2.1.2 (immune system).
Developmental neurotoxicity: Fredriksson et al. (1993) conducted a study in male NMRI
weanling mice (12/group, selected from 3-4 litters), which were exposed to trichloroethylene by
oral gavage at doses of 0 (vehicle), 50, or 290 mg/kg-day TCE in a fat emulsion vehicle, on
postnatal days (PND) 10-16. Locomotor behavior (horizontal movement, rearing and total
activity) were assessed over three 20-minute time periods at postnatal days 17 and 60. There
were no effects of treatment in locomotor activity at PND 17. At PND 60, the mice treated with
50 and 290 mg/kg-day TCE showed a significant (p < 0.01) decrease in rearing behavior at the
0-20 and 20-40 minute time points, but not at the 40-60 minute time point. Mean rearing
counts were decreased by over 50% in treated groups as compared to control. Horizontal activity
and total activity were not affected by treatment.
Open field testing was conducted in control and high-dose F1 weanling Fischer 344 rat
pups in an NTP reproduction and fertility study with continuous breeding (George et al., 1986).
In this study, TCE was administered at dietary levels of 0, 0.15, 0.30, or 0.60%. The open field
testing revealed a significant (p < 0.05) dose-related trend toward an increase in the time required
for male and female pups to cross the first grid in the testing device, suggesting an effect on the
ability to react to a novel environment.
Taylor et al. (1985) administered TCE in drinking water (0, 312, 625, or 1,250 ppm) to
female Sprague-Dawley rats for 14 days prior to breeding, and from gestation day 0 through
offspring postnatal day 21. The number of litters/group was not reported, nor did the study state
how many pups per litter were evaluated for behavioral parameters. Exploratory behavior was
measured in the pups in an automated apparatus during a 15-minute sampling period on PND 28,
60, and 90. Additionally, wheel-running, feeding, and drinking behavior was monitored 24
hours/day on PND 55-60. The number of exploratory events was significantly increased by
6/22/2009
838
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
approximately 25-50% in 60- and 90-day old male TCE-treated rats at all dose levels, with the
largest effect observed at the highest dose level tested, although there were no effects of
treatment on the number of infrared beam-breaks. No difference between control and treated rats
was noted for pups tested on PND 28. Wheel-running activity was increased approximately 40%
in 60-day old males exposed to 1,250 ppm TCE as compared to controls. It is notable that
adverse outcomes reported in the developmentally-exposed offspring on this study were
observed long after treatment ceased.
Using a similar treatment protocol, the effects of TCE on development of myelinated
axons in the hippocampus was evaluated by Isaacson and Taylor (1989) in Sprague-Dawley rats.
Female rats (6/group) were exposed in the drinking water from 14 days prior to breeding and
through the mating period; then the dams and their pups were exposed throughout the prenatal
period and until PND 21, when they were sacrificed. The dams received 0, 312 or 625 ppm (0, 4
or 8.1 mg/day TCE in the drinking water. Myelinated fibers were counted in the hippocampus of
2-3 pups per treatment group at PND 21, revealing a decrease of approximately 40% in
myelinated fibers in the CA1 area of the hippocampus of pups from dams at both treatment
levels, with no dose-response relationship. There was no effect of TCE treatment on myelination
in several other brain regions including the internal capsule, optic tract or fornix.
A study by Noland-Gerbec et al. (1986) examined the effect of pre- and perinatal
exposure to TCE on 2-deoxyglucose (2-DG) uptake in the cerebellum, hippocampus and whole
brain of neonatal rats. Sprague Dawley female rats (9-11/group) were exposed via drinking
water to 0 or 312 mg TCE/liter distilled water from 14 days prior to mating until their pups were
euthanized at postnatal day 21. The total TCE dose received by the dams was 825 mg over the
61 day exposure period. Pairs of male neonates were euthanized on PND 7, 11, 16, and 21.
There was no significant impairment in neonatal weight or brain weight attributable to treatment,
nor were other overt effects observed. 2- DG uptake was significantly reduced from control
values in neonatal whole brain (9-11%) and cerebellum (8—16%) from treated rats at all ages
studied, and hippocampal 2-DG uptake was significantly reduced (7—21% from control) in
treated rats at all ages except at PND 21.
In a study by Blossom et al. (2008), MRL +/+ mice were treated in the drinking water
with 0 or 0.1 mg/mL TCE from maternal GD 0 through offspring PND 42. Based on drinking
water consumption data, average maternal doses of TCE were 25.7 mg/kg-day, and average
offspring (PND 24-42) doses of TCE were 31.0 mg/kg-day. In this study, a subset of offspring
(3 randomly selected neonates from each litter) was evaluated for righting reflex on PNDs 6, 8,
and 10; bar-holding ability on PNDs 15 and 17; and negative geotaxis on PNDs 15 and 17; none
of these were impaired by treatment. In an assessment of offspring nest building on PND 35,
there was a significant association between impaired nest quality and TCE exposure; however,
6/22/2009
839
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE exposure did not have an effect on the ability of the mice to detect social and non-social
odors on PND 29 using olfactory habituation and dishabituation methods. Resident intruder
testing conducted on PND 40 to evaluate social behaviors identified significantly more
aggressive activities (i.e., wrestling and biting) in TCE-exposed juvenile male mice as compared
to controls. Cerebellar tissue homogenates from the male TCE-treated mice had significantly
lower glutathione (GSH) levels and GSH:oxidized GSH (GSH:GSSG) ratios, indicating
increased oxidative stress and impaired thiol status; these have been previously reported to be
associated with aggressive behaviors (Franco et al., 2006). Qualitative histopathological
examination of the brain did not identify alterations indicative of neuronal damage or
inflammation. Although the study author attempted to link the treatment-related alterations in
social behaviors to the potential for developmental exposures to TCE to result in autism in
humans, this association is not supported by data and is considered speculative at this time.
As previously noted, postnatal behavioral studies conducted by Dorfmueller et al. (1979)
did not identify any changes in general motor activity measurements of rat offspring on PND 10,
20, and 100 following maternal gestational inhalation exposure to TCE at 1,800 ± 200 ppm.
Developmental immunotoxicity: Peden-Adams et al. (2006) assessed the potential for
developmental immunotoxicity following TCE exposures. In this study, B6C3F1 mice
(5/sex/group) were administered TCE via drinking water at dose levels of 0, 1,400 or 14,000 ppb
from maternal gestation day 0 to either postnatal 3 or 8, when offspring lymphocyte
proliferation, NK cell activity, SRBC-specific IgM production (PFC response), splenic B220+
cells, and thymus and spleen T-cell immunophenotypes were assessed. (A total of 5-7 pups per
group were evaluated at week 3, and the remainder were evaluated at week 8.) Observed
positive responses consisted of suppressed PFC responses in males at both ages and both TCE
treatment levels, and in females at both ages at 14,000 ppb and at 8 weeks of age at 1,400 ppb.
Spleen numbers of B220+ cells were decreased in 3-week old pups at 14,000 ppb. Pronounced
increases in all thymus T-cell subpopulations (CD4+, CD8+, CD4+/CD8+, and CD4-/CD8-)
were observed at 8-weeks of age. Delayed hypersensitivity response, assessed in offspring at 8
weeks of age, was increased in females at both treatment levels and in males at 14,000 ppb only.
No treatment-related increase in serum anti-dsDNA antibody levels was found in the offspring at
8 weeks of age.
In a study by Blossom and Doss (2007), TCE was administered to groups of pregnant
MRL +/+ mice in drinking water at levels of 0, 0.5 or 2.5 mg/mL. TCE was continuously
administered to the offspring until young adulthood (i.e., 7-8 weeks of age). Offspring post-
weaning body weights were significantly decreased in both treated groups. Decreased spleen
cellularity and reduced numbers of CD4+, CD8+, and B220+ lymphocyte subpopulations were
observed in the post-weaning offspring. Thymocyte development was altered by TCE exposures
6/22/2009
840
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(significant alterations in the proportions of double-negative subpopulations and inhibition of in
vitro apoptosis in immature thymocytes). A dose-dependent increase in CD4+ and CD8+ T-
lymphocyte IFNy was observed in peripheral blood by 4-5 weeks of age, although these effects
were no longer observed at 7-8 weeks of age. Serum anti-histone autoantibodies and total IgG2a
were significantly increased in treated offspring; however, no histopathological signs of
autoimmunity were observed in the liver and kidneys at sacrifice.
Blossom et al. (2008) administered TCE to MRL +/+ mice (8 dams/group) in the drinking
water at levels of 0 or 0.1 mg/mL from GD 0 through offspring postnatal day 42. Average
maternal doses of TCE were 25.7 mg/kg-day, and average offspring (PND 24-42) doses of TCE
were 31.0 mg/kg-day. Subsets of offspring were sacrificed at PND 10 and 20, and thymus
endpoints (i.e., total cellularity, CD4+/CD8+ ratios, CD24 differentiation markers, and double-
negative subpopulation counts) were evaluated. Evaluation of the thymus identified a significant
treatment-related increase in cellularity, accompanied by alterations in thymocyte subset
distribution, at PND 20 (sexes combined). TCE treatment also appeared to promote T cell
differentiation and maturation at PND 42, Indicators of oxidative stress were measured in the
thymus at PND 10 and 20, and in the brain at PND 42,.and ex vivo evaluation of cultured
thymocytes indicated increased reactive oxygen species (ROS) generation. Mitogen-induced
intracellular cytokine production by splenic CD4+ and CD8+ T cells was evaluated in juvenile
mice and brain tissue was examined at PND 42 for evidence of inflammation. Evaluation of
peripheral blood indicated that splenic CD4+ T cells from TCE-exposed PND 42 mice produced
significantly greater levels of IFN-y and IL-2 in males and TNF-a in both sexes. There was no
effect on cytokine production on PND 10 or 20.
Peden-Adams et al. (2008) administered TCE to MRL+/+ mice (unspecified number of
dams/group) in drinking water at levels of 0, 1,400, or 14,000 ppb from gestation day (GD) 0 and
continuing until the offspring were 12 months of age. At 12 months of age, final body weight;
spleen, thymus, and kidney weights; spleen and thymus lymphocyte immunophenotyping (CD4
or CD8); splenic B-cell counts; mitogen-induced splenic lymphocyte proliferation; serum levels
of autoantibodies to dsDNA and glomerular antigen (GA), periodically measured from 4 to 12
months of age; and urinary protein measures were recorded. Reported sample sizes for the
offspring measurements varied from 6 to 10 per sex per group; the number of source litters
represented within each sample was not specified. The only organ weight alteration was an 18%
increase in kidney weight in the 1,400 ppb males. Splenic CD4-/CD8- cells were altered in
female mice (but not males) at 1,400 ppm only. Splenic T-cell populations, numbers of B220+
cells, and lymphocyte proliferation were not affected by treatment. Populations of thymic T-cell
subpopulations (CD8+, CD4-/CD8-, and CD4+) were significantly decreased in male but not
female mice following exposure to 14,000 ppb TCE, and CD4+/CD8+ cells were significantly
6/22/2009
841
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
reduced in males by treatment with both TCE concentrations. Autoantibody levels (anti-dsDNA
and anti-GA) were not increased in the offspring over the course of the study.
Although all of the developmental immunotoxicity studies with TCE (Peden-Adams et
al., 2006, 2008; Blossom and Doss, 2007; Blossom et al., 2008) exposed the offspring during
critical periods of pre- and postnatal immune system development, they were not designed to
assess issues such as post-treatment recovery, latent outcomes, or differences in severity of
response that might be attributed to the early life exposures.
Table 4.7.17. Summary of mammalian in vivo developmental toxicity studies - oral
exposures
Reference
Species/strain/
sex/number
Dose level/
Exposure
duration
Route/vehicle
NOAEL;
LOAEL a
Effects
Blossom &
Mouse, MRL +/+,
0, 0.5, or 2.5
Drinking
Dev.
At 0.5 mg/mL: Sig [
Doss, 2007
dams and both sexes
mg/mL
water
LOAEL=
postweaning weight; sig.f
offspring, 3
0.5 mg/mL 0
IFNy produced by splenic
litters/group, 8-12
CD4+ cells at 5-6 wks; sig
offspring/group
Parental mice
and/or offspring
exposed from GD
0 to 7-8 months
of age
I splenic CD8+and B220+
lymphocytes; sig.f IgG2a
and histone; sig. altered
CD4-/CD8- and
CD4+/CD8+ thymocyte
profile
At 2.5 mg/mL: Sig j
postweaning weight; sig.f
IFNy produced by splenic
CD4+ and CD8+ cells at
4-5 and 5-6 wks; sig j
splenic CD4+, CD8+, and
B220+ lymphocytes; sig.
altered CD4+/CD8+
thymocyte profile
6/22/2009
842
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
sex/number
Dose level/
Exposure
duration
Route/vehicle
NOAEL;
LOAEL a
Effects
Blossom et
al., 2008
Mouse, MRL +/+,
dams and both sexes
offspring, 8
litters/group, 3-8
offspring/group
0 or 0.1
mg/mL
(maternal dose
= 25.7 mg/kg-
day; offspring
PND 24-42
dose - 31.0
mg/kg-day
Parental mice
and/or offspring
exposed from GD
0 to PND 42
Drinking
water
Dev.
LOAEL=
1,400 ppb c
At 0.1 mg/mL: at PND 20,
sig. t thymocyte cellularity
and distribution, associated
with sig. t in thymocyte
subset distribution; sig. t
reactive oxygen species
generation in total
thymocytes; sig. t in
splenic CD4+ T cell
production of IFN-y and
IL-2 in females and TNF-a
in males at PND 42.
Significantly impaired nest-
building behaviors at PND
35. Increased aggressive
activities, and increased
oxidative stress and
impaired thiol status in the
cerebellar tissue of male
offspring at PND 40.
Collier et
al., 2003
Rat, Sprague-Dawley,
female, no. dams/
group not reported
0, 0.11, or 1.1
mg/mL
(0, 830, or 8,300
ligM)b
GD 0-11
Drinking
water
Dev. LOEL:
0.11 mg/mL
Embryos collected between
GD 10.5 and 11. Gene
expression at 1.1 mg/mL
TCE: 8 housekeeping
genes T, and one gene j; 3
stress response genes T, IL
10 |; 2 cyto-skeletal/cell
adhesion/blood related
genes T, 3 genes >1; 2 heart-
specific genes t- Effects at
0.11 mg/mL reduced
considerably. Two
possible markers for fetal
TCE exposure identified as
Serca-2 Ca+2 ATPase and
GPI-pl37.
6/22/2009
843
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
sex/number
Dose level/
Exposure
duration
Route/vehicle
NOAEL;
LOAEL a
Effects
Cosby &
Mouse, B6D2F1,
0, 24, or 240
Gavage in
Mat.
No maternal toxicity.
Dukelow,
female, 28-62
mg/kg-day
corn oil
NOAEL:
1992
dams/group
GD 1-5, 6-10, or
240 mg/kg-
day
11-15
Dev.
NOAEL:
240 mg/kg-
day
No effects on embryonic or
fetal development.
Dawson, et
Rat, Sprague-Dawley,
0, 1.5, or 1,100
Drinking
Mat.
No maternal toxicity.
al., 1993
116 females allocated
to 11 groups
ppm
water
NOAEL:
1,100 ppm
2 mo before
Dev.
Sig. T in heart defects,
mating and/or
LOAEL: 1.5
primarily atrial septal
during gestation
ppm
defects, found at both dose
levels in groups exposed
prior to pregnancy and
during pregnancy, as well
as in group exposed to
1,100 ppm dose during
pregnancy only. No sig. T
in congenital heart defects
in groups exposed prior to
pregnancy only.
Fisher et
Rat, Sprague-Dawley,
0, 500 mg/kg-day
Gavage in
Mat.
No maternal toxicity.
al., 2001;
female, 20-25
soybean oil
NOAEL:
Warren et
dams/group
GD 6-15
500 mg/kg-
al., 2006
day
Dev.
No developmental toxicity.
NOAEL:
The incidence of heart
500 mg/kg-
malformations for fetuses
day
from TCE-treated dams
(3-5%) did not differ from
neg. controls. No eye
defects observed.
6/22/2009
844
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
Dose level/
Route/vehicle
NOAEL;
Effects
sex/number
Exposure
duration
LOAEL
a
Fredriksson
Mouse, NMRI, male
0, 50, or 290
Gavage in a
Dev.
Rearing activity sig. 1 at
etal., 1993
pups, 12 pups from
mg/kg-day
20% fat
LOAEL:
50
both dose levels on PND
3-4 different
emulsion
mg/kg-day
60
litters/group
PND 10-16
prepared from
egg lecithin
and peanut oil
George et
Rat, F334, male and
0, 0.15,0.30 or
Dietary
LOAEL:
Open field testing in pups:
al., 1986
female, 20 pairs/
treatment group,
40 controls/sex
0.60%
microencapsulated
TCE
Breeders exposed
1 wk pre-mating,
then for 13 wk;
pregnant s
throughout
pregnancy (i.e., 18
wk total)
0.15%
a sig. dose-related trend
toward T time required for
male and female pups to
cross the first grid in the
test devise
Isaacson &
Taylor,
1989
Rat, Sprague-Dawley,
females, 6 dams/group
0,312, or 625
mg/L.
(0,4.0, or 8.1
mg/day) b
Dams (and pups)
exposed from 14
days prior to
mating until end
of lactation.
Drinking
water
Dev.
LOAEL:
mg/L 0
312
Sig. X myelinated
fibers in the stratum
lacunosum-moleculare
of pups. Reduction in
myelin in the
hippocampus.
Johnson et
Rat, Sprague-Dawley,
0, 2.5 ppb, 250
Drinking
Dev.
Sig. T in percentage of
al., 2003
female, 9-13/group, 55
ppb, 1.5 ppm, or
water
NOAEL:
2.5
abnormal hearts and the
in control group
1,100 ppm
ppb
Dev.
percentage of litters with
abnormal hearts at >250
(0, 0.00045,
LOAEL:
250
ppb
0.048, 0.218, or
PPb °
129 mg/kg-day) b
GD 0-22
6/22/2009
845
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
sex/number
Dose level/
Exposure
duration
Route/vehicle
NOAEL;
LOAEL a
Effects
Narotsky et
Rat, Fischer 344,
0, 10.1,32, 101,
Gavage in
Mat.
Sig. dose-related -l dam
al., 1995
females, 8-12
320, 475, 633, 844
corn oil
LOAEL: 475
BW gain at all dose levels
dams/group
or 1,125 mg/kg-
day
GD 6-15
mg/kg-day
on GD 6-8 and 6-20.
Delayed parturition at >475
mg/kg-day; ataxia at >633
mg/kg-day; mortality at
1,125 mg/kg-day.
Dev.
T full litter resorption and
NOAEL: 32
postnatal mortality at >425
mg/kg-day
mg/kg-day. Sig. prenatal
Dev.
loss at 1,125 mg/kg/day.
LOAEL: 101
Pup BW -l (not sig.) on
mg/kg-day
PND 1 and 6. Sig. T in
pups with eye defects at
1,125 mg/kg-day. Dose-
related (n.s.) T in pups with
eye defects at >101 mg/kg-
day
Narotsky &
Rat, Fischer 344,
0, 1,125, or
Gavage in
Mat.
Ataxia, 1 activity,
Kavlock,
females, 16-21
1,500 mg/kg-day
corn oil
LOAEL:
piloerection; dose-related 1
1995
dams/group
GD 6-19
1,125 mg/kg-
day
BW gain
Dev.
Sig. T Ml litter resorptions,
LOAEL:
1 live pups/litter; sig. 1 pup
1,125 mg/kg-
BW on PND 1; sig. t
day
incidences of
microophthalmia and
anophthalmia.
Noland-
Rat, Sprague-Dawley,
0, 312 mg/L
Drinking
Dev. LOEL:
Sig. i uptake of 3H-2-DG
Gerbec et
females, 9-11 dams/
(Avg. total intake
water
312 mg/L c
in whole brains and
al., 1986
group
of dams: 825 mg
TCE over 61
days.)b
Dams (and pups)
exposed from 14
days prior to
mating until end
of lactation.
cerebella (no effect in
hippocampus) of exposed
pups at 7, 11, and 16 days,
but returned to control
levels by 21 days.
6/22/2009
846
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
sex/number
Dose level/
Exposure
duration
Route/vehicle
NOAEL;
LOAEL a
Effects
Peden-
Adams et
al., 2006
Mouse, B6C3F1,
dams and
both sexes
offspring, 5
dams/group; 5-7
pups/group at 3
wks; 4-5
pups/sex/group at 8
weeks
0, 1,400, or
14,000 ppb
Parental mice
and/or
offspring
exposed during
mating, and
from GD 0 thru
3 or 8 wks of
age
Drinking
water
Dev.
LOAEL:
1,400 ppb
C
At 1,400 ppb: Suppressed
plaque-forming cell (PFC)
responses in males at 3 and
8 wks of age and in females
at 8 wks of age. Delayed
hypersensitivity response
increased at 8 wks of age in
females.
At 14,000 ppb: Suppressed
PFC responses in males
and females at 3 and 8 wks
of age. Splenic cell
population decreased in 3
wk old pups. Increased
thymic T-cells at 8 wks of
age. Delayed
hypersensitivity response
increased at 8 wks of age in
males and females.
Peden-
Adams et
al., 2008
Mouse, MRL +/+,
dams and both
sexes offspring,
unknown no.
litters/group, 6-10
offspring/sex/group
0, 1,400, or
14,000 ppb
(vehicle = 1%
emulphore)
Parental mice
and/or
offspring
exposed from
GD 0 to 12
months of age
Drinking
water
Dev.
LOAEL=
1,400 ppb
C
At 1,400 ppb: splenic CD4-
/CD8- cells sig.t in
females; thymic
CD4+/CD8+ cells sig. -i- in
males; 18% f in male
kidney weight
At 14,000 ppb: thymic T-
cell subpopulations (CD8+,
CD4/CD8-, CD4+) sig. ^
in males
6/22/2009
847
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Reference
Species/strain/
sex/number
Dose level/
Exposure
duration
Route/vehicle
NOAEL;
LOAEL a
Effects
Taylor et
al., 1985
Rat, Sprague-
Dawley, females,
no. dams/group not
reported
0, 312, 625,
and
1,250 mg/L
Dams (and pups)
exposed from 14
days prior to
mating until end
of lactation.
Drinking
water
Dev.
LOAEL:
312 mg/L c
Exploratory behavior sig. t
in 60- and 90-day old male
rats at all treatment levels.
Locomotor activity was
higher in rats from dams
exposed to 1,250 ppm
TCE.
a NOAEL (No Observed Adverse Affect Level), LOAEL (Lowest Observed Adverse Affect Level), and LOEL
(Lowest Observed Effect Level) are based upon reported study findings. Mat. = Maternal; Dev. = Developmental.
b Dose conversions provided by study author(s).
0 Maternal observations not reported.
4.7.3.2.1.3 Intraperitoneal exposures
The effect of TCE on pulmonary development was evaluated in a study by Das and Scott
(1994). Pregnant Swiss-Webster mice (5/group) were administered a single intraperitoneal
injection of TCE in peanut oil at doses of 0 or 3,000 mg/kg on gestation day 17 (where mating =
day 1). Lungs from GD 18 and 19 fetuses and from neonates on postnatal days (PND) 1, 5, and
10 were evaluated for phospholipid content, DNA, and microscopic pathology. Fetal and
neonatal (PND 1) mortality was significantly increased (p < 0.01) in the treated group. Pup body
weight and absolute lung weight were significantly decreased (p < 0.05) on PND 1, and mean
absolute and relative (to body weight) lung weights were significantly decreased on GD 18 and
19. Total DNA content ((J-g/mg lung) was similar between control and treated mice, but lung
phospholipid was significantly (p < 0.05) reduced on GD 19 and significantly increased
(p < 0.05) on PND 10 in the TCE-treated group. Microscopic examination revealed delays in
progressive lung morphological development in treated offspring, first observed at GD 19 and
continuing at least through PND 5.
6/22/2009
848
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.7.3.2.2 Studies in non-mammalian species
4.7.3.2.2.1 Avian
Injection of White Leghorn chick embryos with 1, 5, 10, or 25 [j,mol TCE per egg on
Days 1 and 2 of embryogenesis demonstrated mortality, growth defects, and morphological
anomalies at evaluation on Day 14 (Bross et al., 1983). These findings were consistent with a
previous study that had been conducted by Elovaara et al. (1979). Up to 67% mortality was
observed in the treated groups, and most of the surviving embryos were malformed (as compared
to a complete absence of malformed chicks in the untreated and mineral-oil-treated control
groups). Reported anomalies included subcutaneous edema, evisceration (gastroschisis), light
dermal pigmentation, beak malformations, club foot, and patchy feathering. Retarded growth
was observed as significantly (p < 0.05) reduced crown-rump, leg, wing, toe, and beak lengths as
compared to untreated controls. This study did not identify any liver damage or cardiac
anomalies.
In a study by Loeber et al. (1988), 5, 10, 15, 20, or 25 [j,mol TCE was injected into the air
space of White Longhorn eggs at embryonic stages 6, 12, 18, or 23. Embryo cardiac
development was examined in surviving chicks in a double-blinded manner at stages 29, 34, or
44. Cardiac malformations were found in 7.3% of TCE-treated hearts, compared to 2.3% of
saline controls and 1.5% of mineral oil controls. The observed defects included septal defects,
cor biloculare, conotruncal abnormalities, atrioventricular canal defects, and abnormal cardiac
muscle.
Drake et al. (2006a) injected embryonated White Leghorn chicken eggs (Babcock or
Bovan strains) with 0, 0.4, 8, or 400 ppb TCE per egg during the period of cardiac valvuloseptal
morphogenesis (i.e., 2-3.3 days incubation). The injections were administered in four aliquots at
Hamberger and Hamilton (HH) stages 13, 15, 17, and 20, which spanned the major events of
cardiac cushion formation, from induction through mesenchyme transformation and migration.
Embryos were harvested 22 hours after the last injection (i.e., HH 24 or HH 30) and evaluated
for embryonic survival, apoptosis, cellularity and proliferation, or cardiac function. Survival was
significantly reduced for embryos at 8 and 400 ppb TCE at HH 30. Cellular morphology of
cushion mesenchyme, cardiomyocytes, and endocardiocytes was not affected by TCE treatment;
however, the proliferative index was significantly increased in the atrioventricular canal (AVC)
cushions at both treatment levels and in the outflow tract (OFT) cushions at 8 ppb. This resulted
in significant cushion hypercellularity for both the OFT and AVC of TCE-treated embryos.
Similar outcomes were observed in embryos when TCA or TCOH was administered, and the
effects of TCA were more severe than for TCE. Doppler ultrasound assessment of cardiac
hemodynamics revealed no effects of TCE exposure on cardiac cycle length or heart rate;
6/22/2009
849
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
however there was a reduction in dorsal aortic blood flow, which was attributed to a 30.5%
reduction in the active component of atrioventricular blood flow. Additionally the passive-to-
active atrioventricular blood flow was significantly increased in treated embryos, and there was a
trend toward lower stroke volume. The overall conclusion was that exposure to 8 ppb TCE
during cushion morphogenesis reduced the cardiac output of the embryos in this study. The
findings of cardiac malformations and/or mortality following in ovo exposure to chick embryos
with 8 ppb TCE during the period of valvuloseptal morphogenesis has also been confirmed by
Rufer et al. (2008).
In a follow-up study, Drake et al. (2006b) injected embryonated White Leghorn chicken
eggs with TCE or TCA during the critical window of avian heart development, beginning at HH
stage 3+ when the primary heart field is specified in the primitive streak and ending
approximately 50 hours later at HH stage 17, at the onset of chambering. Total dosages of 0, 0.2,
2, 4, 20, or 200 nmol (equivalent to 0, 0.4, 4, 8, 40, or 400 ppb) were injected in four aliquots
into each egg yolk during this window (i.e., at stages 3+, 6, 13, and 17: hours 16, 24, 46, and 68).
Embryos were harvested at 72 hours, 3.5 days, 4 days or 4.25 days (HH stages 18, 21, 23, or 24,
respectively) and evaluated for embryonic survival, cardiac function, or cellular parameters.
Doppler ultrasound technology was utilized to assess cardiovascular effects at HH 18, 21, and
23. In contrast with the results of Drake et al. (2006a), all of the functional parameters assessed
(i.e., cardiac cycle length, heart rate, stroke volume, and dorsal aortic and atrioventricular blood
flow) were similar between control and TCE- or TCA-treated embryos. The authors attributed
this difference in response between studies to dependence upon developmental stage at the time
of exposure. In this case, the chick embryo is relatively resistant to TCE when exposure
occurred during early cardiogenic stages, but was extremely vulnerable when TCE exposure
occurred during valvuloseptal morphogenesis. It was opined that this could explain why some
researchers have observed no developmental cardiac effects after TCE exposure to mammalian
models, while others have reported positive associations.
4.7.3.2.2.2 Amphibian
The developmental toxicity of TCE was evaluated in the Frog Embryo Teratogenesis
Assay: Xenopus (FETAX) by Fort et al. (1991, 1993). Late Xenopus laevis blastulae were
exposed to TCE, with and without exogenous metabolic activation systems, or to TCE
metabolites (dichloroacetic acid, trichloroacetic acid, trichloroethanol, or oxalic acid), and
developmental toxicity ensued. Findings included alterations in embryo growth, and increased
types and severity of induced malformations. Findings included cardiac malformations that were
reportedly similar to those that had been observed in avian studies. It was suggested that a mixed
6/22/2009
850
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
function oxidase-mediated reactive epoxide intermediate (i.e., TCE-oxide) may play a significant
role in observed developmental toxicity in in vitro tests.
Likewise, McDaniel et al. (2004) observed dose-dependent increases in developmental
abnormalities in embryos of four North American amphibian species (wood frogs, green frogs,
American toads, and spotted salamanders) following 96-hour exposures to TCE. Median
effective concentrations (EC50) for malformations was 40 mg/L for TCE in green frogs, while
American toads were less sensitive (with no EC50 at the highest concentration tested - 85 mg/L).
Although significant mortality was not observed, the types of malformations noted would be
expected to compromise survival in an environmental context.
4.7.3.2.2.3 Invertebrate
The response of the daphnid Ceriodaphnia dubia to six industrial chemicals, including
TCE, was evaluated by Niederlehner et al. (1998). Exposures were conducted for 6-7 days,
according to standard EPA testing guidelines. Lethality, impairment of reproduction, and
behavioral changes, such as narcosis and abnormal movement, were observed with TCE
exposures. The reproductive sublethal effect concentration (IC50) value for TCE was found to be
82 |xM.
4.7.3.2.3 In vitro studies
Rat whole embryo cultures were used by Saillenfait et al. (1995) to evaluate the
embryotoxicity of TCE, tetrachloroethylene, and four metabolites (trichloroacetic acid,
dichloroacetic acid, chloral hydrate, and trichloroacetyl chloride). In this study, explanted
embryos of Sprague-Dawley rats were cultured in the presence of the test chemicals for 46 hours
and subsequently evaluated. Concentration-dependant decreases in growth and differentiation,
and increases in the incidence of morphologically abnormal embryos were observed for TCE at
>5 mM.
Whole embryo cultures were also utilized by Hunter et al. (1996) in evaluating the
embryotoxic potential of a number of disinfection by-products, including the TCE metabolites
dichloroacetic acid (DCA) and trichloroacetic acid (TCA). CD-I mouse conceptuses (GD 9; 3-6
somites) were cultured for 24-26 hours in treated medium. DCA levels assessed were 0, 734,
I,468, 4,403, 5,871, 7,339, 11,010, or 14,680 |xM; TCA levels assessed were 0, 500, 1,000,
2,000, 3,000, 4,000, 5,000 [xM. For DCA, neural tube defects were observed at levels of > 5,871
|iM, heart defects were observed at > 7,339 [xM, and eye defects were observed at levels of >
II,010 [jM. For TCA, neural tube defects were observed at levels of > 2,000 [xM, heart and eye
defects were observed at > 3,000 [xM. The heart defects for TCA were reported to include
6/22/2009
851
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
incomplete looping, a reduction in the length of the heart beyond the bulboventricular fold, and a
marked reduction in the caliber of the heart tube lumen. Overall benchmark concentrations (i.e.,
the lower limit of the 95% confidence interval required to produce a 5% increase in the number
of embryos with neural tube defects) were 2,451.9 |iM for DCA and 1,335.8 |iM for TCA
(Richard and Hunter, 1996).
Boyer et al. (2000) used an in vitro chick-atrioventricular (AV) canal culture to test the
hypothesis that TCE might cause cardiac valve and septal defects by specifically perturbing
epithelial-mesenchymal cell transformation of endothelial cells in the AV canal and outflow tract
areas of the heart. AV explants from Stage 16 White Leghorn chick embryos were placed in
hydrated collagen gels, with medium and TCE concentrations of 0, 50, 100, 150, 200, or 250
ppm. TCE was found to block the endothelial cell-cell separation process that is associated with
endothelial activation as well as to inhibit mesenchymal cell formation across all TCE
concentrations tested. TCE did not, however, have an effect on the cell migration rate of fully
formed mesenchymal cells. TCE-treatment was also found to inhibit the expression of
transformation factor Mox-1 and extracellular matrix protein fibrillin 2, two protein markers of
epithelial-mesenchyme cell transformation.
4.7.3.3 Discussion/synthesis of developmental data
In summary, an overall review of the weight of evidence in humans and experimental
animals is suggestive of the potential for developmental toxicity with TCE exposure. A number
of developmental outcomes have been observed in the animal toxicity and the epidemiological
data, as discussed below. These include adverse fetal/birth outcomes including death
(spontaneous abortion, perinatal death, pre- or post-implantation loss, resorptions), decreased
growth (low birth weight, small for gestational age, intrauterine growth restriction, decreased
postnatal growth), and congenital malformations, in particular cardiac defects. Postnatal
developmental outcomes include developmental neurotoxicity, developmental immunotoxicity,
and childhood cancer.
4.7.3.3.1 Adverse fetal and early neonatal outcomes
Studies that demonstrate adverse fetal or early neonatal outcomes are summarized in
Table 4.7-18. In human studies of prenatal TCE exposure, increased risk of spontaneous
abortion was observed in some studies (ATSDR, 2001; Taskinen et al., 1994; Windham et al.,
1991), but not in others (ATSDR, 2001, 2008; Goldberg et al., 1990; Lagakos et al., 1986;
Lindbohm et al., 1990; Taskinen et al., 1989). In addition, perinatal deaths were observed after
1970, but not before 1970 (Lagakos et al., 1986). In rodent studies that examined offspring
6/22/2009
852
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
viability and survival, there was an indication that TCE exposure may have resulted in increased
pre-and/or postimplantation loss (Kumar et al., 2000a; Healy et al., 1982; Narotsky and Kavlock,
1995), and in reductions in live pups born as well as in postnatal and postweaning survival
(George etal., 1985, 1986).
Decreased birth weight and small for gestational age was observed (ATSDR, 1998, 2006;
Rodenbeck et al., 2000; Windham et al., 1991), however no association was observed in other
studies (Bove, 1996; Bove et al., 1995; Lagakos et al., 1986). While comprising both
occupational and environmental exposures, these human studies are overall not highly
informative due to their small numbers of cases and limited exposure characterization or to the
fact that exposures to mixed solvents were involved. However, decreased fetal weight, live birth
weights and postnatal growth were also observed in rodents (George et al., 1985, 1986; Healy et
al., 1982; Narotsky and Kavlock, 1995), adding to the weight of evidence for this endpoint. It is
noted that the rat studies reporting effects on fetal or neonatal viability and growth used Fischer
344 or Wistar rats, while several other studies, which used Sprague-Dawley rats, reported no
increased risk in these developmental measures (Carney et al., 2006; Hardin et al., 1981;
Schwetz et al., 1975).
Overall, based on weakly suggestive epidemiologic data and fairly consistent laboratory
animal data, it can be concluded that TCE exposure poses a potential hazard for prenatal losses
and decreased growth or birth weight of offspring.
Table 4.7.18. Summary of adverse fetal and early neonatal outcomes associated with TCE
exposures
Positive Finding
Species
Citation
Spontaneous abortion, miscarriage,
Human
ATSDR, 2001 a
pre-and/or postimplantation loss
Taskinen et al., 1994 a
Windham et al., 1991
Rat
Kumar et al., 2000a
Healy et al., 1982
Narotsky and Kavlock, 1995
Narotsky et al., 1995
Perinatal death, reduction in live
Human
Lagakos et al., 1986 b
births
Mouse
George et al., 1985
Rat
George et al., 1986
Postnatal and postweaning survival
Mouse
George et al., 1985
Rat
George et al., 1986
6/22/2009
853
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Decreased birth weight, small for
Human
ATSDR, 1998
gestational age, postnatal growth
ATSDR, 2006
Rodenbeck et al., 2000 c
Windham et al., 1991
Mouse
George et al., 1985
Rat
George et al., 1986
Healy et al., 1982
Narotsky and Kavlock, 1995
Narotsky et al., 1995
a Not significant.
b Observed for exposures after 1970, but not before.
0 Increased risk for very low birth weight but not low birth weight or full-term low birth weight.
4.7.3.3.2 Cardiac malformations
A discrete number of epidemiological studies and studies in laboratory animal models
have identified an association between TCE exposures and cardiac defects in developing
embryos and/or fetuses. These are listed in Table 4.7-19. Additionally, a number of avian and
rodent in vivo studies and in vitro assays have examined various aspects of the induction of
cardiac malformations.
In humans, an increased risk of cardiac defects has been observed after exposure to TCE
in studies reported by ATSDR (2006, 2008) and Yauck et al. (2004), although others saw no
significant effect (Bove et al., 1995; Bove, 1996; Goldberg et al., 1990; Lagakos et al., 1986),
possibly due to a small number of cases. In addition, altered heart rate was seen in one study
(Jasinka, 1965, translation). A cohort of water contamination in Santa Clara County, California
is often cited as a study of TCE exposure and cardiac defects; however, the chemical of exposure
is in fact trichloroethane, not TCE (Deane et al., 1989; Swan et al., 1989).
In laboratory animal models, avian studies were the first to identify adverse effects of
TCE exposure on cardiac development. As described in Section 4.7.2.2.1, cardiac malformations
have been reported in chick embryos exposed to TCE (Bross et al., 1983; Loeber et al., 1988;
Boyer et al., 2000; Drake et al., 2006a, b; Mishima et al., 2006; Rufer et al., 2008). Additionally,
a number of studies were conducted in rodents in which cardiac malformations were observed in
fetuses following the oral administration of TCE to maternal animals during gestation (Dawson
et al., 1990, 1993; Johnson et al., 2003, 2005; see Section 4.7.2.2.1.2). Cardiac defects were also
observed in rats following oral gestational treatment with metabolites of TCE (Johnson et al.,
1998a, b; Smith et al., 1989, 1992; Epstein et al., 1992).
6/22/2009
854
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
However, cardiac malformations were not observed in a number of other studies in
laboratory animals in which TCE was administered during the period of cardiac organogenesis
and fetal visceral findings were assessed. These included inhalation studies in rats (Dorfmueller
et al., 1979; Schwetz et al., 1975; Hardin et al., 1981; Healy et al., 1982; Carney et al., 2006) and
rabbits (Hardin et al., 1981), and oral gavage studies in rats (Narotsky et al., 1995; Narotsky and
Kavlock, 1995; Fisher et al., 2001) and mice (Cosby and Dukelow, 1992).
It is generally recognized that response variability among developmental bioassays
conducted with the same chemical agent may be related to factors such as the study design (e.g.,
the species and strain of laboratory animal model used, the day(s) or time of day of dose
administration in relation to critical developmental windows, the route of exposure, the vehicle
used, the day of study termination), or the study methodologies (e.g., how fetuses were
processed, fixed, and examined; what standard procedures were used in the evaluation of
morphological landmarks or anomalies, and whether there was consistency in the fetal
evaluations that were conducted). In the case of studies that addressed cardiac malformations,
there is additional concern as to whether detailed visceral observations were conducted, whether
or not cardiac evaluation was conducted using standardized dissection procedures (e.g., with the
use of a dissection microscope or including confirmation by histopathological evaluation, and
whether the examinations were conducted by technicians who were trained and familiar with
fetal cardiac anatomy). Furthermore, interpretation of the findings can be influenced by the
analytical approaches applied to the data as well as by biological considerations such as the
historical incidence data for the species and strain of interest. These issues have been critically
examined in the case of the TCE developmental toxicity studies (Hardin et al., 2005; Watson et
al., 2006).
In the available animal developmental studies with TCE, differences were noted in the
procedures used to evaluate fetal cardiac morphology following TCE gestational exposures
across studies, and some of these differences may have resulted in inconsistent fetal outcomes
and/or the inability to detect cardiac malformations. Most of the studies that did not identify
cardiac anomalies used a traditional free-hand sectioning technique (as described in Wilson,
1965) on fixed fetal specimens (Dorfmueller et al., 1979; Schwetz et al., 1975; Hardin et al.,
1981; Healy et al., 1982). Detection of cardiac anomalies can be enhanced through the use of a
fresh dissection technique as described by Staples (1974) and Stuckhardt and Poppe (1984); a
significant increase in treatment-related cardiac heart defects was observed by Dawson et al.
(1990) when this technique was used. Further refinement of this fresh dissection technique was
employed by Dawson and colleagues at the University of Arizona (UA), resulting in several
additional studies that reported cardiac malformations (Dawson et al., 1993; Johnson et al., 2003,
2005). However, two studies conducted in an attempt to verify the teratogenic outcomes of the
6/22/2009
855
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
UA laboratory studies used the same or similar enhanced fresh dissection techniques and were
unable to detect cardiac anomalies (Fisher et al., 2001; Carney et al., 2001). Although the
Carney et al. study was administered via inhalation (a route which has not previously been
shown to produce positive outcomes), the Fisher et al. study was administered orally and
included collaboration between industry and UA scientists. It was suggested that the apparent
differences between the results of the Fisher et al. study and the Dawson et al. (1993) and
Johnson et al. studies may be related to factors such as differences in purity of test substances or
in the rat strains, or differences in experimental design (e.g., oral gavage versus drinking water,
exposure only during the period of organogenesis versus during the entire gestation period, or the
use of a staining procedure).
It is notable that all studies that identified cardiac anomalies following gestational
exposure to TCE or its metabolites were: 1) conducted in rats and 2) dosed by an oral route of
exposure (gavage or drinking water). Cross-species and route-specific differences in fetal
response may be due in part to toxicokinetic factors. Although a strong accumulation and
retention of TCA was found in the amniotic fluid of pregnant mice following inhalation
exposures to TCE (Ghantous et al., 1986), other toxicokinetic factors may be critical. The
consideration of toxicokinetics in determining the relevance of murine developmental data for
human risk assessment is briefly discussed by Watson et al. (2006). There are differences in the
metabolism of TCE between rodent and humans in that TCE is metabolized more efficiently in
rats and mice than humans, and a greater proportion of TCE is metabolized to DCA in rodents
versus to TCA in humans. Studies that examined the induction of cardiac malformations with
gestational exposures of rodents to various metabolites of TCE identified TCA and DCA as
putative cardiac teratogens. Johnson et al. (1998a, b) and Smith et al. (1989) reported increased
incidences of cardiac defects with gestational TCA exposures, while Smith et al. (1992) and
Epstein et al. (1992) reported increased incidences following DCA exposures.
In all studies that observed increased cardiac defects, either TCE or its metabolites were
administered during critical windows of in utero cardiac development, primarily during the entire
duration of gestation, or during the period of major organogenesis (e.g., GD 6-15 in the rat).
The study by Epstein et al. (1992) used dosing with DCA on discrete days of gestation and had
identified gestation days 9 through 12 as a particularly sensitive period for eliciting high
interventricular septal defects associated with exposures to TCE or its metabolites.
In the oral studies that identified increased incidences of cardiac malformations following
gestational exposure to TCE, there was a broad range of administered doses at which effects
were observed. In drinking water studies, Dawson et al. (1993) observed cardiac anomalies at
1.5 and 1,100 ppm (with no NOAEL) and Johnson et al. (2003, 2005) reported effects at 250 ppb
(with a NOAEL of 2.5. ppb). Once concern is the lack of a clear dose-response for the incidence
6/22/2009
856
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
of any specific cardiac anomaly or combination of anomalies was not identified, a disparity for
which no reasonable explanation for this disparity has been put forth.
The analysis of the incidence data for cardiac defects observed in the Johnson et al.
(2003, 2005) studies has been critiqued (Watson et al., 2006). Issues of concern that have been
raised include the statistical analyses of findings on a per-fetus (rather than the more appropriate
per-litter) basis (Benson, 2004), and the use of non-concurrent control data in the analysis
(Hardin et al., 2004). In response, the study author has further explained procedures used
(Johnson, 2004) and has provided individual litter incidence data to the USEPA for independent
statistical analysis (P. Johnson, personal communication, 2008) (see Section 6, dose-response).
In sum, while the studies by Dawson et al. (1993) and Johnson et al. (2003, 2005) have
significant limitations, there is insufficient reason to dismiss their findings.
Table 4.7.19. Summary of studies that identified cardiac malformations associated with
TCE exposures
Finding
Species
Citations
Cardiac defects
Human
AT SDR, 2006, 2008;
Yauck et al., 2004;
Rat
Dawson etal., 1990, 1993
Johnson et al., 2003, 2005
Johnson et al., 1998a, b a
Smith etal., 1989 a, 1992 a
Epstein et al., 1992 a
Chicken
Bross et al., 1983
Boyer et al., 2000
Loeber et al., 1988
Drake et al., 2006a, b
Mishima et al., 2006
Rufer et al., 2008
Altered heart rate
Human
Jasinka, 1965, translation
a Metabolites of TCE.
4.7.3.3.2.1 Mode of action for cardiac malformations
A number of in vitro studies have been conducted to further characterize the potential for
alterations in cardiac development that have been attributed to exposures with TCE and/or its
metabolites. It was noted that many of the cardiac defects observed in humans and laboratory
species (primarily rats and chickens) involved septal and valvular structures.
6/22/2009
857
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
During early cardiac morphogenesis, outflow tract and atrioventricular (A-V) endothelial
cells differentiate into mesenchymal cells. These mesenchymal cells have characteristics of
smooth muscle-like myofibroblasts and form endocardial cushion tissue, which is the primordia
of septa and valves in the adult heart. Events that take place in cardiac valve formation in
mammals and birds are summarized by NRC (2006) and reproduced in Table 4.7-20.
Table 4.7.20. Events in cardiac valve formation in mammals and birds a
Stage and Event
Structural Description b
Early cardiac development
The heart is a hollow, linear, tube-like structure with two cell layers.
The outer surface is a myocardial cell layer, and the inner luminal
surface is an endothial layer. Extracellular matrix is between the two
cell layers.
Epithelial-mesenchymal cell
A subpopulation of endothelial sells lining the atrioventricular canal
transformation
detaches from adjacent cells and invades the underlying extracellular
matrix.
Three events occur:
> Endothelial cell activation (avian stage 14)
> Mesenchymal cell formation (avian stage 16)
r Mesenchymal cell migration into the extracellular matrix (avian
stages 17 and 18)
Mesenchymal cell migration and
Endothelial-derived mesenchymal cells migrate toward the
proliferation
surrounding myocardium and proliferate to populate the
atrioventricular (A-V) canal extracellular matrix.
Development of septa and
Cardiac mesenchyme provides cellular constituents for:
valvular structures
> Septum intermedium
> Valvular leaflets of the mitral and tricuspid A-V valves
The septum intermedium subsequently contributes to:
> Lower portion of the interatrial septum
> Membranous portion of the interventricular septum.
a As summarized in NRC (2006)
b Markwald et al., 1984, 1996; Boyer et al., 2000
Methods have been developed to extract the chick stage 16 atrioventricular canal from
the embryo and culture it on a hydrated collagen gel for 24-48 hours, allowing evaluation of the
described stages of cardiac development and their response to chemical treatment. Factors that
have been shown to influence the induction of endocardial cushion tissue include molecular
components such as fibronectin, laminin, and galactosyltransferase (Mjaatvedt et al., 1987;
Loeber and Runyan, 1990), components of the extracellular matrix (Mjaatvedt et al., 1991), and
6/22/2009
858
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
smooth muscle a-actin and transforming growth factor (TGF) P3 (Nakajima et al., 1997;
Ramsdell andMarkwald, 1997).
Boyer et al. (2000) utilized the in vitro chick A-V canal culture system to examine the
molecular mechanism of TCE effects on cardiac morphogenesis. A-V canal explants from stage
16 chick embryos (15/treatment level) were placed onto collagen gels and treated with 0, 50,
100, 150, 200, or 250 ppm TCE and incubated for a total of 54 hours. Epithelial-mesenchymal
transformation, endothelial cell density, cell migration, and immunohistochemistry were
evaluated. TCE treatment was found to inhibit endothelial cell activation and normal
mesenchymal cell transformation, endothelial cell-cell separation, and protein marker expression
(i.e., transcription factor Mox-1 and extracellular matrix protein fibrillin 2). Mesenchymal cell
migration was not affected, nor was the expression of smooth muscle a-actin. The study authors
proposed that TCE may cause cardiac valvular and septal malformations by inhibiting
endothelial separation and early events of mesenchymal cell formation. Hoffman et al. (2004)
has proposed alternatively that TCE may be affecting the adhesive properties of the endocardial
cells. No experimental data are currently available that address the levels of TCE in cardiac
tissue in vivo, resulting in some questions (Dugard, 2000) regarding the relevance of these
mechanistic findings to human health risk assessment.
In a study by Mishima et al. (2006), White Leghorn chick whole embryo cultures (stage
13 and 14) were used to assess the susceptibility of endocardial epithelial-mesenchymal
transformation in the early chick heart to TCE at analytically determined concentrations of 0, 10,
20, 40, or 80 ppm. This methodology maintained the anatomical relationships of developing
tissues and organs, while exposing precisely staged embryos to quantifiable levels of TCE and
facilitating direct monitoring of developmental morphology. Following 24 hours of incubation
the numbers of mesenchymal cells in the inferior and superior AV cushions were counted. TCE
treatment significantly reduced the number of mesenchymal cells in both the superior and
inferior AV cushions at 80 ppm.
Ou et al. (2003) examined the possible role of endothelial nitric oxide synthase (which
generates nitric oxide that has an important role in normal endothelial cell proliferation and
hence normal blood vessel growth and development) in TCE-mediated toxicity. Cultured
proliferating bovine coronary endothelial cells were treated with TCE at 0-100 |iM and
stimulated with a calcium ionophore to determined changes in endothelial cells and the
generation of endothelial nitric oxide synthase, nitric oxide, and superoxide anion. TCE was
shown to alter heat shock protein interactions with endothelial nitric oxide synthase and induce
endothelial nitric oxide synthase to shift nitric oxide to superoxide-anion generation. These
findings provide insight into how TCE impairs endothelial proliferation.
6/22/2009
859
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Several studies have also identified a TCE-related perturbation of several proteins
involved in regulation of intracellular Ca2+. After 12 days of maternal exposure to TCE in
drinking water, Serca2a (sarcoendoplasmic reticulum Ca2+ ATPase) mRNA expression was
reduced in rat embryo cardiac tissues (Collier et al., 2003). Selmin et al. (2008) conducted a
microarray analysis of a P19 mouse stem cell line exposed to 1 ppm TCE in vitro, identifying
altered expression of Ryr (ryanodine receptor isoform 2). Caldwell et al. (2008) used real-time
PCR and digital imaging microscopy to characterize the effects of various doses of TCE on gene
expression and Ca2+ response to vasopressin in rat cardiac myocytes (H9c2) Serca2a and Ryr 2
expression were reduced at 12 and 48 hours following exposure to TCE. Additionally, Ca2+
response to vasopressin was altered following TCE treatment. Overall, these data suggest that
TCE may disrupt the ability to regulate cellular Ca2+ fluxes, leading to morphogenic
consequences in the developing heart. This remains an open area of research.
Thus, in summary, a number of studies have been conducted in an attempt to characterize
the MOA for TCE-induced cardiac defects. A major research focus has been on disruptions in
cardiac valve formation, using avian in ovo and in vitro studies. These studies demonstrated
treatment-related alterations in endothelial cushion development that could plausibly be
associated with defects involving septal and valvular morphogenesis in rodents and chickens.
However, a broad array of cardiac malformations has been observed in animal models following
TCE exposures (Dawson et al., 1993; Johnson et al., 2003, 2005), and other evidence of
molecular disruption of Ca2+ during cardiac development has been examined (Caldwell et al.,
2008; Collier et al., 2003; Selmin et al., 2008) suggesting the possible existence of multiple
MOAs.
4.7.3.3.2.2 Association of PPAR with developmental outcomes
The peroxisome proliferators activated receptors (PPARs) are ligand activated receptors
that belong to the nuclear hormone receptor family. Three isotypes have been identified
(PPARa, PPARS [also known as PPARP], and PPARy). These receptors, upon binding to an
activator, stimulate the expression of target genes implicated in important metabolic pathways.
In rodents, all three isotypes show specific time and tissue-dependent patterns of expression
during fetal development and in adult animals. In development, they have been especially
implicated in several aspects of tissue differentiation, e.g., of the adipose tissue, brain, placenta
and skin. Epidermal differentiation has been linked strongly with PPARa and PPARS (Michalik
et al., 2002). PPARa starts late in development, with increasing levels in organs such as liver,
kidney, intestine, and pancreas; it is also transiently expressed in fetal epidermis and CNS
(Braissant and Wahli, 1998) and has been linked to phthalate-induced developmental and
6/22/2009
860
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
testicular toxicity (Corton and Lapinskas, 2005). Liver, kidney, and heart are the sites of highest
PPARa expression (Toth et al., 2007). PPAR8 and PPARy have been linked to placental
development and function, with PPARy found to be crucial for vascularization of the
chorioallantoic placenta in rodents (Wendling et al., 1999), and placental anomalies mediated by
PPARy have been linked to rodent cardiac defects (Barak et al., 2008). While it might be
hypothesized that there is some correlation between PPAR signaling, fetal deaths, and/or cardiac
defects observed following TCE exposures in rodents, no definitive data have been generated
that elucidate a possible PPAR-mediated MOA for these outcomes.
4.7.3.3.2.3 Summary of the weight of evidence on cardiac malformations
The evidence for an association between TCE exposures in the human population and the
occurrence of congenital cardiac defects is not particularly strong. Many of the epidemiological
study designs were not sufficiently robust to detect exposure-related birth defects with a high
degree of confidence. However, two well-conducted studies by ATSDR (2006, 2008) clearly
demonstrated an elevation in cardiac defects. It could be surmised that the identified cardiac
defects were detected because they were severe, and that additional cases with less severe
cardiac anomalies may have gone undetected.
The animal data provide strong, but not unequivocal, evidence of the potential for TCE-
induced cardiac malformations following oral exposures during gestation. Strengths of the
evidence are the duplication of the adverse response in several studies from the same laboratory
group, detection of treatment-related cardiac defects in both mammalian and avian species (i.e.,
rat and chicken), general cross-study consistency in the positive association of increased cardiac
malformations with test species (i.e., rat), route of administration (i.e., oral), and the
methodologies used in cardiac morphological evaluation (i.e., fresh dissection of fetal hearts).
Furthermore, when differences in response are observed across studies they can generally be
attributed to obvious methodological differences, and a number of in ovo and in vitro studies
demonstrate a consistent and biologically plausible MOA for one type of malformation observed.
Weaknesses in the evidence include lack of a clear dose-related response in the incidence of
cardiac defects, and the broad variety of cardiac defects observed, such that they cannot all be
grouped easily by type or etiology.
Taken together, the epidemiological and animal study evidence raise sufficient concern
regarding the potential for developmental toxicity (increased incidence of cardiac defects) with
in utero TCE exposures.
6/22/2009
861
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.7.3.3.3 Other structural developmental outcomes
A summary of other structural developmental outcomes that have been associated with
TCE exposures is presented in Table 4.7-21.
In humans, a variety of birth defects other than cardiac have been observed. These
include total birth defects (Bove, 1996; Bove et al., 1995; ADHS, 1988; AT SDR, 2001), CNS
birth defects (ATSDR, 2001; Bove, 1996; Bove et al., 1995; Lagakos et al., 1986), eye/ear birth
anomalies (Lagakos et al., 1986); oral cleft defects (Bove, 1996; Bove et al., 1995; Lagakos et
al., 1986; Lorente et al., 2000); kidney/urinary tract disorders (Lagakos et al., 1986);
musculoskeletal birth anomalies (Lagakos et al., 1986); anemia/blood disorders (Burg and Gist,
1999); and lung/respiratory tract disorders (Lagakos et al., 1986). While some of these results
were statistically significant, they have not been reported elsewhere. Occupational cohort
studies, while not reporting positive results, are generally limited by the small number of
observed or expected cases of birth defects (Lorente et al., 2000; Tola et al., 1980; Taskinen et
al., 1989).
In experimental animals, a statistically significant increase in the incidence of fetal eye
defects, primarily micropththalmia and anopththalmia, manifested as reduced or absent eye
bulge, was observed in rats following gavage administration of 1,125 mg/kg-day TCE during the
period of organogenesis (Narotsky et al., 1995; Narotsky and Kavlock, 1995). Dose-related non-
significant increases in the incidence of Fischer 344 rat pups with eye defects were also observed
at lower dose levels (101, 320, 475, 633, and 844 mg/kg-day) in the Narotsky et al. (1995) study
(also reported in Barton and Das [1996]). However, no other developmental or reproductive
toxicity studies identified abnormalities of eye development following TCE exposures. For
example, in a study reported by Warren et al. (2006), extensive computerized morphometric
ocular evaluation was conducted in Sprague-Dawley rat fetuses that had been examined for
cardiac defects by Fisher et al. (2001); the dams had been administered TCE (500 mg/kg-day),
DC A (300 mg/kg-day), or TCA (300 mg/kg-day) during gestation days 6-15. No ocular defects
were found with TCE exposures; however, significant reductions in the lens area, globe area, and
interocular distance were observed with DCA exposures, and non-significant decreases in these
measures as well as the medial canthus distance were noted with TCA exposures.
Developmental toxicity studies conducted by Smith et al. (1989, 1992) also identified orbital
defects (combined soft tissue and skeletal abnormalities) in Long Evans rat fetuses following GD
6-15 exposures with TCA and DCA (statistically or biologically significant at >800 mg/kg-day
and >900 mg/kg-day, respectively). Overall, the study evidence indicates that TCE and its
oxidative metabolites can disrupt ocular development in rats. In addition to the evidence of
alteration to the normal development of ocular structure, these findings may also be an indicator
of disruptions to nervous system development. It has been suggested by Warren et al. (2006) and
6/22/2009
862
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Williams and DeSesso (2008) that the effects of concern (defined as statistically significant
outcomes) are observed only at high dose levels and are not relevant to risk assessment for
environmental exposures. On the other hand, Barton and Das (1996) point out that benchmark
dose modeling of the quantal eye defect incidence data provides a reasonable approach to the
development of oral toxicity values for TCE human health risk assessment. It is also noted that
concerns may exist not only for risks related to low level environmental exposures, but also for
risks resulting from acute or short-term occupational or accidental exposures, which may be
associated with much higher inadvertent doses.
It was also notable that a study using a single intraperitoneal dose of 3,000 mg/kg TCE to
mice during late gestation (GD 17) identified apparent delays in lung development and increased
neonatal mortality (Das and Scott, 1994). No further evaluation of this outcome has been
identified in the literature.
Healy et al. (1982) did not identify any treatment-related fetal malformations following
inhalation exposure of pregnant inbred Wistar rats to 0 or 100 ppm (535 mg/m3) on GD 8-21. In
this study, significant differences between control and treated litters were observed as an
increased incidence of minor ossification variations (p = 0.003) (absent or bipartite centers of
ossification).
Table 4.7.21. Summary of other structural developmental outcomes associated with
TCE exposures
Finding
Species
Citations
Eye/ear birth anomalies
Human
Lagakos et al., 1986
Rat
Narotsky, 1995
Narotsky and Kavlock, 1995
Oral cleft defects
Human
Bove, 1996
Bove et al., 1995
Lagakos et al., 1986
Lorente et al., 2000
Kidney/urinary tract
disorders
Human
Lagakos et al., 1986
Musculoskeletal birth
anomalies
Human
Lagakos et al., 1986
Anemia/blood disorders
Human
Burg and Gist, 1999
Lung/respiratory tract
disorders
Human
Lagakos et al., 1986
Mouse
Das and Scott, 1994
Skeletal
Rat
Healy et al., 1982
6/22/2009
863
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Other
Human
AT SDR, 2001
a As reported by the authors.
4.7.3.3.4 Developmental neurotoxicity
Studies that address effects of TCE on the developing nervous system are discussed in
detail in Section 4.2, addressed above in the sections on human developmental toxicity (Section
4.7.2) and on mammalian studies (Section 4.7.2.2.1) by route of exposure, and summarized in
Table 4.7-22. The available data collectively suggest that the developing brain is susceptible to
TCE exposures.
In humans, CNS birth defects were observed in a few studies (ATSDR, 2001; Bove,
1996; Bove et al., 1995; Lagakos et al., 1986). Postnatally, observed adverse effects in humans
include delayed newborn reflexes following use of TCE during childbirth (Beppu, 1968),
impaired learning or memory (Bernad et al., 1987, abstract; White et al., 1997); aggressive
behavior (Bernad et al., 1987, abstract); hearing impairment (Beppu, 1968; Burg et al., 1995;
Burg and Gist, 1999; ATSDR, 2003a); speech impairment (Berg et al., 1995; Burg and Gist,
1999; White et al., 1997); encephalopathy (White et al., 1997); impaired executive and motor
function (White et al., 1997); attention deficit (Bernad et al., 1987, abstract; White et al., 1997),
and autism spectrum disorder (Windham et al., 2006). While there are broad developmental
neurotoxic effects that have been associated with TCE exposure, there are many limitations in
the studies.
More compelling evidence for the adverse effect of TCE exposure on the developing
nervous system is found in the animal study data, although a rigorous evaluation of potential
outcomes has not been conducted. For example, there has not been an assessment of cognitive
function (i.e., learning and memory) following developmental exposures to TCE, nor have most
of the available studies characterized the pre- or postnatal exposure of the offspring to TCE or its
metabolites. Nevertheless, there is evidence of treatment-related alterations in brain
development and in behavioral parameters (e.g., spontaneous motor activity and social
behaviors) associated with exposures during neurological development. The animal study
database includes the following information: Following inhalation exposures of 150 ppm to mice
during mating and gestation, the specific gravity of offspring brains were significantly decreased
at postnatal time points through the age of weaning; however, this effect did not persist to 1
month of age (Westergren et al., 1984). In studies reported by Taylor et al. (1985), Isaacson and
Taylor (1989), and Noland-Gerbec et al. (1986), 312 mg/L exposures in drinking water that were
initiated 2 weeks prior to mating and continued to the end of lactation resulted, respectively, in:
a) significant increases in exploratory behavior at postnatal days 60 and 90, b) reductions in
myelination in the brains of offspring at weaning, and c) significantly decreased uptake of 2-
6/22/2009
864
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
deoxyglucose in the neonatal rat brain (suggesting decreased neuronal activity). Ocular
malformations in rats observed by Narotsky (1995) and Narotsky and Kavlock (1995) following
maternal gavage doses of 1,125 mg/kg-day during gestation may also be indicative of alterations
of nervous system development. Gestational exposures to mice (Fredriksson et al., 1993)
resulted in significantly decreased rearing activity on postnatal day 60, and dietary exposures
during the course of a continuous breeding study in rats (George et al., 1986) found a significant
trend toward increased time to cross the first grid in open field testing. In a study by Blossom et
al. (2008), alterations in social behaviors (deficits in nest-building quality and increased
aggression in males) were observed in pubertal-age MRL +/+ mice that had been exposed to 0.1
mg/mL TCE via drinking water during prenatal and postnatal development (until PND 42).
Dorfmueller et al. (1979) was the only study that assessed neurobehavioral endpoints following
in utero exposure (maternal inhalation exposures of 1,800 ± 200 ppm during gestation) and
found no adverse effects that could be attributed to TCE exposure. Specifically, an automated
assessment of ambulatory response in a novel environment on postnatal days 10, 20 and 100, did
not identify any effect on general motor activity of offspring.
Table 4.7.22. Summary of developmental neurotoxicity associated with TCE exposures
Positive Findings
Species
Citations
CNS defects, neural tube defects
Human
AT SDR, 2001
Bove, 1996; Bove et al., 1995
Lagakos et al., 1986
Eye defects
Rat
Narotsky, 1995;
Narotsky and Kavlock, 1995
Delayed newborn reflexes
Human
Beppu, 1968
Impaired learning or memory
Human
Bernad et al., 1987, abstract
White et al., 1997
Aggressive behavior
Human
Bernad et al., 1987, abstract
Rat
Blossom et al., 2008
Hearing impairment
Human
AT SDR, 2003 a;
Burg et al., 1995;
Burg and Gist, 1999
Beppu, 1968
Speech impairment
Human
AT SDR, 2003 a;
Burg et al., 1995;
Burg and Gist, 1999
White et al., 1997
6/22/2009
865
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Encephalopathy
Human
White et al., 1997
Impaired executive function
Human
White et al., 1997
Impaired motor function
Human
White et al., 1997
Attention deficit
Human
Bernad et al., 1987, abstract
Autism spectrum disorder (ASD)
Human
Windham et al., 2006
Delayed or altered biomarkers of
CNS development
Rat
Isaacson & Taylor, 1989
Noland-Gerbec et al., 1986
Westergren et al., 1984
Behavioral alterations
Mice
Blossom et al., 2008
Fredriksson et al., 1993
Rat
George et al., 1986
Taylor et al., 1985
4.7.3.3.5 Developmental immunotoxicity
Studies that address the developmental immunotoxic effects of TCE are discussed in
detail in Section 4.5, addressed above in the sections on human developmental toxicity (Section
4.7.2) and on mammalian studies (Section 4.7.2.2.1) by route of exposure, and summarized in
Table 4.7-23.
Two epidemiological studies that addressed potential immunological perturbations in
children that were exposed to TCE were reported by Lehmann et al. (2001, 2002). In the 2001
study, no association was observed between TCE and allergic sensitization to egg white and
milk, or to cytokine producing peripheral T cells, in premature neonates and 36-month-old
neonates that were at risk of atopy. In the 2002 study, there was a significant reduction in Thl
IL-2 producing cells. Another study observed altered immune response in family members of
those diagnosed with childhood leukemia, including 13 siblings under age 19 at the time of
exposure, but an analysis looking at only these children was not done (Byers et al., 1988).
Several studies were identified (Peden-Adams et al., 2006, 2008; Blossom and Doss,
2007; Blossom et al., 2008) which assessed the potential for developmental immunotoxicity in
mice following oral (drinking water) TCE exposures during critical pre- and postnatal stages of
immune system development. Peden-Adams et al. (2006) noted evidence of immune system
perturbation (suppression of PFC responses, increased T-cell subpopulations, decreased spleen
cellularity, and increased hypersensitivity response) in B6C3F1 offspring following in utero and
8 weeks of postnatal exposures to TCE. Evidence of autoimmune response was not observed in
the offspring of this non-autoimmune-prone strain of mice. However, in a study by Peden-
6/22/2009
866
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Adams et al. (2008) MRL +/+ mice, which are autoimmune-prone, were exposed from
conception until 12 months of age. Consistent with the Peden-Adams et al. (2006) study, no
evidence of increased autoantibody levels was observed in the offspring. In two other studies
focused on autoimmune responses following drinking water exposures of MRL +/+ mice to TCE
during in utero development and continuing until the time of sexual maturation, Blossom and
Doss (2007) and Blossom et al. (2008) reported some peripheral blood changes that were
indicative of treatment-related autoimmune responses in offspring. Positive response levels were
0.5 and 2.5 mg/mL for Blossom and Doss (2007) and 0.1 mg/mL for Blossom et al. (2008).
None of these studies were designed to extensively evaluate recovery, latent outcomes, or
differences in severity of response that might be attributed to the early life exposures.
Consistency in response in these animal studies was difficult to ascertain due to the variations in
study design (e.g., animal strain used, duration of exposure, treatment levels evaluated, timing of
assessments, and endpoints evaluated). Likewise, the endpoints assessed in the few
epidemiological studies that evaluated immunological outcomes following developmental
exposures to TCE were dissimilar from those evaluated in the animal models, and so provided no
clear cross-species correlation. The most sensitive immune system response noted in the studies
that exposed developing animals were the decreased PFC and increased hypersensitivity
observed by Peden-Adams et al. (2006); treatment-related outcomes were noted in mice exposed
in the drinking water at a concentration of 1,400 ppb. None of the other studies that treated mice
during immune system development assessed these same endpoints; therefore direct
confirmation of these findings across studies was not possible. It is noted, however, that similar
responses were not observed in studies in which adult animals were administered TCE (e.g.,
Woolhiser et al., 2006), suggesting increased susceptibility in the young. Differential lifestage-
related responses have been observed with other diverse chemicals (e.g., diethylstilbestrol;
diazepam; lead; 2,3,7,8-tetrachlorobenzo-p-dioxin; and tributyltin oxide) in which immune
system perturbations were observed at lower doses and/or with greater persistence when tested in
developing animals as compared to adults (Luebke et al., 2006). Thus, such an adverse response
with TCE exposure is considered biologically plausible and an issue of concern for human health
risk assessment.
Table 4.7.23. Summary of developmental immunotoxicity associated with TCE exposures
Finding
Species (Strain)
Citations
Significant reduction in Thl
IL-2 producing cells
Human
Lehmann et al., 2002
Altered immune response
Human
Byers et al., 1988
6/22/2009
867
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Suppression of PFC
responses, increased T-cell
subpopulations, decreased
spleen cellularity, and
increased hypersensitivity
response
Mouse (B6C3F1)
Peden-Adams et al., 2006
Altered splenic and thymic
T-cell subpopulations
Mouse (MRL +/+)
Peden-Adams et al., 2008
Altered thymic T-cell
subpopulations; transient
increased proinflammatory
cytokine production by T-
cells; increased
autoantibody levels and IgG
Mouse (MRL +/+)
Blossom and Doss, 2007
Increased proinflammatory
cytokine production by T-
cells
Mouse (MRL +/+)
Blossom et al., 2008
4.7.3.3.6 Childhood Cancers
A summary of childhood cancers that have been associated with TCE exposures
discussed above is presented in Table 4.7-24. A summary of studies that observed childhood
leukemia is also discussed in detail in Section 4.5.1.3.
A non-significant increased risk of leukemia diagnosed during childhood has been
observed in a number of studies examining TCE exposure (ADHS, 1998, 1990a, c; Cohn et al.,
1994; Costas et al., 2002; Lagakos et al., 1986; Lowengart et al., 1987; MADPH, 1997;
McKinney et al., 1991; Shu et al., 1999). However, other studies did not observed an increased
risk for childhood leukemia after TCE exposure (ADHS, 1990b, 1997; Morgan and Cassady,
2002), possibly due to the limited number of cases or the analysis based on multiple solvents.
CNS cancers during childhood have been reported on in a few studies. Neuroblastomas were not
statistically elevated in one study observing parental exposure to multiple chemicals, including
TCE (De Roos et al., 2001). Brain tumors were observed in another study, but the odds ratio
could not be determined (Peters et al., 1981, 1985). CNS cancers were not elevated in other
studies (ADHS, 1990c; Morgan and Cassady, 2002). Other studies did not see an excess risk of
total childhood cancers (ATSDR, 2006; Morgan and Cassady, 2002).
6/22/2009
868
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
A follow-up study of the Camp Lejeune cohort that will examine childhood cancers
(along with birth defects) was initiated in 1999 (ATSDR, 2003b) and expected to be completed
soon (GAO 2007a, b) may provide additional insight.
No studies of cancers in experimental animals in early lifestages have been identified.
Table 4.7.24. Summary of childhood cancers associated with TCE exposures
Finding
Species
Citations
Leukemia
Human
ADHS, 1988, 1990a
ADHS, 1990c
Cohn et al., 1994
Cutler et al., 1986; Costas et al., 2002;
Lagakos et al., 1986; MADPH, 1997
Lowengart et al., 1987
McKinney et al., 1991
Shu et al., 1999
Neuroblastoma
Human
De Roos et al., 2001
Peters et al., 1981, 1985
6/22/2009
869
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
References
ACS (American Cancer Society). (2008) What are the risk factors for cervical cancer?
http://www.cancer.org/docroot/CRI/content/CRI 2 4 2X What are the risk factors for cervi
cal cancer 8.asp. Accessed October 8, 2008.
ADHS (Arizona Department of Health Services). (1988) Report on mortality in Maricopa
County 1966-1986.
ADHS. (1990a) Incidence study of childhood cancer in Maricopa County: 1965-1986.
ADHS. (1990b) Cancer incidence and mortality in an East Phoenix area overlaying groundwater
contaminated with volatile organic compounds.
ADHS. (1990c) The incidence of childhood leukemia and testicular cancer in Pima County:
1970-1986.
ADHS. (1997) Case-referent study of childhood leukemia in Maricopa County, Arizona 1965-
1990.
Anttila A, Pukkala E, Sallmen M, Hernberg S, Hemminki K. (1995) Cancer incidence among
Finnish workers exposed to halogenated hydrocarbons. J Occup Environ Med 37:797-806.
ATSDR (Agency for Toxic Substances and Disease Registry). (1997) Toxicological profile for
trichloroethylene. U.S. Department of Health and Human Services, September 1997, Atlanta,
GA.
ATSDR. (1998) Volatile organic compounds in drinking water and adverse pregnancy outcomes.
U.S. Department of Health and Human Services, August 1998, Atlanta, GA.
ATSDR. (2001) Final Report: evaluation of priority health conditions in a community with
historical contamination by trichloroethylene. U.S. Department of Health and Human Services,
August 2001, Atlanta, GA.
ATSDR. (2003a) Impact of trichloroethylene exposure on oral motor, speech and hearing in
children. U.S. Department of Health and Human Services, July 2003, Atlanta, GA.
ATSDR. (2003b) Survey of specific childhood cancers and birth defects among children whose
mothers were pregnant while living at US Marine Corps Base Camp Lejeune, North Carolina,
6/22/2009
870
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1968-1985. U.S. Department of Health and Human Services, July 2003, Atlanta, GA.
ATSDR. (2004) Feasibility Investigation of Worker Exposure to Trichloroethylene at the View-
Master Factory in Beaverton, Oregon. Prepared by the Oregon Department of Human Services.
Report No. PB2005-102173.
ATSDR. (2006) Health Consultation: Cancer and Birth Outcome Analysis, Endicott Area, Town
of Union, Broome County, New York. U.S. Department of Health and Human Services, May
2006, Atlanta, GA.
ATSDR. (2007) Analyses of groundwater flow, contaminant fate and transport and distribution
of drinking water at Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North
Carolina: historical reconstruction and present-day conditions. U.S. Department of Health and
Human Services, Atlanta, GA.
ATSDR. (2008) Health statistics review follow-up: cancer and birth outcome analysis. Endicott
Area Investigation, Endicott Area, Town of Union, Broome County, New York. U.S.
Department of Health and Human Services, May 2008, Atlanta, GA.
Axelson O, Selden A, Andersson K, Hogstedt C. (1994). Updated and expanded 1 Swedish
cohort study on trichloroethylene and cancer risk. J Occup Environ 36:556-562.
Band PR, Le ND, Fang R, Deschamps M, Gallagher RP, Yang P. (2000) Identification of
occupational cancer risks in British Columbia. J Occup Environ Med 42:284-310.
Barak Y; Sadovsky Y; Shalom-Barak T. (2008) PPAR signaling in placental development and
function. PPAR Research Vol. 2008, article ID 142082, 11 pages. Doi:10.1155/2008/142082.
Bardodej Z, Vyskocil J. (1956) The problem of trichloroethylene in occupational medicine;
trichloroethylene metabolism and its effect on the nervous system evaluated as a means of
hygienic control. AMA Arch Ind Health 13(6):581—592.
Barton, H; Das, S. (1996) Alternatives for a risk assessment on chronic noncancer effects from
oral exposure to trichloroethylene. Regulatory Toxicology and Pharmacology 24:269-285.
Benson, R. (2004) Correspondence from Robert Benson (U.S. Environmental Protection
Agency, Region 8, Denver, CO 80202) to Dr. P.D. Johnson (University of Arizona, UAC,
Tucson, AZ 85721), December 27, 2004.
6/22/2009
871
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Beppu, K. (1968) Transmission of the anesthetic agents through the placenta in painless delivery
and their effects on newborn infants. Keio J Med 17(2):81—107.
Berger, T and Horner, CM. (2003) In vivo exposure of female rats to toxicants may affect oocyte
quality. Reproductive Toxicology 17:273-281.
Bernad PG, Newell S, Spyker DA. (1987) Neurotoxicity and behavior abnormalities in a cohort
chronically exposed to trichloroethylene. Abstract. Vet Hum Toxicol 29(6):475.
Blackshaw, AW. (1970) Histochemical localization of testicular enzyme. In Johnson AD;
Gomes, WR; and Vandermark, NL (eds.) The Testes. New York: Academic Press, pp 73-173.
Blair A, Hartge P, Stewart PA, McAdams M, Lubin J. (1998) Mortality and cancer incidence of
aircraft maintenance workers exposed to trichloroethylene and other organic solvents and
chemicals: extended follow up. Occup Environ Med 55:161-171.
Blossom, SJ; Doss, JC. (2007) Trichloroethylene alters central and peripheral immune function
in autoimmune-prone MRL+/+ mice following continuous developmental and early life
exposure. Journal of Immunotoxicology 4:129-141.
Blossom, SJ; Doss, JC; Hennings LJ; Jernigan S; Melnyk S; James, SJ. (2008) Developmental
exposure to trichloroethylene promotes CD4+ T cell differentiation and hyperactivity in
association with oxidative stress and neurobehavioral deficits in MRL +/+ mice. Toxicol Appl
Pharmacol 231(3):344-53.
Boice JD, Marano DE, Fryzek JP, Sadler CJ, McLaughlin JK. (1999) Mortality among aircraft
manufacturing workers. Occup Environ Med 56:581-597.
Boice JD, Marano DE, Cohen SS, Mumma MR, Blott WJ, Brill AB, Fryzek JP, Henderson BE,
McLaughlin JK. (2006) Mortality among Rocketdyne worker who tested rocket engines, 1948-
1999. J Occup Environ Med 48:1070-1092.
Bove FJ, Fulcomer MC, Klotz JB, Esmart J, Dufficy EM, Savrin JE. (1995) Public drinking
water contamination and birth outcomes. Am J Epidemiol 141(9):850-862.
Bove FJ. (1996) Public drinking water contamination and birthweight, prematurity, fetal deaths,
and birth defects. Toxicol Ind Health 12(2):255-266.
6/22/2009
872
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Boyer, AS; Finch, WT; Runyan, RB. (2000) Trichloroethylene inhibits development of
embryonic heart valve precursors in vitro. Toxicol Sci 53:109-117.
Braissant O; Wahli W. (1998) Differential expression of peroxisome proliferator-activated
receptor-a, -P, and -y during rat embryonic development. Endocrinology 139(6):2748-2754.
Bross, G; DiFranceisco, D; Desmond, ME. (1983) The effects of low dosages of
trichloroethylene on chick development. Toxicology 28:283-294.
Burg JA, Gist GL. (1997) The potential impact on women from environmental exposures.
Journal of women's health / the official publication of the Society for the Advancement of
Women's Health Research 6(2): 159-161.
Burg JR, Gist GL. (1999) Health effects of environmental contaminant exposure: an intrafile
comparison of the Trichloroethylene Subregistry. Archives of environmental health
54(4):231-241.
Burg JR, Gist GL, Allred SL, Radtke TM, Pallos LL, Cusack CD. (1995) The National Exposure
Registry - Morbidity analyses of noncancer outcomes from the trichloroethylene subregistry
baseline data. Int J Occup Med and Toxicol. 4(2):237-257.
Byers VS, Levin AS, Ozonoff DM, Baldwin RW. (1988) Association between clinical symptoms
and lymphocyte abnormalities in a population with chronic domestic exposure to industrial
solvent-contaminated domestic water supply and a high incidence of leukaemia. Cancer
Immunol Immunother 27(1):77-81.
Caldwell, PT; Thorne, PA; Johnson, PD; Boitano, S; Runyan, RB; Selmin, O. (2008)
Trichloroethylene disrupts cardiac gene expression and homeostasis in rat myocytes. Toxicol Sci
104(1): 135-143.
Carney, EW; Thorsrud, BA; Dugard, PH; Zablotny, CL. (2006) Developmental toxicity studies
in Crl:Cd (SD) rats following inhalation exposure to trichloroethylene and perchloroethylene.
Birth Defects Research (Part B) 77:405-412.
Chang YM, Tai CF, Yang SC, Chen CJ, Shih TS, Lin RS, Liou SH. (2003) A cohort mortality
study of workers exposed to chlorinated organic solvents in Taiwan. Ann Epidemiol
13:652-660.
6/22/2009
873
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Chang YM, Tai CF, Yang SC, Lin R, Sung FC, Shin TS, Liou SH. (2005) Cancer incidence
among workers potentially exposed to chlorinated solvents in an electronics factory. J Occup
Health 47:171-180.
Chapin, RE; Sloane, RA. (1996) Reproductive assessment by continuous breeding: evolving
study design and summaries of eighty-eight studies. Environ Health Perspect 104(suppl. 2): 1-7.
Chapin, RE; Norton, RM; Popp, JA; Bus, JS. (1982) The effect of 2,5-hexanedione on
reproductive hormones and testicular enzyme activities in the F344 rat. Toxicol Appl Pharmacol
62:262-272.
Chia SE, Ong CN, Tsakok MF, Ho A. (1996) Semen parameters in workers exposed to
trichloroethylene. Reprod Toxicol 10(4):295-299.
Chia SE, Goh VH, Ong CN. (1997) Endocrine profiles of male workers with exposure to
trichloroethylene. Am J Ind Med 32(3):217-222.
Clapp RW, Hoffman K. 2008. Cancer mortality in IBM Endicott plant workers, 1969-2001: an
update on a NY production plant. Environ Health 7:13.
Clegg, ED; Cook, JC, Chapin, RE; Foster, PMD; Daston, GP. (1997) Leydig cell hyperplasia and
adenoma formation: mechanisms and relevance to humans. Reproductive Toxicology
11(1): 107—121.
Coberly, S; Oudiz, DJ; Overstreet, JW; Wiley, LM. (1992) Effects of maternal exposure to
trichloroethylene (TCE) on cell proliferation in the mouse preimplantation embryo. Reprod
Toxicol 6:241-245.
Cohn P, Klotz J, Bove F, Berkowitz M, Fagliano J. (1994) Drinking water contamination and the
incidence of leukemia and non-Hodgkin's lymphoma. Environ Health Perspect
102(6-7):556-61.
Collier, JM; Selmin, O; Johnson, PD; Runyan, RB. (2003) Trichloroethylene effects on gene
expression during cardiac development. Birth Defects Res A Clin Mol Teratol 67:488-495.
Cook, JC; Klinefelter, GR; Hardisty, JF; Sharpe, RM; Foster, PM. (1999) Rodent Leydig cell
tumorigenesis: a review of the physiology, pathology, mechanisms, and relevance to humans.
Crit Rev Toxicol 29:169-261.
6/22/2009
874
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Corton JC; Lapinskas PJ. (2005) Peroxisome proliferator receptors: mediators of phthalate ester-
induced effects in the male reproductive tract? Toxicol Sci 83:4-17.
Cosby, NC and Dukelow, WR. (1992) Toxicology of maternally ingested trichloroethylene
(TCE) on embryonal and fetal development in mice and of TCE metabolites on in vitro
fertilization. Fundam Appl Toxicol 19:268-274.
Costa G, Merletti F, Segnan N. (1989) A mortality cohort study in a north Italian aircraft factory.
Br JInd Med 46:738-743.
Costas K, Knorr RS, Condon SK. (2002) A case-control study of childhood leukemia in Woburn,
Massachusetts: the relationship between leukemia incidence and exposure to public drinking
water. Sci Total Environ 300(l-3):23-35.
Cutler JJ, Parker GS, Rosen S, Prenney B, Healey R, Caldwell GG. (1986) Childhood leukemia
in Woburn, Massachusetts. Public Health Rep 101(2):201-205.
Das, RM; Scott, JE. (1994) Trichloroethylene-induced pneumotoxicity in fetal and neonatal
mice. Toxicology Letters 73:227-239.
Dawson, BV; Johnson, PD; Goldberg, SJ; Ulreich, JB. (1990) Cardiac teratogenesis of
trichloroethylene and dichloroethylene in a mammalian model. J Am Coll Cardiol
16(5): 1304—1309.
Dawson, BV; Johnson, PD; Goldberg, SJ; Ulreich, JB. (1993) Cardiac teratogenesis of
halogenated hydrocarbon-contaminated drinking water. J Am Coll Cardiol 21(6): 1466-1472.
De Roos AJ, Olshan AF, Teschke K, Poole C, Savitz DA, Blatt J, Bondy ML, Pollock BH.
(2001) Parental occupational exposures to chemicals and incidence of neuroblastoma in
offspring. Am J Epidemiol 154(2): 106-14.
Deane M, Swan SH, Harris JA, Epstein DM, Neutra RR. (1989) Adverse pregnancy outcomes in
relation to water contamination, Santa Clara County, California, 1980-1981. Am J Epidemiol
129(5):894-904.
Dorfmueller, MA; Henne, SP; York,G; Bornschein, RL; Manson, JM. (1979) Evaluation of
teratogenicity and behavioral toxicity with inhalation exposure of maternal rats to
trichloroethylene. Toxicology 14:153-166.
6/22/2009
875
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Drake, VJ; Koprowski, SL; Lough, J; Hu, N; Smith, SM. (2006a) Trichloroethylene exposure
during cardiac valvuloseptal morphogenesis alters cushion formation and cardiac hemodynamics
in the avian embryo. Environ Health Perspect 114(6):842-847.
Drake, VJ; Koprowski, SL; Hu, N; Smith, SM; Lough, J. (2006b) Cardiogenic effects of
trichloroethylene and trichloroacetic acid following exposure during heart specification of avian
development. Toxicol Sci 94(1): 153-162.
Dugard, PH. (2000) Letters to the Editor, Toxicol Sci 56:437-438.
DuTeaux, SB; Hengel, MJ; DeGroot, DE; Jelks, KA; Miller, MG. (2003) Evidence for
trichloroethylene bioactivation and adduct formation in the rat epididymis and efferent ducts.
Biol Reprod 69:771-779.
DuTeaux, SB; Newman, JW; Morisseau, C; Fairbairn, EA; Jelks, K; Hammock, BD; Miller,
MG. (2004a) Epoxide hydrolases in the rat epididymis: possible roles in xenobiotic and
endogenous fatty acid metabolism. Toxicol Sci 78:187-195.
DuTeaux, SB; Berger, T; Hess, RA; Sartin, BL; Miller, MG. (2004b) Male reproductive toxicity
of trichloroethylene: sperm protein oxidation and decreased fertilizing ability. Biol Reprod
70:1518-1526.
DuTeaux, SB; Berger, T; Hess, RA; Sartin, BL; Miller, MG. (2008) CORRECTION: Male
reproductive toxicity of trichloroethylene: sperm protein oxidation and decreased fertilizing
ability. Biol Reprod 79:787. dosage error in text.
El Ghawabi, SM; Mansoor, MB; El Gamel, MS; El Saharti, AA; El Enany, FF. (1973) Chronic
trichloroethylene exposure. J Egypt Med Assoc 56(11-12):715-724.
Elovaara G; Hemminki, K; Vainio, H. (1979) Effects of methylene chloride, trichl or or ethane,
trichloroethylene, tetrachloroethylene and toluene on the development of chick embryos.
Toxicology 12:111-119.
Environmental Health Strategies. (1997) Final Report. Correspondence from Paul A. Cammer,
Ph.D., Trichloroethylene Issues Group, to Cheryl Siegel Scott, US Environmental Protection
Agency. December 22, 1997.
Epstein, DL; Nolen, GA; Randall, JL; Christ, SA; Read, EJ; Stober, JA; Smith, MK. (1992)
Cardiopathic effects of dichloroacetate in the fetal Long-Evans rat. Teratology 46:225-235.
6/22/2009
876
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Fisher, JW; Channel, SR; Eggers, JS; Johnson, PD; MacMahon, KL; Goodyear, CD; Sudberry,
GL; Warren, DA; Latendesse, JR; Graeter, LJ. (2001) Trichloroethylene, trichloroacetic acid,
and dichloroacetic acid: do they affect fetal rat heart development? Int J Toxicol 20:257-267.
Forkert, P-G; Lash, LH; Nadeau V; Tardif, R; Simmonds, A. (2002) Metabolism and toxicity of
trichloroethylene in epididymis and testis. Toxicol Appl Pharmacol 182:244-254.
Forkert PG, Lash L, Tardif R, Tanphaichitr N, Vandevoort C, Moussa M. (2003) Identification
of trichloroethylene and its metabolites in human seminal fluid of workers exposed to
trichloroethylene. Drug Metab Dispos 31(3):306—311.
Fort, DJ; Rayburn, JR; DeYoung, DJ; Bantle, JA. (1991) Assessing the efficacy of an aroclor
1254-induced exogenous metabolic activation system for FETAX. Drug Chem Toxicol
14(1&2): 143—160.
Fort, DJ; Stover, EL; Rayburn, JR; Hull, M; Bantle, JA. (1993) Evaluation of the developmental
toxicity of trichloroethylene and detoxification metabolites using Xenopus. Teratog Carcinog
Mutagen 13:35-45.
Franco, JL; Teixeira, A; Meotti, FC; Ribas, CM; Stringari, J; Pomblum, SCG; Moro, AM;
BohrerD; Bairros, AV; Dafre, AL; Santos, ARS; Farina, M. (2006) Cerebellar thiol status and
motor deficit after lactational exposure to methylmercury. Environmental Research 102:22-28.
Fredriksson, A; Danielsson, BRG; Eriksson, P. (1993) Altered behavior in adult mice orally
exposed to tri- and tetrachloroethylene as neonates. Toxicology Letters 66:13-19.
GAO (U.S. Government Accountability Office). (2007a) Defense Health Care: Activities related
to past drinking water contamination at Marine Corps Base Camp Lejeune. Report to
Congressional Committees. GAO-07-276.
GAO (U.S. Government Accountability Office). (2007b) Defense Health Care: Issues related to
past drinking water contamination at Marine Corps Base Camp Lejeune. Statement of Marcia
Crosse, Director, Health Care. Testimony before the Subcommittee on Oversight and
Investigations, Committee on Energy and Commerce, House of Representatives. GAO-07-933T.
Garabrant DH, Held J, Langholz B, Bernstein L. (1988) Mortality of aircraft manufacturing
workers in southern California. Am J Ind Med 13:683-693.
6/22/2009
877
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
George, JD; Reel, JR; Myers, CB; Lawton, AD; Lamb, JC. (1985) Trichloroethylene:
reproduction and fertility assessment in CD-I mice when administered in the feed. RTI Project
No. 310-2344, NTP-86-068. National Institute of Environmental Health Sciences, National
Toxicology Program, RTP, NC.
George, JD; Reel, JR; Myers, CB; Lawton, AD; Lamb, JC. (1986) Trichloroethylene:
reproduction and fertility assessment in F344 rats when administered in the feed. RTI Project No.
310-2344, NTP-86-085. National Institute of Environmental Health Sciences, National
Toxicology Program, RTP, NC.
Ghantous, H; Danielsson, BRG; Dencker, L; Gorczak J; Vesterberg, O. (1985) Trichloroacetic
acid accumulates in murine amniotic fluid after tri- and tetrachloroethylene inhalation. Acta
Pharmacol Toxicol (Copenh) 58:105-114.
Goh VH, Chia SE, Ong CN. (1998) Effects of chronic exposure to low doses of trichloroethylene
on steroid hormone and insulin levels in normal men. Environ Health Perspect 106(1): 41-44.
Goldberg SJ, Lebowitz MD, Graver EJ, Hicks S. (1990) An association of human congenital
cardiac malformations and drinking water contaminants. J Am Coll Cardiol 16(1): 155-164.
Greenland S, Salvan A, Wegman DH, Hallock MF, Smith TH. (1994) A case-control study of
cancer mortality at the transformer-assembly facility. Int Arch Occup Environ Health 66:49-54.
Hansen J, Raaschou-Nielsen O, Christensen JM, Johansen I, McLaughlin JK, Lipworth L, Blot
WJ, Olsen JH. (2001) Cancer incidence among Danish workers exposed to trichloroethylene. J
Occup Environ Med 43:133-139.
Hardin, BD; Bond, GP; Sikov, MR; Andrew, FD; Beliles, RP; and Niemeier, RW. (1981)
Testing of selected work place chemicals for teratogenic potential. Scand J Work Environ Hlth
7(Suppl 45):66-75.
Hardin, BD; Kelman BJ; Brent RL. (2004) Trichloroethylene and cardiac malformations.
Environ Health Perspect 112(11): A607-A608.
Hardin, BD; Kelman, BJ; Brent, RL. (2005) Trichloroethylene and dichloroethylene: a critical
review of teratogenicity. Birth Defects Res A Clin Mol Teratol 73:931-955.
Healy, TEJ; Poole, TR; Hopper, A. (1982) Rat fetal development and maternal exposure to
trichloroethylene 100 p.p.m. British Journal of Anaesthesiology 54:337-340.
6/22/2009
878
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Henschler, D; Romen, W; Elsasser HM; Reichert D; Eder E; Radwan Z. (1980) Carcinogenicity
study of trichloroethylene by longterm inhalation in three animal species. Arch Toxicol
43:237-248.
Henschler, D; Elsasser H; Romen, W; Eder, E. (1984) Carcinogenicity study of
trichloroethylene, with and without epoxide stabilizers, in mice. J Cancer Res Clin Oncol
107:149-156.
Henschler D, Vamvakas S, Lammert M, Dekant W, Kraus B, Thomas B, Ulm K. (1995)
Increased incidence of renal cell tumors in a cohort of cardboard workers exposed to
trichloroethene. Arch Toxicol 69:291-299.
Hodgen, GD; Sherins, RJ. (1973) Enzymes as markers of testicular growth and development in
the rat. Endocrinology 93(4):985-989.
Hoffman, S; Mishima, N; Krug, EL. (2004) An improved model for evaluating trichloroethylene
and its metabolites as cardiac-specific teratogens. In: Mohr, LC; Hoel, DG; Jollow, D (ed.)
Trichloroethylene - The Scientific Basis of Risk Assessment. The Medical University of South
Carolina Press, Charleston, SC, pp. 69-79.
Hunter, ES; Rogers, EH; Schmid, JE; Richard, A. (1996) Comparative effects of haloacetic acids
in whole embryo culture. Teratology 54:57-64.
Isaacson, LG; Taylor, DH. (1989) Maternal exposure to 1,1,2-trichloroethylene affects myelin in
the hippocampal formation of the developing rat. Brain Research 488:403-407.
Jasinska M. (1965) [General anesthesia with trichloroethylene in children with congenital
developmental defects of the jaw and face], Czas Stomatol 18(8):935-40. Translated.
Ji B-T, Blair A, Shu X-O, Chow W-H, Hauptmann M, Dosemeci M, Yang G, Lubin J, Gao Y-
T,m Rothman N, Zheng W. (2008) Occupation and breast cancer risk among Shanghai women in
a population-based cohort study. Am J Ind Med 51:100-110.
Johnson, PD. (2004) Trichloroethylene: Johnson et al.'s response. Environ Health Perspect
112(11): A608-A609.
Johnson, AD; VanDemark WL; Gomes, WR; Butler, WR; Hodgen, GD. (1967) The effect of
antispermatogenic and hormone treatments on testicular cholesterol. Fed Proc 26:645.
6/22/2009
879
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Johnson, PD; Dawson, BV; Goldberg, SJ. (1998a) Cardiac teratogenicity of trichloroethylene
metabolites. JACC 72(2):540-545.
Johnson, PD; Dawson, BV; Goldberg, SJ. (1998b) A review: trichloroethylene metabolites:
potential cardiac teratogens. Environ Health Perspect 106(suppl. 4):995-999.
Johnson, PD; Goldberg, SJ; Mays, MZ; Dawson, BV. (2003) Threshold of trichloroethylene
contamination in maternal drinking waters affecting fetal heart development in the rat. Environ
Health Perspect 111(3):289-292.
Johnson, PD; Goldberg, SJ; Mays, MZ; Dawson, BV. (2005) Errata. Threshold of
trichloroethylene contamination in maternal drinking waters affecting fetal heart development in
the rat. Environ Health Perspect 113(1): A18.
Kan, FW; Forkert, PG; Wade, MG. (2007) Trichloroethylene exposure elicits damage in
epididymal epithelium and spermatozoa in mice. Histol Histopathol 22(9):977-988.
Krishnadasan A, Kennedy N, Zhao Y, Morgenstern H, Ritz B. (2007) Nested case-control study
of occupational chemical exposures and prostate cancer in aerospace and radiation workers. Am
JInd Med 50:383-390.
Krishnadasan A, Kennedy N, Zhao Y, Morgenstern H, Ritz B. (2008) Nested case-control study
of occupational physical activity and prostate cancer among workers using a job exposure
matrix. Cancer Causes Control 19:107-114.
Kumar, P; Prasad, AK; Saxena, DK; Mani, U; Maji, BK; Dutta, KK. (2000a) Fertility and
general reproduction studies in trichloroethylene exposed rats. Indian J Occup Health
43(3): 117—125.
Kumar, P; Prasad, AK; Dutta, KK. (2000b) Steroidogenic alterations in testes and sera of rats
exposed to trichloroethylene (TCE) by inhalation. Hum Exp Toxicol 19:117-121.
Kumar, P; Prasad, AK; Mani, U; Maji, BK; Dutta, KK. (2001) Trichloroethylene induced
testicular toxicity in rats exposed by inhalation. Hum Exp Toxicol 20:585-589.
Lagakos SW, Wessen B, Zelen M. (1986) Analysis of contaminated well water and health effect
in Woburn, Massachusetts. J Am Statist Assoc 81:583-596.
6/22/2009
880
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Lamb, JC; Hentz, KL. (2006) Toxicological review of male reproductive effects and
trichloroethylene exposure: Assessing the relevance to human male reproductive health. Reprod
Toxicol 22:557-563.
Land, PC; Owen, EL; Linde, HW. (1981) Morphologic changes in mouse spermatozoa after
exposure to inhalational anesthetics during early spermatogenesis. Anesthesiology 54:53-56.
Laslo-Baker D, Barrera M, Knittel-Keren D, Kozer E, Wolpin J, Khattak S, Hackman R, Rovet
J, Koren G. (2004) Child neurodevelopmental outcome and maternal occupational exposure to
solvents. Arch Pediatr Adolesc Med 158(10):956—61.
Lehmann I, Rehwagen M, Diez U, Seiffart A, Rolle-Kampczyk U, Richter M, et al. (2001)
Enhanced in vivo IgE production and T cell polarization toward the type 2 phenotype in
association with indoor exposure to VOC: results of the LARS study. Int J Hyg Environ Health
204(4):211-221.
Lehmann I, Thoelke A, Rehwagen M, Rolle-Kampczyk U, Schlink U, Schulz R, et al. (2002)
The influence of maternal exposure to volatile organic compounds on the cytokine secretion
profile of neonatal T cells. Environ Toxicol 17(3):203-210.
Lindbohm ML, Taskinen H, Sallmen M, Hemminki K. (1990) Spontaneous abortions among
women exposed to organic solvents. Am J Ind Med 17(4):449-463.
Lipscomb, JC; Fisher, JC; Confer, PD; Byczkowski, JZ. (1998) In vitro to in vivo extrapolation
for trichloroethylene metabolism. Toxicol Appl Pharmacol 152:376-387.
Loeber, CP; Hendrix, MJC; dePinos, SD; Goldberg, SJ. (1988) Trichloroethylene: a cardiac
teratogen in developing chick embryos. Pediatric Research 24(6):740-744.
Loeber, CP; Runyan, RB. (1990) A comparison of fibronectin, laminin, and galactosyltransferase
adhesion mechanisms during embryonic cardiac mesenchymal cell migration in vitro.
Developmental Biology 140:401-412.
Lorente C, Cordier S, Bergeret A, De Walle HE, Goujard J, Ayme S, et al. (2000) Maternal
occupational risk factors for oral clefts. Occupational Exposure and Congenital Malformation
Working Group. Scand J Work Environ Health 26(2): 137-145.
Lowengart RA, Peters JM, Cicioni C, Buckley J, Bernstein L, Preston-Martin S, Rappaport E.
(1987) Childhood leukemia and parents' occupational and home exposures. J Natl Cancer Inst
6/22/2009
881
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
79(l):39-46.
Luebke, RW; Chen, DH; Dietert R; Yang, Y; King, M. (2006) The comparative immunotoxicity
of five selected compounds following developmental or adult exposure. J Toxicol Environ
Health B 9:1-26.
MADPH (Massachusetts Department of Public Health). (1997) Woburn childhood leukemia
follow-up study. Volume 1: Analyses.
Maltoni C; Lefemine G; Cotti, G (1986) Experimental research on trichloroethylene
carcinogenesis. In: Archives of Research on Industrial Carcinogenesis. Maltoni, C; Mehlman,
MA eds. Princeton, NJ: Princeton Scientific Publishing Co., pp. 1-393.
Manson, JM; Murphy, M; Richdale, N; Smith, MK. (1984) Effects of oral exposure to
trichloroethylene on female reproductive function. Toxicology 32:229-242.
Markwald, RR; Runyan, RB; Kitten, GT; Funderburg, FM; Bernanke, DH; Brauer, PR. (1984)
Use of collagen gel cultures to study heart development: proteoglycan and glycoprotein
interactions during the formation of endocardial cushion tissue. The Role of Extracellular Matrix
in Development, Alan R. Liss, Inc., New York, pp 323-350.
Markwald, R; Eisenberg, C; Eisenberg, L; Trusk, T; Sugi, Y. (1996) Epithelial-mesenchymal
transformations in early avian heart development. Acta Anatomica 156:173-186.
McDaniel, TV; Martin, PA; Ross, N; Brown, S; Lesage, S; Pauli, BD. (2004) Effects of
chlorinated solvents on four species of North American amphibians. Archives of Environmental
Contamination Toxicology 47:101-109.
McKinney PA, Alexander FE, Cartwright RA, Parker L. (1991) Parental occupations of children
with leukaemia in west Cumbria, north Humberside, and Gateshead. BMJ 302(6778):681-7.
Michalik L; Desvergne B; Dreyer C; Gavillet M; Laurini RN; Wahli W. (2002) PPAR
expression and function during vertebrate development. Int J Develop Biol 46:105-114.
Mills, NC; Means, AR. (1970) Sorbitol dehydrogenase of rat testis changes activity during
development after hypophysectomy and following gonadotropic hormone administration.
Endocrinology 91(1): 147-156.
6/22/2009
882
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Mishima, N; Hoffman, S; Hill, EG; Krug, EL. (2006) Chick embryos exposed to
trichloroethylene in an ex ovo culture model show selective defects in early endocardial cushion
tissue formation. Birth Defects Res A Clin Mol Teratol 76:517-527.
Mjaatvedt, CH; Lepera, RC; Markwald, RR. (1987) Myocardial specificity for initiating
endothelial-mesenchymal cell transition in embryonic chick heart correlates with a particulate
distribution of fibronectin. Develop Biol 119:59-67.
Mjaatvedt, CH; Krug, EL; Markwald, RR. (1991) An antiserum (ESI) against a particulate form
of extracellular matrix blocks the transition of cardiac endothelium into mesenchyme in culture.
Developmental Biology 145:219-230.
Morgan JW, Cassady RE. (2002) Community cancer assessment in response to long-time
exposure to perchlorate and trichloroethylene in drinking water. J Occup Environ Med.
44(7):616-621.
Morgan RW, Kelsh MA, Zhao K, Heringer S. (1998) Mortality of aerospace workers exposed to
trichloroethylene. Epidemiology 9:424-431.
Morgan RW, Kelsh MA, Zhao K, Heringer S. (2000). Mortality of aerospace workers exposed to
trichloroethylene. Erratum. Epidemiology 9:424-431.
Nakajima, Y; Mironov, V; Yamagishi, T; Nakamura, H. (1997) Expression of smooth muscle
alpha-actin in mesenchymal cells during formation of avian endocardial cushion tissue: a role for
transforming growth factor P3. Developmental Dynamics 209:296-309.
Narotsky, MG; Kavlock, RJ. (1995) A multidisciplinary approach to toxicological screening: II.
Developmental toxicity. J Toxicol Environ Health 45:145-171.
Narotsky, MG; Weller, EA; Chinchilli, VM; Kavlock, RJ. (1995) Nonadditive developmental
toxicity in mixtures of trichloroethylene, di(2-ethylhexyl) phthalate, and heptachlor in a 5 x 5 x 5
design. Fundam Appl Toxicol 27:203-216.
Nelson, JL; Zenick, H. (1986) The effect of trichloroethylene on male sexual behavior: possible
opioid role. Neurotoxicol Teratol 8(5):441-445.
Niederlehner, BR; Cairns, J, Jr.; Smith, EP. (1998) Modeling acute and chronic toxicity of
nonpolar narcotic chemicals and mixtures to Ceriodaphnia dubia. Ecotoxicol Environ Saf
39:136-146.
6/22/2009
883
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Noland-Gerbec, EA; Pfohl, RJ; Taylor, DH; Bull RJ. (1986) 2-Deoxyglucose uptake in the
developing rat brain upon pre- and postnatal exposure to trichloroethylene. Neurotoxicology
7(3):157-164.
NRC (National Research Council). (2006) Assessing the Human Health Risks of
Trichloroethylene - Key Scientific Issues. The National Academies Press, Washington, DC.
NTP (National Toxicology Program). (1988) Toxicology and Carcinogenesis Studies of
Trichloroethylene (CAS No. 79-01-6) in Four Strains of Rat (ACI, August, Marshall, and
Osborne-Mendel) (Gavage Studies). NTP-TR-273.
NTP (National Toxicology Program). (1990) Carcinogenesis Studies of Trichloroethylene
(Without Epichlorohydrin) (CAS No. 79-01-6) in F344/N Rats and B6C3F1 Mice (Gavage
Study). NTP TR 243. Research Triangle Park, NC: U.S Department of Health and Human
Services.
O'Donnell L, McLachlan RI, Wreford NG, Robertson DM. (1994) Testosterone promotes the
conversion of round spermatids between stages VII and VIII of the rat spermatogenic cycle.
Endocrinology 135:2608-2614.
Ou, J; Ou, Z; McCarver, G; Hines, RN; Oldham, KT; Ackerman, AW; Pritchard, KA. (2003)
Trichloroethylene decreases heat shock protein 90 interactions with endothelial nitric oxide
synthase: implications for endothelial cell proliferation. Toxicol Sci 73:90-97.
Peden-Adams, MM; EuDaly, JG; Heesemann, LM; Smythe, J; Miller, J; Gilkeson, GS; Keil, DE.
(2006) Developmental immunotoxicity of trichloroethylene (TCE): Studies in B6C3F1 mice. J
Environ Sci Health A 41:249-271.
Peden-Adams MM, EuDaly JG, Lee AM, Miller DE, Keil DE, Gilkeson GS. (2008)
Developmental effects of trichloroethylene (TCE) on immunity: studies in MRL (+/+) mice. J
Environ Sci Health A Tox Hazard Subst Environ Eng A43(12): 1402-9.
Peters FM, Preston-Martin S, Yu MC. (1981) Brain tumors in children and occupational
exposure of parents. Science 213(4504):235-7.
Peters JM, Garabrant DH, Wright WE, Bernstein L, Mack TM. (1985) Uses of a cancer registry
in the assessment of occupational cancer risks. Natl Cancer Inst Monogr 69:157-61.
Phillips TJ, Macdonald RR. (1971) Comparative effect of pethidine, trichloroethylene, and
6/22/2009
884
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Entonox on fetal and neonatal acid-base and P02. BMJ 3(5774):558-560.
Raaschou-Nielsen O, Hansen J, McLaughlin JK, Kolstad H, Christensen JM, Tarone RE, Olsen
JH. (2003) Cancer risk among workers at Danish companies using trichloroethylene: a cohort
study. Am J Epidemiol 158:1182-1192.
Radican L, Blair A, Stewart P, Wartenberg D. 2008. Mortality follow-up in hydrocarbon-
exposed workers. J Occup Environ Med 50:1306-1319.
Ramsdell, AF; Markwald. RR. (1997) Induction of endocardial cushion tissue in the avian heart
is regulated, in part, by TGFP-3-mediated autocrine signaling. Develop Biol 188:64-74.
Rasmussen K, Sabroe S, Wohlert M, Ingerslev HJ, Kappel B, Nielsen J. (1988) A genotoxic
study of metal workers exposed to trichloroethylene. Sperm parameters and chromosome
aberrations in lymphocytes. Int Arch Occup Environ Health 60(6):419-423.
Ray, A; Chatteijee, S; Bagchi, P; Das, TK; Deb, C. (1987) Effect of Quinalphos on testicular
steroidogenesis in rats. Andrologia 19(3): 163-168.
Rennix CP, Quinn MM, Amoroso PJ, Eisen E, Wegman DH. (2005) Risk of breast cancer among
enlisted Army women occupational exposed to volatile organic compounds. Am J Ind Med
48:157-167.
Richard, AM; Hunter, ES. (1996) Quantitative structure-activity relationships for the
developmental toxicity of haloacetic acids in mammalian whole embryo culture. Teratology
53:352-360.
Ritz B. (1999) Cancer mortality among workers exposed to chemicals during uranium
processing. J Occup Environ Med. 41:556-566.
Rodenbeck SE, Sanderson LM, Rene A. (2000) Maternal exposure to trichloroethylene in
drinking water and birth-weight outcomes. Arch Environ Health 55(3): 188-194.
Rufer, ES; Hacker, T; Lough, J; Smith, SM. (2008) Low-dose trichloroethylene exposure during
valvuloseptal morphogenesis causes ventricular septal defects in hatched chicks. Toxicologist
102(S-1):316.
Sagawa K, Nishitani H, Kawai H, Kuge Y, Ikeda M. (1973) Transverse lesion of spinal cord
after accidental exposure to trichloroethylene. Int Arch Arbeitsmed 31(4):257-64
6/22/2009
885
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Saihan EM, Burton JL, Heaton KW. (1978) A new syndrome with pigmentation, scleroderma,
gynaecomastia, Raynaud's phenomenon and peripheral neuropathy. Br J Dermatol
99(4):437-440.
Saillenfait, AM; Langonne, I; Sabate, JP. (1995) Developmental toxicity of trichloroethylene,
tetrachloroethylene and four of their metabolites in rat whole embryo culture. Arch Toxicol
70:71-82.
Sallmen M, Lindbohm ML, Kyyronen P, Nykyri E, Anttila A, Taskinen H, et al. (1995) Reduced
fertility among women exposed to organic solvents. Am J Ind Med 27(5):699-713.
Sallmen M, Lindbohm ML, Anttila A, Kyyronen P, Taskinen H, Nykyri E, et al. (1998) Time to
pregnancy among the wives of men exposed to organic solvents. Occup Environ Med
55(l):24-30.
Saxena, DK; Srivasta, RS; Lai, B; Chandra, SV. (1987) The effect of lead exposure on the testis
of growing rats. Exp Pathol 31:249-252.
Schwetz, BA; Leong, BKJ; Gehring, PJ. (1975) The effect of maternally inhaled
trichloroethylene, perchloroethylene, methyl chloroform, and methylene chloride on embryonal
and fetal development in mice and rats. Toxicol Appl Pharmacol 32:84-96.
Selmin, O; Thorne, PA; Caldwell, PT; Taylor, MR. (2008) Trichloroethylene and trichloroacetic
acid regulate calcium signaling pathways in murine embryonal carcinoma cells PI9.
Cardiovascular Toxicology. Doi:10.1007/s 12012-008-9014-2.
Sherins, RJ; Hodgen, GD. (1976) Testicular GT and SDH indices of Sertoli and germ cell
function in man. J Reprod Fertil 48:191-193.
Shu XO, Stewart P, Wen WQ, Han D, Potter JD, Buckley JD, Heineman E, Robison LL. (1999)
Parental occupational exposure to hydrocarbons and risk of acute lymphocytic leukemia in
offspring. Cancer Epidemiol Biomarkers Prev 8(9):783-791.
Siemiatycki J. (1991) Risk factors for cancer in the workplace. Boca Raton: CRC Press.
Sinks T, Lushniak B, Haussler BJ, Sniezek J, Deng J-F, Roper P, Dill P, Coates R. (1992) Renal
cell cancer among paperboard printing workers. Epidemiol 3:483-489.
6/22/2009
886
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Smith MK; Randall, JL; Read, EJ; Stober, JA. (1989) Teratogenic activity of trichloroacetic acid
in the rat. Teratology 40:445-451.
Smith MK; Randall, JL; Read, EJ; Stober, JA. (1992) Developmental toxicity of dichloroacetate
in the rat. Teratology 46:217-223.
Srivastava, S; Srivastava, SP. (1991) Effect of di(2-ethylhexyl)phthalate on 17 P-hydroxy steroid
dehydrogenase. Toxicology Letters 57:235-239.
Staples, RE. (1974) Detection of visceral alterations in mammalian fetuses. Teratology
9:A37-A38.
Stuckhardt, JL; Poppe, SM. (1984) Fresh visceral examination of rat and rabbit fetuses used in
teratogenicity testing. Teratog Carcinog Mutagen 4:181-188.
Sung TI, Chen PC, Jvuhn-Hsiarn Lee L. Lin YP, Hsieh GY, Wang JD. (2007) Increased
standardized incidence ratio of breast cancer in female electronics workers. BMC Public Health
7:102.
Swan SH, Shaw G, Harris JA, Neutra RR. (1989) Congenital cardiac anomalies in relation to
water contamination, Santa Clara County, California, 1981-1983. Am J Epidemiol
129(5):885-93.
Taskinen H, Anttila A, Lindbohm ML, Sallmen M, Hemminki K. (1989) Spontaneous abortions
and congenital malformations among the wives of men occupationally exposed to organic
solvents. Scand J Work Environ Health 15(5): 345-352.
Taskinen H, Kyyronen P, Hemminki K, Hoikkala M, Lajunen K, Lindbohm ML. (1994)
Laboratory work and pregnancy outcome. J Occup Med 36(3): 311-319.
Taylor, DH; Lagory, KE; Zaccaro, DJ; Pfohl, RJ; Laurie, RD. (1985) Effect of trichloroethylene
on the exploratory and locomotor activity of rats exposed during development. Sci Total Environ
47:415-420.
Till C, Westall CA, Rovet JF, Koren G. (2001) Effects of maternal occupational exposure to
organic solvents on offspring visual functioning: a prospective controlled study. Teratology
64:134-141.
Tola S, Vilhunen R, Jarvinen E, Korkala ML. (1980) A cohort study on workers exposed to
6/22/2009
887
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
trichloroethylene. J Occup Med 22(11):737-740.
Toth B; Hornung D; Scholz C; Djalali S; Friese K; Jeschke U. (2007) Peroxisome proliferator-
activated receptors: new players in the field of reproduction. Am J Reprod Immunol 58:298-310.
US EPA. (2001) Trichloroethylene Health Risk Assessment: Synthesis and Characterization.
EPA/600/P-01/002A, Office of Research and Development, Washington, DC.
Vartiainen T, Pukkala E, Rienoja T, Strandman T, Kaksonen K. (1993) Population exposure to
tri-and tetrachloroethene and cancer risk: two cases of drinking water pollution. Chemosphere
27:1171-1181.
Veeramachaneni, DNR; Palmer, JS; Amann, RP. (2001) Long-term effects on male reproduction
of early exposure to common chemical contaminants in drinking water. Hum Reprod
16(5):979-987.
Warren, DA; Graeter, LJ, Channel, SR; Eggers, JS; Goodyear, CD; Macmahon, KL; Sudberry,
GL; Latendresse, JR; Fisher, JW; Baker, WH. (2006) Trichloroethylene, trichloroacetic acid, and
dichloroacetic acid: do they affect eye development in the Sprague-Dawley rat? Int J Toxicol
25(4):279-284.
Watson, RE; Jacobson, CF; Williams, AL; Howard, WB; DeSesso, JM. (2006)
Trichloroethylene-contaminated drinking water and congenital heart defects: a critical analysis of
the literature. Reprod Toxicol 21:117-147.
Weiderpass E, Pukkala E, Kauppinen T, Mutanen P, Paakkulainen H, Vasama-Neuvonen K,
Boffetta P, Partanen T. (1999) Breast cancer and occupational exposure in women in Finland.
Am JInd Med 36:48-53.
Weiss JR, Moysich KB, Swede H. 2005. Epidemiology of male breast cancer. Cancer
Epidemiol Biomarkers Prev 14:20-26.
Wendling O; Chambon P; Mark M. (1999) Retinoid X receptors are essential for early mouse
development and placentogenesis. Proc Nat Acad Sci USA 96:547-551.
Westergren, I; Kjellstrand, P; Linder, LE; Johansson, BB. (1984) Reduction of brain specific
gravity in mice prenatally exposed to trichloroethylene. Toxicology Letters 23:223-226.
White RF, Feldman RG, Eviator, II, Jabre JF, Niles CA. (1997) Hazardous waste and
6/22/2009
888
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
neurobehavioral effects: a developmental perspective. Environ Res 73(1-2): 113-124.
Wigle DT, Turner MC, Gomes J, Parent ME. (2008) Role of hormonal and other factors in
human prostate cancer. J Toxicol Environ Health B Crit Rev 11:242-259.
Wilcosky TC, Checkoway H, Marshall EG, Tyroler HA. (1984) Cancer mortality and solvent
exposures in the rubber industry. Am Ind Hyg Assoc J 45(12):809-11.
Williams, AL; DeSesso, JM. (2008) Trichloroethylene and ocular malformations: analysis of
extant literature. Int J Toxicol 27:81-95.
Wilson, JG. (1965) Methods for administering agents and detecting malformations in
experimental animals. In: Wilson, JG; Warkany, T (ed.) Teratology Principles and Methods. The
University of Chicago Press, Chicago, IL, pp.262-267.
Windham GC, Shusterman D, Swan SH, Fenster L, Eskenazi B. (1991) Exposure to organic
solvents and adverse pregnancy outcome. Am J Ind Med 20(2):241-259.
Windham GC, Zhang L, Gunier R, Croen LA, Grether JK. (2006) Autism spectrum disorders in
relation to distribution of hazardous air pollutants in the San Francisco Bay area. Environ Health
Perspect 114(9): 1438-1444.
Woolhiser, MR; Krieger, SM; Thomas, J; Hotchkiss, JA. (2006) Trichloroethylene (TCE):
Immunotoxicity potential in CD rats following a 4-week vapor inhalation exposure. Dow
Chemical Company, Toxicology & Environmental Research and Consulting, Midland, MI, Study
ID 031020, July 5, 2006, unpublished.
Wu, KL; Berger, T. (2007) Trichloroethylene metabolism in the rat ovary reduces oocyte
fertilizibility. Chemico-Biological Interactions 170:20-30.
Wu, KL; Berger, T. (2008) Ovarian gene expression is stable after exposure to trichloroethylene.
Toxicology Letters 177:59-65.
Xu, H; Tanphaichitr, N; Forkert, P-G; Anupriwan, A; Weerachatyanukul, W; Vincent, R;
Leader, A; Wade, MG. (2004) Exposure to trichloroethylene and its metabolites causes
impairment of sperm fertilizing ability in mice. Toxicol Sci 82:590-597.
Yauck JS, Malloy ME, Blair K, Simpson PM, McCarver DG. (2004) Proximity of residence to
trichloroethylene-emitting sites and increased risk of offspring congenital heart defects among
6/22/2009
889
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
older women. Birth Defects Res A Clin Mol Teratol 70(10) 808-814.
Zenick, H; Blackburn, K; Hope, E; Rischdale, N; Smith, MK. (1984) Effects of trichloroethylene
exposure on male reproductive function in rats. Toxicology 31:237-250
Zhao Y, Krishnadasan A, Kennedy N, Morgenstern H, Ritz B. (2005) Estimated effects of
solvents and mineral oils on cancer incidence and Mortality in a cohort of aerospace workers.
Am J Ind Med 48:249-258.
Zielinski, A. (1973) General health state of women professionally exposed to trichloroethylene
vapours. Med Pr 24:263-271.
1
6/22/2009
890
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.8 Other site-specific cancers
4.8.1 Esophageal cancer
Increasing esophageal cancer incidence has been observed in males, but not females in
the United States between 1975 and 2002, a result of increasing incidence of esophageal
adenocarcinoma (Ward et al., 2006). Males also have higher age-adjusted incidence and
mortality rates (incidence, 7.8 per 100,000; mortality, 7.8 per 100,000) than females (incidence,
2.0 per 100,000; mortality, 1.7 per 100,000) (Ries et al., 2008). Survival for esophageal cancer
remains poor and age-adjusted mortality rates are just slightly lower than incidence rates. Major
risk factors associated with esophageal cancer are smoking and alcohol for squamous cell
carcinoma, typically found in the upper third of the esophagus, and obesity, gastroesophageal
reflux, and Barrett's esophagus for adenocarcinoma that generally occurs in the lower esophagus
(Ward et al., 2006).
Sixteen epidemiologic studies on TCE exposure reported relative risks for esophageal
cancer (Garabrant et al., 1988; Costa et al., 1989; Siemiatycki, 1991; Greenland et al., 1994;
Blair et al., 1998; Boice et al., 1999, 2006; Ritz, 1999; Hansen et al., 2001; Raaschou-Nielsen et
al., 2003; AT SDR, 2004, 2006; Zhao et al., 2005; Sung et al., 2007; Clapp and Hoffman, 2008;
Radican et al., 2008). Ten studies had high likelihood of TCE exposure in individual study
subjects and were judged to have met, to a sufficient degree, the standards of epidemiologic
design and analysis (Siemiatycki, 1991; Greenland et al., 1994; Blair et al., 1998; Boice et al.,
1999, 2006; Ritz, 1999; Hansen et al., 2001; Raaschou-Nielsen et al., 2003; Zhao et al., 2005;
Radican et al., 2008). Four studies with high quality information (Axelson et al., 1994; Anttila et
al., 1995; Blair et al., 1998 [Incidence]; Morgan et al., 1998) do not present relative risk
estimates for esophageal cancer and TCE exposure nor do two other studies which carry less
weight in the analysis because of design limitations (Sinks et al., 1992; Henschler et al., 1995).
Only Raaschou-Nielsen et al. (2003) examines esophageal cancer histologic type, an important
consideration given differences between suspected risk factors for adenocarcinoma and those for
squamous cell carcinoma. Appendix C identifies these study's design and exposure assessment
characteristics.
Several population case-control studies (Yu et al., 1988; Gustavsson et al., 1998; Parent
et al., 2000; Weiderpass et al., 2003; Engel et al., 2002; Ramanakumar et al., 2008; Santibanez et
al., 2008 [In press]) examine esophageal cancer and organic solvents or occupational job titles
with past TCE use documented (Bakke et al., 2006). Relative risk estimates in case-control
studies that examine metal occupations or job titles, or solvent exposures are found in Table
6/22/2009
891
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.8.1. The lack of exposure assessment to TCE, low prevalence of exposure to chlorinated
hydrocarbon solvents, or few exposed cases and controls in those studies lowers their sensitivity
for informing evaluations of TCE and esophageal cancer.
Table 4.8.2 presents risk estimates for TCE exposure and esophageal cancer observed in
cohort, case-control, and geographic based studies. Ten studies in which there is a high
likelihood of TCE exposure in individual study subjects (e.g., based on job-exposure matrices or
biomarker monitoring) reported risk estimates for esophageal cancer (Siemtiatycki, 1991;
Greenland et al., 1994; Blair et al., 1998; Boice et al., 1999; Ritz et al., 1999; Hansen et al.,
2001; Raaschou-Nielsen et al., 2003; Zhao et al., 2005; Boice et al., 2006; Radican et al., 2008).
Some evidence for association with esophageal cancer and overall TCE exposure comes from
studies with high likelihood of TCE exposure (5.6, 95% CI: 0.7, 44.5 [Blair et al., 1998] and
1.88, 95% CI: 0.61, 5.79 (Radican et al., 2008, which was an update of Blair et al., 1998 with an
additional 10 years of follow-up); 4.2, 95% CI: 1.5, 9.2, (Hansen et al., 2001); 1.2, 95% CI: 0.84,
1.57 (Raaschou-Nielsen et al., 2003)]. Two studies support an association with adenocarcinoma
histologic type of esophageal cancer and TCE exposure (five of the six observed esophageal
cancers were adenocarcinomas (less than 1 expected) (Hansen et al., 2001); 1.8, 95% CI: 1.2, 2.7
(Raaschou-Nielsen et al., 2003)]. Risk estimates in other high-quality studies are based on few
deaths, low statistical power to detect a doubling of esophageal cancer risk, and confidence
intervals which include a risk estimate of 1.0 [no increased risk].
Six other studies (Garabrant et al., 1988; Costa et al., 1989; Sung et al., 2007; ATSDR,
2004, 2006; Clapp and Hoffman, 2008) with lower likelihood for TCE exposure, in addition to
limited statistical power and other design limitations, observed relative risk estimates between
0.21 (95% CI: 0.0.01, 1.17) (Costa et al., 1989) to 1.14 (95% CI: 0.62, 1.92) (Garabrant et al.,
1988). For these reasons, esophageal cancer observations in these studies are not inconsistent
with Blair et al. (1998) and its update Radican et al. (2008), Hansen et al., (2001), or Raaschou-
Nielsen et al. (2003). No study reported a statistically significant deficit in the esophageal cancer
risk estimate and overall of TCE exposure.
Of those studies with exposure-response analyses, a pattern of increasing esophageal
cancer relative risk with increasing exposure metric is not generally noted (Siemiatycki, 1991;
Blair et al., 1998; Boice et al., 1999; Zhao et al., 2005; Radican et al., 2008) except for Hansen et
al. (2001) and Raaschou-Nielsen et al. (2003). In these last two studies, esophageal cancer
relative risk estimates associated with long employment duration were slightly higher [SIR = 6.6,
95% CI: 1.8, 7.0.8, 3.7 (Hansen et al., 2001); SIR = 1.9, 95% CO: 0.8, 3.7 (Raaschou-Nielsen et
al., 2003)] than those for short employment duration [SIR = 4.4, 95% CI: 0.5, 19 (Hansen et al.,
2001); SIR = 1.7, 95% CI: 0.6, 3.6 (Raaschou-Nielsen et al., 2003). Hansen et al. (2001) also
6/22/2009
892
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
reports risk for two other TCE exposure surrogates, average intensity and cumulative exposure,
and in both cases observed lower risk estimates with the higher exposure surrogate.
Meta-analysis is not adopted as a tool for examining the body of epidemiologic evidence
on esophageal cancer and TCE exposure given the absence of reported relative risk estimates in
several of the high-quality studies.
Overall, three high-quality cohort studies provide some evidence of association for
esophageal cancer and TCE exposure. The finding in two of these studies of esophageal risk
estimates among subjects with long employment duration were higher than those associated with
low employment duration provides additional evidence (Hansen et al., 2001; Raaschou-Nielsen
et al., 2003). The cohort studies are unable to directly examine possible confounding due to
suspected risk factors for esophageal cancer such as smoking, obesity and alcohol. The use of an
internal referent group, similar in socioeconomic status as exposed subjects, is believed to
minimize but may not completely control for possible confounding related to smoking and health
status (Blair et al., 1998; its follow-up Radican et al., 2008; Zhao et al., 2005; Boice et al, 2006).
Observation of a higher risk for adenocarcinoma histologic type than for a combined category of
esophageal cancer in Raaschou-Nielsen et al. (2003) also suggests minimal confounding from
smoking. Smoking is not identified as a possible risk factor for the adenocarcinoma histologic
type of esophageal cancer but is believed a risk factor for squamous cell histologic type.
Furthermore, the magnitude of lung cancer risk in Raaschou-Nielsen et al. (2003) suggests a high
smoking rate is unlikely. The lack of association with overall TCE exposure and the absence of
exposure-response patterns in the other studies of TCE exposure may reflect limitations in
statistical power, the possibility of exposure misclassification, and differences in measurement
methods. These studies do not provide evidence against an association between TCE exposure
and esophageal cancer.
6/22/2009
893
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TABLE 4.8.1: Selected observations from case-control studies of TCE exposure and esophageal cancer
Moderate exposure
Study
High exposure
Population Exposure Group
Test for trend
0.5 (0.1, 3.9)
1
0.4 (0.1, 1.5)
2
All Esophageal Cancers
Relative Risk No. obs.
0.4(0.1, 1.8)1 2
Squgnjcjus Cell Cancer
Relative Risk (95% No. obs.
0.9 (0.5, 1.6)1
Adc^iq^ircinoma
Relative Risk (95%
12
No.
(95% CI)
events
CI)
events
CI)
ever
Population of Regions in Eastern Spain
Metal molders, welders, etc.
0.94 (0.14,6.16)
3
0.40 (0.05,3.18)
2
3.55 (0.28, 44.70)
1
Metal-processing plant operators
1.14(0.29,4.44)
5
1.23 (0.23,6.51)
4
0.86 (0.08, 8.63)
1
Chlorinated hydrocarbon solvents
Low exposure
1.05 (0.15,7.17)
2
0
4.92 (0.69, 34.66)
2
High exposure
1.76 (0.40, 7.74)
6
2.18(0.41, 11.57)
5
3.03 (0.28, 32.15)
1
Population of Montreal, Canada
Painter, Metal coatings
Any exposure
1.3 (0.4,4.2)
6
Substantial exposure
4.2(1.1, 17.0)
4
Solvents
Any exposure
1.1 (0.7, 1.7)
39
1.4 (0.8, 2.5)
30
Non-substantial exposure
1.0(0.5, 1.9)
16
1.3 (0.6, 2.6)
12
Substantial exposure
1.1 (0.6, 1.9)
39
1.4 (0.8, 2.5)
30
Population of Sweden
Organic solvents
No exposure
1.0
145
1.0
128
Moderate exposure
0.7 (0.4, 1.5)
15
1.2 (0.6, 2.3)
14
High exposure
1.3 (0.7, 2.3)
21
1.4 (0.7,2.5)
18
Test for trend
p = 0.47
p = 0.59
No exposure
1.0
1.0
Reference
Santibanez et al., 2008
Ramanakumar et al., 2008;
Parent et al., 2000
Janssen et al., 2006a, b
6/22/2009
894
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Population of Finland (Females)
Chlorinated hydrocarbon solvents
Low level exposure
High level exposure
0.95 (0.54, 1.66)
0.62 (0.34, 1.13)
Not reported
Not reported
Weiderpass et al., 2003
Population of NJ, CT, WA State
Precision metal workers
Metal product manufacturing
Not reported
Not reported
0.7 (0.3, 1.5)
0.8(0.3, 1.8)
12
15
1.4 (0.8,2.3)
1.3 (0.8,2.3)
Engel et al., 2002
25
26
1 Jansson et al. (2006b) is a registry-based study of the Swedish Construction Worker Cohort. Relative risks are incidence rate ratios
from Cox regression analysis using calendar time and adjustment for attained age, calendar period at entry into the cohort, tobacco
smoking status at entry into the cohort and BMI at entry into the cohort.
6/22/2009 895
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.8.2. Summary of human studies on TCE exposure and esophageal cancer
Exposure Group
Cohort Studies - Incidence
Aerospace workers (Rocketdyne)
Any exposure to TCE
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
Relative Risk
(95% CI)
Not reported
1.001
1.66 (0.62, 4.41f
0.82 (0.17, 3.95)2
p = 0.974
No. obs.
events
Reference
Zhao et al., 2005
All employees at electronics factory (Taiwan)
Males
Females
Not reported
1.16 (0.0.14, 4.20)3
Sung et al., 2007
Danish blue-collar worker w/TCE exposure
Any exposure, all subjects 1.2 (0.84,1.57) 44
Any exposure, males 1.1(0.81,1.53) 40
Any exposure, females 2.0(0.54,5.16) 4
Raaschou-Nielsen et al., 2003
Exposure Lag Time
Employment duration
Any exposure, males 1.8(1.15,2.73)
Any exposure, females
20 years 1.7 (0.8, 3.0)4
<1 year 1.7 (0.6, 3.6)4
1-4.9 years 1.9 (0.9, 3.6)4
>5 years 1.9 (0.8, 3.7)4
Subcohort w/higher exposure
Any TCE exposure 1.7(0.9,2.9)
Employment duration
1-4.9 years 1.6 (0.6, 3.4)4
>5 years 1.9 (0.8, 3.8)4
23
0 (0.4 exp)4
10
6
9
8
13
6
7
Biologically-monitored Danish workers
Any TCE exposure, males
Adenocarcinoma histologic type
Any TCE exposure, females
Cumulative exp (Ikeda)
<17 ppm-yr
>17 ppm-yr
4.0 1.5, 8.72)
4.2(1.5, 9.2)
3.6(1.2, 8.3)5
6.5 (1.3, 19)
4.2(1.5,9.2)
6
6
5
0 (0.1 exp)
Hansen etal., 2001
6/22/2009
896
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Mean concentration (Ikeda)
<4 ppm
4+ ppm
Employment duration
< 6.25 yr
>6.25
Aircraft maintenance workers from Hill Air Force Base
TCE subcohort
Males, Cumulative exp
0
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
Females, Cumulative exp
0
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
Biologically-monitored Finnish workers
All subjects
Mean air-TCE (Ikeda extrapolation)
<6 ppm
6+ ppm
Cardboard manufacturing workers in Arnsburg, Germany
Exposed workers
Biologically-monitored Swedish workers
Any TCE exposure, males
Any TCE exposure, females
Cardboard manufacturing workers, Atlanta area, GA
All subjects
Cohort Studies-Mortality
Computer manufacturing workers (IBM), NY
Males
Aerospace workers (Rocketdyne)
Any TCE (utility/eng flush)
8.0 (2.6, 19) 5
1.3(0.02,7.0) 1
4.4 (0.5, 16) 2
6.6(1.8,17) 4
Blair etal., 1998
Not reported
1.01
Not reported
Not reported
Not reported
1.01
Not reported
Not reported
Not reported
Anttila et al., 1995
Not reported
Not reported
Not reported
Henschler et al., 1995
Not reported
Axelson et al., 1994
Not reported
Not reported
Sinks et al., 1992
Not reported
Clapp and Hoffman, 2008
1.12 (0.30, 2.86)6
5.24 (0.13, 29.2)6
0.88 (0.18,2.58) 3 Boice et al., 2006
6/22/2009
897
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Any exposure to TCE
Low cum TCE score
Not reported
1.001
Zhao et al., 2005
18
Med cum TCE score
1.40 (0.70, 2.82)2
15
High TCE score
1.27 (0.52, 3.13)2
7
p for trend
p = 0.535
View-Master employees
ATSDR, 2004
Males
0.62 (0.02, 3.45)6
1
Females
0 (1.45 exp)6
All employees at electronics factory (Taiwan)
Chang et al., 2003
Males
0 (3.34 exp)
Females
0 (0.83 exp)
US Uranium-processing workers (Fernald)
Ritz, 1999
Any TCE exposure
Not reported
Light TCE exposure, >2 years duration
2.61 (0.99, 6.88)7
12
Mod TCE exposure, >2 years duration
0
Aerospace workers (Lockheed)
Boice et al., 1999
Routine Exp
0.83 (0.34, 1.72)
7
Routine-Intermittent1
Not presented
11
Duration of exposure
0 years
1.01
28
< 1 year
0.23 (0.05, 0.99)
2
1-4 years
0.57 (0.20, 1.67)
4
> 5 years
0.91 (0.38, 2.22)
7
p for trend
p > 0.20
Aerospace workers (Hughes) Morgan et al., 1998
TCE Subcohort Not reported
Low Intensity (<50 ppm)
High Intensity (>50 ppm)
TCE Subcohort (Cox Analysis) Not reported
Never exposed
Ever exposed
Peak Not reported
No/Low
Med/Hi
Cumulative Not reported
Referent
6/22/2009 898
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Low
High
Aircraft maintenance workers (Hill AFB, Utah)
TCE subcohort
5.6(0.7, 44.5)1
10
Males, Cumulative exp
0
1.01
< 5 ppm-yr
Not reported8
3
5-25 ppm-yr
Not reported8
2
>25 ppm-yr
Not reported8
4
Females, Cumulative exp
0
1.01
< 5 ppm-yr
3.6 (0.2, 58)
1
5-25 ppm-yr
0
>25 ppm-yr
0
TCE subcohort
1.88 (0.61,5.79)
17
Males, Cumulative exp
1.66 (0.48, 5.74)
15
0
1.01
< 5 ppm-yr
1.84 (0.48,7.14)
7
5-25 ppm-yr
1.33 (0.27, 6.59)
3
>25 ppm-yr
1.67 (0.40, 7.00)
5
Females, Cumulative exp
2.81 (0.25, 31.10)
2
0
1.01
< 5 ppm-yr
3.99 (0.25, 63.94)
1
5-25 ppm-yr
9,.59 (0.60, 154.14)
1
>25 ppm-yr
0
Cardboard manufacturing workers in
Arnsburg, Germany
TCE exposed workers
Not reported
Unexposed workers
Not reported
Deaths reported to among GE pension fund (Pittsfield, MA) 0.95 (0.1, 3.17)
Cardboard manufacturing workers, Atlanta area, GA Not reported
13
Blair etal., 1998
Radican et al„ 2008
Henschler et al., 1995
Greenland et al., 1994
Sinks et al., 1992
Aircraft manufacturing plant employees (Italy) Costa et al., 1989
All subjects 0.21(0.01,1.17) 1
Rubber Workers Not reported9 Wilcosky et al., 1984
Aircraft manufacturing plant employees (San Diego, CA) Garabrant et al., 1988
6/22/2009 899
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
All subjects
1.14(0.62,1.92) 14
Case-control Studies
Population of Montreal, Canada
Any TCE exposure
Substantial TCE exposure
Siemiatycki et al., 1991; Parent
et al., 2000
0.5 (0.1, 2.5)10
0.8 (0.1, 4.6)10
1
1
Geographic Based Studies
Residents in two study areas in Endicott, NY
0.78 (0.29, 1.70) 6
ATSDR, 2006
Residents of 13 census tracts in Redlands, CA
Not reported
Morgan and Cassidy, 2002
Finnish residents
Residents of Hausjarvi
Residents of Huttula
Not reported
Not reported
Vartiainen et al., 1993
1 Internal referents, workers not exposed to TCE
2 Ritz (1999) and Zhao et al. (2005) reported relative risks for the combined site of esophagus and stomach
3 Sung et al. (2007) Chang et al. (2005) - Standardized incidence ratio (SIR) for females and reflects a 10-year lag
period
4 SIR for adenocarcinoma of the esophagus
5 The SIR for adenocarcinoma histologic type can not be calculated because Hansen et al. (2001) do not present
expected numbers for adenocarcinoma histologic type of esophageal cancer. An approximation of the SIR for
adenocarcinoma histologic type is presented using the expected number of total number of expected esophageal
cancers for males (n = 1.4). The expected numbers of esophageal adenocarcinomas in males will be lower; Hansen
et al. (2001) noted the proportion of adenocarcinomas among the comparable Danish male population during the
later period of the study (1990-1996) as 38%. A rough approximation of the expected number of esophageal
carcinomas would be 0.5 expected cases and an approximated SIR of 9.4 (3.1, 22).
6 Proportional mortality ratio
7 Adjusted relative risks for >2 year exposure duration and 15 year lag from 1st exposure
8 No esophageal cancer deaths occurred in the referent population in Blair et al. (1998) and relative risk in could not
be calculated for this reason
9 Odds ratio from nested case-control analysis
10 90% Confidence Interval
6/22/2009
900
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.8.2 Bladder Cancer
Twenty-three epidemiologic studies present risk estimates for bladder cancer (Garabrant
et al., 1988; Costa et al., 1989; Mallin, 1990; Siemiatycki, 1991; Sinks et al., 1992; Axelson et
al., 1994; Greenland et al., 1994; Anttila et al., 1995; Blair et al., 1998; Morgan et al., 1998;
Boice et al., 1999, 2006; Pesch et al., 2000b; Hansen et al., 2001; Cassidy and Morgan, 2002;
Chang et al., 2003, 2005; Raaschou-Nielsen et al., 2003; ATSDR, 2004, 2006; Zhao et al., 2005;
Sung et al., 2007; Radican et al., 2008). Thirteen studies, all either cohort or case-control
studies, which there is a high likelihood of TCE exposure in individual study subjects (e.g.,
based on job-exposure matrices or biomarker monitoring) or which met, to a sufficient degree,
the standards of epidemiologic design and analysis in a systematic review, reported relative risk
estimates for bladder or urothelial cancer between 0.6 (Siemiatycki, 1991) and 1.7 (Boice et al.,
2006) and overall TCE exposure. Relative risk estimates were generally based on small numbers
of cases or deaths, except for one study (Raaschou-Nielsen et al., 2004), with the result of wide
confidence intervals on the estimates. Of high-quality studies, two reported statistically
significant elevated bladder or urothelial cancer risks with the highest cumulative TCE exposure
category [2.71, 95% CI: 1.10, 6.65 (Morgan et al., 1998); 1.8, 95% CI: 1.2, 2.7 (Pesch et al.,
2000b)] and five presented risk estimates and categories of increasing cumulative TCE exposure
(Blair et al., 1998; Morgan et al., 1998; Pesch et al., 2000b; Zhao et al., 2005; Radican et al.,
2008). Risk estimates in Morgan et al. (1998), Pesch et al. (2000b), and Zhao et al. (2005)
appeared to increase with increasing cumulative TCE exposure with the p-value for trend of 0.07
in Zhao et al. (2005), the only study to present a formal statistical test for linear trend. Risk
estimates did not appear to either increase or decrease with increasing cumulative TCE exposure
in Blair et al. (1998) or its update Radican et al. (2008), which added another 10 years of follow-
up. Ten additional studies were given less weight because of their lesser likelihood of TCE
exposure and other design limitations that would decrease statistical power and study sensitivity
(Garabrant et al., 1988; Costa et al., 1989; Mallin, 1990; Sinks et al., 1992; Cassidy and Morgan,
2002; Chang et al., 2003, 2005; ATSDR, 2004, 2006; Sung et al., 2007).
Meta-analysis is not adopted as a tool for examining the body of epidemiologic evidence
on bladder cancer and TCE.
Overall, three high-quality cohort or case-control studies provide some evidence of
association for bladder or urothelial cancer and high cumulative TCE exposure (Morgan et al.,
1998; Pesch et al., 2000b; Zhao et al., 2005). The case-control study of Pesch et al. (2000b)
adjusted for age, study center, and cigarette smoking, with a finding of a statistically significant
6/22/2009
901
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
risk estimate between urothelial cancer and the highest TCE exposure category. Cancer cases in
this study are of several sites, bladder, ureter, and renal pelvis, and grouping different site-
specific cancers with possible etiologic heterogeneity may introduce misclassification bias. The
cohort studies are unable to directly examine possible confounding due to suspected risk factors
for esophageal cancer such as smoking, obesity, and alcohol. The use of an internal referent
group, similar in socioeconomic status as exposed subjects, by Morgan et al. (1998) and Zhao et
al. (2005) is believed to minimize but may not completely control for possible confounding
related to smoking and health status. The lack of association with overall TCE exposure in other
studies and the absence of exposure-response patterns with TCE exposure in Blair et al. (1998)
and Radican et al. (2008) may reflect limitations in statistical power, the possibility of exposure
misclassification, and differences in measurement methods. These studies do not provide
evidence against an association between TCE exposure and bladder cancer.
6/22/2009
902
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.8.3 Summary of human studies on TCE exposure and bladder cancer
Relative Risk No. obs.
Exposure Group (95% CI) events
Cohort Studies - Incidence
Aerospace workers (Rocketdyne)
Any exposure to TCE
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
Not reported
1.001
1.54 (0.81, 2.92)2
1.98 (0.93, 4.22)2
p = 0.069
20
19
11
Reference
Zhao et al., 2005
TCE, 20 years exposure lag
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
1.001 20
1.76 (0.61, 5.10)3 20
3.68 (0.87, 15.5)3 10
p = 0.064
All employees at electronics factory (Taiwan)
Males
Females
Males
Females
Not reported
0.34 (0.07, 1.00) 10
1.06 (0.45, 2.08)4 8
1.09 (0.56, 1.91)4 12
Sung et al., 2007
Chang et al., 2005
Danish blue-collar worker w/TCE exposure Raaschou-Nielsen et al., 2003
Any exposure, all subjects 1.1 (0.92,1.21) 220
Any exposure, males 1.0(0.89,1.18) 203
Any exposure, females 1.6 (0.93,2.57) 17
Biologically-monitored Danish workers
Any TCE exposure, males
Any TCE exposure, females
1.0 (0.48, 1.86)
1.1 (0.50, 2.0)
0.5 expected
10
10
0
Hansen et al., 2001
Aircraft maintenance workers from Hill Air Force Base Blair et al., 1998
TCE subcohort Not reported
Males, Cumulative exp
Females, Cumulative exp
0
1.01
< 5 ppm-yr
1.7 (0.6,4.4)
13
5-25 ppm-yr
1.7 (0.6,4.9)
9
>25 ppm-yr
1.4 (0.5,4.1)
9
0
1.01
< 5 ppm-yr
1.1 (0.1, 10.8)
1
5-25 ppm-yr
0
6/22/2009 903
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
>25 ppm-yr 1.0(0.1,9.1)
TCE subcohort
0.80 (0.41, 1.58)
25
Males, Cumulative exp
1.05 (0.47, 2.35)
24
0
1.01
< 5 ppm-yr
0.96 (0.37,2.51)
9
5-25 ppm-yr
1.77 (0.70, 4.52)
10
>25 ppm-yr
0.67 (0.15,2.95)
5
Females, Cumulative exp
0.22 (0.03, 1.83)
1
0
1.01
< 5 ppm-yr
0
5-25 ppm-yr
2.86 (0.27, 29.85)
1
>25 ppm-yr
0
Biologically-monitored Finnish workers
All subjects
0.82 (0.27, 1.90)
5
Biologically-monitored Swedish workers
Any TCE exposure, males
1.02 (0.44, 2.00)
8
Any TCE exposure, females
Not reported
Cohort Studies-Mortality
Aerospace workers (Rocketdyne)
Any TCE (utility/eng flush)
1.66 (0.54, 3.87)
5
Any exposure to TCE
Not reported
Low cum TCE score
1.001
8
Med cum TCE score
1.27 (0.43, 3.73)2
6
High TCE score
1.15 (0.29, 4.51)2
3
p for trend
p = 0.809
TCE, 20 years exposure lag
Low cum TCE score
1.001
8
Med cum TCE score
0.95 (0.15, 6.02)3
7
High TCE score
1.85 (0.12, 27.7)3
2
p for trend
p = 0.533
View-Master employees
Males
1.22 (0.15,4.40)
Females
0.78 (0.09, 2.82)
Radican et al., 2008
Anttila et al., 1995
Axelson et al., 1994
Boice et al., 2006
Zhao et al., 2005
ATSDR, 2004
US Uranium-processing workers (Fernald)
Any TCE exposure
Not reported
Ritz, 1999
6/22/2009
904
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE
O
o
Light TCE exposure, >2 years duration
Not reported
Mod TCE exposure, >2 years duration
Not reported
Aerospace workers (Lockheed)
Routine Exp
0.55 (0.18, 1.28)
5
Routine-Intermittent1
Not reported
Aerospace workers (Hughes)
TCE Subcohort
1.36 (0.59, 2.68)
8
Low Intensity (<50 ppm)
0.51 (0.01,2.83)
1
High Intensity (>50 ppm)
1.79 (0.72, 3.69)
7
TCE Subcohort (Cox Analysis)
Never exposed
1.01
Ever exposed
2.05 (0.86, 4.85)5
8
Peak
No/Low
1.01
Med/Hi
1.41 (0.52, 3.81)
5
Cumulative
Referent
1.01
Low
0.69 (0.09, 5.36)
1
High
2.71 (1.10,6.65)
7
Aircraft maintenance workers (Hill AFB, Utah)
TCE subcohort
1.2 (0.5, 2.9)1
17
Males, Cumulative exp
0
1.01
< 5 ppm-yr
1.8 (0.5, 6.2)
7
5-25 ppm-yr
2.1 (0.6,8.0)
5
>25 ppm-yr
1.0 (0.2,5.1)
3
Females, Cumulative exp
0
1.01
< 5 ppm-yr
0
5-25 ppm-yr
0
>25 ppm-yr
0.8 (0.1,7.5)
1
Cardboard manufacturing workers in Arnsburg, Germany
TCE exposed workers
Not reported
Unexposed workers
Not reported
Deaths reported to GE pension fund (Pittsfield, MA)
Cardboard manufacturing workers, Atlanta area, GA
0.85 (0.32, 2.23)°
0.3 (0.0, 1.6)
20
Boice et al., 1999
Morgan etal., 1998
Blair etal., 1998
Henschler et al., 1995
Greenland et al., 1994
Sinks et al., 1992
6/22/2009
905
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Aircraft manufacturing plant employees (Italy)
All subjects
0.74 (0.30, 1.53)
Costa etal., 1989
Aircraft manufacturing plant employees (San Diego, CA)
All subjects
1.26 (0.74, 2.03) 17
Garabrant et al., 1988
Case-control Studies
Population of 5 regions in Germany
Any TCE Exposure
Males
Population of Montreal, Canada
Not reported
Males Not reported
Females Not reported
Medium
00
©
1.2)7
47
High
1.3 (0.8,
1.7)7
74
Substantial
1.8(1.2,
2.7)7
36
Any TCE exposure 0.6 (0.3, 1.2)
Substantial TCE exposure 0.7 (0.3, 1.6)
Geographic Based Studies
Residents in two study areas in Endicott, NY
Residents of 13 census tracts in Redlands, CA
0.71 (0.38, 1.21) 13
0.98 (0.71, 1.29) 82
Pesch et al., 2000b
Siemiatycki, 1991; Siemiatycki
etal., 1994
ATSDR, 2006
Morgan and Cassidy, 2002
Finnish residents
Residents of Hausjarvi
Residents of Huttula
Not reported
Not reported
Vartiainen et al., 1993
Residents of 9 county area in Northwestern Illinois
All zip codes in study area
Cluster community
Mallin, 1990
Adjacent community
Males 1.4(1.1,1.9) 47
Females 1.8(1.2,2.7) 21
Males 1.7(1.1,2.6) 21
Females 2.6 (1.2, 4.7) 10
Males 1.2(0.6,2.0) 12
6/22/2009
906
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Females 1.6 (0.5, 3.8)
5
Remainder of zip code areas
Males 1.4(0.8,2.2)
Females 1.4 (0.5, 3.0)
14
6
1 Internal referents, workers not exposed to TCE
2 Relative risk estimates for TCE exposure after adjustment for 1st employment, socioeconomic status, and age at
event.
3 Relative risk estimates for TCE exposure after adjustment for 1st employment, socioeconomic status, age at event,
and all other carcinogen exposures, including hydrazine.
4 Chang et al. (2005) and Costa et al. (1989) report estimated risks for a combined site of all urinary organ cancers.
5 Risk ratio from Cox Proportional Hazard Analysis, stratified by age, sex and decade (Environmental Health
Strategies, 1997)
6 Odds ratio from nested case-control analysis
7 Odds ratio for urothelial cancer, a category of bladder, ureter, and renal pelvis cancers) and cumulative TCE
exposure, as assigned using a job-task-exposure matrix (JTEM) approach (Pesch et al., 2000b).
8 99% Confidence Interval
6/22/2009
907
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.8.3 Central Nervous System and Brain Cancers
Brain cancer is examined in most cohort studies and in one case-control study (Garabrant
et al., 1988; Costa et al., 1989; Greenland et al., 1994; Heineman et al., 1994; Anttila et al., 1995;
Henschler et al., 1995; Blair et al., 1998; Morgan et al., 1998; Boice et al., 1999, 2006; Ritz,
1999; Hansen et al., 2001; Chang et al., 2003, 2005; Raaschou-Nielsen et al., 2003; Zhao et al.,
2005; Sung et al., 2007; Clapp and Hoffman, 2008; Radican et al., 2008). Overall, these
epidemiologic studies do not provide strong evidence for or against association between TCE
and brain cancer in adults (Table 4.8.4). Relative risk estimates in well designed and conducted
cohort studies, Axelson et al. (1994), Anttila et al. (1995), Blair et al. (1998), its follow-up
reported in Radican et al. (2008), Morgan et al. (1998), Boice et al. (1999), Zhao et al. (2005),
and Boice et al. (2006), are near a risk of 1.0 and imprecise, confidence intervals all include a
risk estimate of 1.0. All studies except Raaschou-Nielsen et al. (2003), observations are based
on few events and lowered statistical power. Bias resulting from exposure misclassification is
likely in these studies, although of a lower magnitude compared to other cohort studies identified
in Table 4.8.4, and may partly explain observations. Exposure misclassification is also likely in
the case-control study of occupational exposure of Heineman et al. (1994) who do not report
association with TCE exposure.
Three geographic-based studies and one case-control study examined childhood brain
cancer (ADHS, 1990, 1995; De Roos et al., 2001; Morgan and Cassidy, 2002; ATSDR, 2006).
The strongest study, De Roos et al. (2001), a population case-control study which examined
paternal exposure, used expert judgment to evaluate the probably of TCE exposure from self-
reported information in an attempt to reduce exposure misclassification bias. The odds ratio
estimate in this study was 0.9 (95% CI: 0.3, 2.5). Like many population case-control studies, a
low prevalence of TCE exposure was found, only 9 fathers were identified with probable TCE
exposure by the industrial hygiene review, and greatly impacted statistical power. There is some
concern for childhood brain cancer and organic solvent exposure based on Peters et al. (1981)
whose case-control study of childhood brain cancer reported to the Los Angeles County Cancer
Surveillance Program observed a high odds ratio estimate for paternal employment in the aircraft
industry (OR = go, p < 0.001). This study does not present an odds ratio for TCE exposure only
although it did identify two of the 14 case and control fathers with previous employment in the
aircraft industry reported exposure to TCE.
6/22/2009
908
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Table 4.8.4. Summary of human studies on TCE exposure and brain cancer
Exposure Group
Cohort Studies - Incidence
Aerospace workers (Rocketdyne)
Any exposure to TCE
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
Relative Risk No. obs.
(95% CI) events
Not reported
1.001 7
0.46 (0.09, 2.25)2 2
0.47 (0.06, 3.95)2 1
p= 0.382
Reference
Zhao et al., 2005
All employees at electronics factory (Taiwan)
Males
Females
Males
Females
Not reported
1.07 (0.59, 1.80)3
0.40 (0.05, 1.46)
0.97 (0.54, 1.61)
2
15
Sung et al., 2007
Chang et al., 2005
Danish blue-collar worker w/TCE exposure
Any exposure, all subjects 1.0 (0.84, 1.24) 104
Any exposure, males 1.0(0.76,1.18)) 85
Any exposure, females 1.1 (0.67,1.74) 19
Raaschou-Nielsen et al., 2003
Biologically-monitored Danish workers
Any TCE exposure, males
Any TCE exposure, females
0.3 (0.01, 1.86)
0.4 (0.01,2.1)
0.5 expected
Hansen et al., 2001
Aircraft maintenance workers from Hill Air Force Base
TCE subcohort Not reported
Males, Cumulative exp
Blair etal., 1998
Females, Cumulative exp
0
1.01
< 5 ppm-yr
2.0 (0.2,
19.7)
3
5-25 ppm-yr
3.9 (0.4,
34.9)
4
>25 ppm-yr
00
©
13.2)
1
0
1.01
< 5 ppm-yr
0
5-25 ppm-yr
0
Biologically-monitored Finnish workers
All subjects
Mean air-TCE (Ikeda extrapolation)
1.09 (0.50, 2.07)
<6 ppm 1.52(0.61,3.13)
Anttila et al., 1995
6/22/2009
909
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
6+ppm 0.76 (0.01,2.74)
Biologically-monitored Swedish workers
Any TCE exposure, males
Any TCE exposure, females
Not reported
Not reported
Axelson et al., 1994
Cohort Studies-Mortality
Computer manufacturing workers (IBM), NY
Males
Females
1.90 (0.52, 4.85)
Clapp and Hoffman, 2008
Aerospace workers (Rocketdyne)
Any TCE (utility/eng flush)
Any exposure to TCE
Low cum TCE score
Med cum TCE score
High TCE score
p for trend
0.81 (0.17,2.36)
Not reported
1.001
0.42 (0.12, 1.50)
0.83 (0.23, 3.08)
p = 0.613
12
3
3
Boice et al., 2006
Zhao et al., 2005
View-Master employees
Males
Females
Not reported
Not reported
ATSDR, 2004
All employees at electronics factory (Taiwan)
Males
Females
US Uranium-processing workers (Fernald)
Any TCE exposure
Chang et al., 2003
Light TCE exposure, >2 years duration
Mod TCE exposure, >2 years duration
Aerospace workers (Lockheed)
Routine Exp
Routine-Intermittent1
0.96 (0.01, 5.36) 1
0.96 (0.01, 5.33) 1
Not reported
1.81 (0.49, 6.71)3 6
3.26 (0.37, 28.9)3 1
0.54 (0.15, 1.37)
Not presented
Ritz, 1999
Boice et al., 1999
Aerospace workers (Hughes)
TCE Subcohort
Low Intensity (<50 ppm)
High Intensity (>50 ppm)5
0.99 (0.64, 1.47) 4
0.73 (0.09, 2.64)) 2
0.44 (0.05, 1.58) 2
Morgan etal., 1998
Aircraft maintenance workers (Hill AFB, Utah)
Blair etal., 1998
6/22/2009
910
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
TCE subcohort
Males, Cumulative exp
Females, Cumulative exp
TCE subcohort
Males, Cumulative exp
Females, Cumulative exp
0.8 (0.2,2.2)'
11
0
1.01
< 5 ppm-yr
0.7 (0.7, 3,3)
3
5-25 ppm-yr
2.0 (0.5, 8.4)
5
>25 ppm-yr
0.9 (0.2, 4.4)
2
0
1.01
< 5 ppm-yr
0
5-25 ppm-yr
0
>25 ppm-yr
0
1.02 (0.39, 2.67)
17
1.26 (0.43, 3.75)
17
0
1.01
< 5 ppm-yr
1.46 (0.44, 4.86)
8
5-25 ppm-yr
1.74 (0.49,6.16)
6
>25 ppm-yr
0.66 (0.15,2.95)
3
0
0
< 5 ppm-yr
5-25 ppm-yr
>25 ppm-yr
Cardboard manufacturing workers in Arnsburg, Germany
TCE exposed workers
Unexposed workers
Deaths reported to GE pension fund (Pittsfield, MA)
Cardboard manufacturing workers, Atlanta area, GA
Aircraft manufacturing plant employees (Italy)
All subjects
3.70 (0.09, 20.64)
9.38 (1.93,27.27)
0.93 (0.32, 2.69)5
Not reported
0.79(0.16,2.31)
Aircraft manufacturing plant employees (San Diego, CA)
All subjects 0.78 (0.42,1.34)
Case-control Studies
Children's Cancer Group/Pediatric Oncology Group
Any TCE exposure
Neuroblastoma, <15 years age
1.64 (0.95, 2.84)
1
3
16
16
37
Radican etal., 2008
Henschler et al., 1995
Greenland et al., 1994
Sinks et al., 1992
Costa etal., 1989
Garabrant et al., 1988
De Roos et al., 2001
6/22/2009
911
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Paternal TCE exposure
Self-reported exposure 1.4(0.7,2.9) 22
IH assignment of probable exposure 0.9 (0.3, 2.5) 9
Population of So. LA, NJ, Philadelphia PA
Heineman et al., 1994
Any TCE exposure
1.1 (0.8, 1.6))
128
Low exposure
1.1 (0.7, 1.7))
27
Medium exposure
1.1 (0.6, 1.8)
42
High exposure
1.1 (0.5,2.8)
12
p for trend
0.45
Geographic Based Studies
Residents in two study areas in Endicott, NY
Brain/CNS, <19 years of age
Not reported
<6
Residents of 13 census tracts in Redlands, CA
Brain/CNS, <15 years of age
1.05 (0.24, 2.70)6
6
Resident of Tucson Airport Area, AZ
Brain/CNS, <19 years of age
1970-1986
0.84 (0.23,2.16)
3
1987-1991
0.78 (0.26, 2.39)
2
ATSDR, 2006
Morgan and Cassidy, 2002
ADHS, 1990, 1995
1 Internal referents, workers not exposed to TCE
2 Relative risks for TCE exposure after adjustment for 1st employment, socioeconomic status, and age at event.
3 Standardized incidence ratio from analyses lagging exposure 10 years prior to end of follow-up or date of incident
cancer.
4 Relative risks for TCE exposure after adjustment for time since 1st hired, external and internal radiation dose, and
same chemical at a different level.
5 Odds ratio from nested case-control analysis
6 99% Confidence Interval
6/22/2009
912
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
References
ATSDR (Agency for Toxic Substance and Disease Registry). 2004. Feasibility Investigation of
Worker Exposure to Trichloroethylene at the View-Master Factory in Beaverton, Oregon.
Prepared by the Oregon Department of Human Services. Report No. PB2005-102173.
ATSDR(Agency for Toxic Substances and Disease Registry). 2006. Health Consultation.
Cancer and Birth Outcome Analysis, Endicott Area, Town of Union, Broome County, New
York. Health Statistics Review. Atlanta, GA: U.S. Department of Health and Human Services,
Public Health Service, Agency for Toxic Substances and Disease Registry. May 26, 2006.
Anttila A, Pukkala E, Sallmen M, Hernberg S, Hemminki K. 1995. Cancer incidence among
Finnish workers exposed to halogenated hydrocarbons. J Occup Environ Med 37:797-806.
Axelson O, Selden A, Andersson K, Hogstedt C. 1994. Updated and expanded Swedish cohort
study on trichloroethylene and cancer risk. J Occup Environ 36:556-562.
Bakke B, Stewart PA, Waters MA. 2007. uses of and exposure to trichloroethylene in U.S.
Industry: a systematic literature review. J Occup Environ Hyg 4:375-390.
Blair A, Hartge P, Stewart PA, McAdams M, Lubin J. 1998. Mortality and cancer incidence of
aircraft maintenance workers exposed to trichloroethylene and other organic solvents and
chemicals: extended follow up. Occup Environ Med 55:161-171.
Boice JD, Marano DE, Fryzek JP, Sadler CJ, McLaughlin JK. 1999. Mortality among aircraft
manufacturing workers. Occup Environ Med 56:581-597.
Boice JD, Marano DE, Cohen SS, Mumma MR, Blott WJ, Brill AB, Fryzek JP, Henderson BE,
McLaughlin JK. 2006. Mortality among Rocketdyne worker who tested rocket engines, 1948-
1999. J Occup Environ Med 48:1070-1092.
Chang Y-M, Tai C-F, Yang S-C, Chan C-J, S Shin T-S, Lin RS, Liou S-H. 2003. A cohort
mortality study of workers exposed to chlorinated organic solvents in Taiwan. Ann Epidemiol
13:652-660.
6/22/2009
913
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Chang Y-M, Tai C-F, Yang S-C, Lin R, Sung F-C, Shin T-S, Liou S-H. 2005. Cancer Incidence
among Workers Potentially Exposed to Chlorinated Solvents in An Electronics Factory. J Occup
Health 47:171-180.
Costa G, Merletti F, Segnan N. 1989. A mortality cohort study in a north Italian aircraft factory.
Br JInd Med 46:738-743
Clapp RW, Hoffman K. 2008. Cancer mortality in IBM Endicott IBM workers, 1969-2001; an
update on a NY production plant. Environ Health 7:13.
Engel LS, Vaughan TL, Gammon MD, Chow W-H, Risch HA, Dubrow R, Mayne ST,
Rotterdam H, Schoenberg JE, Stanford JL, West AB, Blot JW, Fraumeni JF. 2002.
Occupational risk of esophageal and gastric cardia adenocarcinoma. Am J Ind Med 42:11-22.
Environmental Health Strategies. 1997. Final Report. Correspondence from Paul A. Cammer,
Ph.D., Trichloroethylene Issues Group, to Cheryl Siegel Scott, US Environmental Protection
Agency. December 22, 1997.
Garabrant DH, Held J, Langholz B, Bernstein L. 1988. Mortality of Aircraft Manufacturing
Workers in Southern California. Am J Ind Med 13:683-693.
Greenland S, Salvan A, Wegman DH, Hallock MF, Smith TH. 1994. A case-control study of
cancer mortality at the transformer-assembly facility. Int Arch Occup Environ Health 66:49-54.
Gustavsson P, Jakobsson R, Johansson H, Lewin F, Norell S, Rutkvist L-E. Occupational
exposure and squamous cell carcinoma of the oral cavity, pharynx, larynx, and oesophagus: a
case-control study in Sweden. Occup Environ Med 55:393-400.
Hansen J, Raaschou-Nielsen O, Christensen JM, Johansen I, McLaughlin JK, Lipworth L, Blot
WJ, Olsen JH. 2001. Cancer incidence among Danish workers exposed to trichloroethylene. J
Occup Environ Med 43:133-139.
Heineman EF, Cocco P, Gomez MR, Dosemeci M, Stewart PA, Hayes RB, Zahm SH, Thomas
TL, Blair A. 1994. Occupational exposure to chlorinated aliphatic hydrocarbons and risk of
astrocytic brain cancer. Am J Ind Med 26:155-169.
6/22/2009
914
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Henschler D, Vamvakas S, Lammert M, Dekant W, Kraus B, Thomas B, Ulm K. 1995.
Increased incidence of renal cell tumors in a cohort of cardboard workers exposed to
trichloroethene. Arch Toxicol 69:291-299.
Jansson C, Johansson AL, Bergdahl IA, Dickman PW, Plato N, Adami J, Boffeta P, Lagergren J.
2006a. Occupational exposures and risk of esophageal and gastric cardia cancers among male
Swedish construction workers. Cancer Causes Control 16:755-764.
Jansson C, Plate N, Johansson AL, Nyren O, Lagergren J. 2006b. Airborne occupational
exposures and risk of oesophageal and cardia adenocarcinoma. Occup Environ Med
63:107-112.
MallinK. 1990. Investigation of a bladder cancer cluster in Northwestern Illinois. Am J
Epidemiol 132 Suppl S96-S106.
Morgan JW, Cassidy RE. 2002. Community cancer assessment in response to long-time
exposure to perchlorate and trichloroethylene in drinking water. J Occup Environ Med.
44:616-621.
Morgan RW, Kelsh MA, Zhao K, Heringer S. 1998. Mortality of aerospace workers exposed to
trichloroethylene. Epidemiology 9:424-431.
Parent M-E, Siemiatycki J, Fritschi L. 2000. Workplace exposure and oesophageal cancer.
Occup Environ Med 57:325-324.
Peters JM, Preston- Martin S, Yu MC. 1981. Brain tumors in children and occupational
exposure of parents. 213:235-237.
Raaschou-Nielsen O, Hansen J, McLaughlin JK, Kolstad H, Christensen JM, Tarone RE, Olsen
JH. 2003. Cancer risk among workers at Danish companies using trichloroethylene: a cohort
study. Am J Epidemiol 158:1182-1192.
Radican L, Blair A, Stewart P, Wartenberg D. 2008. Mortality of aircraft maintenance workers
exposed to trichloroethylene and other hydrocarbons and chemicals: extended follow-up. J
Occup Environ Med 50:1306-1319.
6/22/2009
915
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Ramanakumar A, Nadon L, Siemiatycki J. 2008. Exposure in painting related occupations and
risk of selected cancers: results from a case-control study in Montreal. Am J Ind Med
51:419-427.
Ries LAG, Melbert D, Krapcho M, Stinchcomb DG, Howlader N, Horner MJ, Mariotto A,
Miller BA, Geuer EJ, Altekruse SF, Lewis DR, Clegg L, Reichman M, Edwards BK (eds). Seer
Cancer Statistics Review, 1975-2005. National Cancer Institute, Bethesda, MD,
http://seer.cancer.gov/csr/1975_2005/, based on November 2007 SEER data submission, posted
to the SEER web site, 2008. Accessed September 16, 2008.
Ritz B. 1999. Cancer mortality among workers exposed to chemicals during uranium
processing. J Occup Environ Med. 41:556-566.
Santibanez M, Vioque J, Alguiacil J, Barber X, Hera MG, Kauppinen T for the PANESOES
Study Group. 2008 [In Press], Occupational exposures and risk of oesophageal cancer by
histological type: A case control study in eastern Spain. Occup Environ Med. DOI:
10.1135/oem.2007.037929.
Siemiatycki J. 1991. Risk Factors for Cancer in the Workplace. Boca Raton: CRC Press.
Siemiatycki J, Dewar R, Nadon L, Gerin M. 1994. Occupational risk factors for bladder cancer:
results from a case-control study in Montreal, Quebec, Canada. Am J Epidemiol
140:1061-1080.
Sinks T, Lushniak B, Haussler BJ, Sniezek J, Deng JF, Roper P, et al. 1992. Renal cell cancer
among paperboard printing workers. Epidemiology 3:483-489.
Sung T-I, Chen P C, Lee L J-H, Lin Y-P, Hsieh G-Y, Wang J-D. 2007. Increased standardized
incidence ratio of breast cancer in female electronics workers. BMC Public Health 7:102-lxx.
Accessed July 9, 2007 on Bio Med Central, http://vvvvvv.biomedcentral.com/content/pdf/1471 -
2458-7-102.pdf.
Vartiainen T, Pukkala E, Rienoja T, Strandman T, Kaksonen K. 1993. Population exposure to
tri-and tetrachloroethene and cancer risk: two cases of drinking water pollution. Chemosphere
27:1171-1181.
6/22/2009
916
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Ward EM, Thun MJ, Hannan LM, Jemal A. 2006. Interpreting Cancer Trends. Ann Ny Acad
Sci 1076:29-53.
Weiderpass E, Pukkala E, Kauppinen T, Mutanen P, Paakkulainen H, Vasama-Neuvonen K,
Boffetta P, Partanen T. 1999. Breast cancer and occupational exposure in women in Finland.
Am JInd Med 36:48-53.
Weiderpass E, Vainion H, Kauppinen T, Vasama-Neuvonen K, Partanen T, Pukkala E. 2003.
Occupational exposure and gastrointestinal cancers among Finnish women. J Occup Environ
Med 45:305-315.
Wilcosky TC, Checkoway H, Marshall EG, Tyroler HA. 1984. Cancer mortality and solvent
exposures in the rubber industry. Am Ind Hyg Assoc J 45:809-811.
Yu MC, Garabrant DH, Peters JM, Mack T. 1988. Tobacco, alcohol, diet, occupation, and
carcinoma of the esophagus. Cancer Res 48:3843-3848.
Zhao Y, Krishnadasan A, Kennedy N, Morgenstern H, Ritz B. 2005. Estimated effects of
solvents and mineral oils on cancer incidence and Mortality in a cohort of aerospace workers.
Am J Ind Med 48:249-258.
6/22/2009
917
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.9 Susceptible Lifestages and Populations
Variation in response among segments of the population may be due to age, genetics, and
ethnicity, as well as to differences in lifestyle, nutrition, and disease status. These could be
potential risk factors that play an important role in determining an individual's susceptibility and
sensitivity to chemical exposures. Studies on TCE toxicity in relation to some of these risk
factors including lifestage, gender, genetics, race/ethnicity, pre-existing health status, and
lifestyle are discussed below. Others have also reviewed factors related to human variability and
their potential for susceptibility to TCE (Barton et al., 1996; Clewell et al., 2000; Davidson and
Beliles, 1991; NRC, 2006; Pastino et al., 2000).
4.9.1 Lifestages
Individuals of different lifestages are physiologically, anatomically, and biochemically
different. Early and later lifestages differ greatly from adulthood in body composition, organ
function, and many other physiological parameters that can influence the toxicokinetics of
chemicals and their metabolites in the body (ILSI, 1992). The limited data on TCE exposure
suggest that these segments of the population - particularly individuals in early lifestages - may
have greater susceptibility than does the general population. This section presents and evaluates
the pertinent published literature available to assess how individuals of differing lifestages may
respond differently to TCE.
4.9.1.1 Early Lifestages
4.9.1.1.1 Early Lifestage-Specific Exposures
Section 2.4 describes the various exposure pathways of concern for TCE. For all
postnatal lifestages, the primary exposure routes of concern include inhalation and contaminated
drinking water. In addition, there are exposure pathways to TCE are unique to early lifestages.
Fetal and infant exposure to TCE can occur through placental transfer and breast milk
consumption if the mother has been exposed, and could potentially increase overall TCE
exposure. Placental transfer of TCE has been demonstrated in humans (Beppu, 1968; Laham,
1970), rats (Withey and Karpinski, 1985), mice (Ghantous et al., 1986), rabbits (Beppu, 1968),
and sheep and goats (Helliwell and Hutton, 1950). Similarly, TCE has been found in breast milk
in humans (Fisher et al., 1997; Pellizzari et al., 1982), goats (Hamada and Tanaka, 1995), and
rats (Fisher et al., 1990). Pellizzari et al. (1982) conducted a survey of environmental
contaminants in human milk, using samples from cities in the northeastern region of the U.S. and
6/22/2009
918
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
one in the southern region and detected TCE in 8 milk samples taken from 42 lactating women.
No details of times post-partum, milk lipid content or TCE concentration in milk or blood were
reported. Fisher et al. (1997) predicted that a nursing infant would consume 0.496 mg TCE
during a 24-hour period. In lactating rats exposed to 600 ppm (3,225 mg/m3) TCE for 4 hours
resulted in concentrations of TCE in milk of 110 |ag/m L immediately following the cessation of
exposure (Fisher et al., 1990).
Direct childhood exposures to TCE from oral exposures may also occur. A
contamination of infant formula resulted in levels of 13 ppb (Fan, 1988). Children consume high
levels of dairy products, and TCE has been found in butter and cheese (Wu and Schaum, 2000).
In addition, TCE has been found in food and beverages containing fats such as margarine
(Wallace et al., 1984), grains and peanut butter (Wu and Schaum, 2000), all of which children
consume in high amounts. A number of studies have examined the potential adverse effects of
childhood exposure to drinking water contaminated with TCE (ATSDR, 1998, 2001; Bernad et
al., 1987; Bove, 1996; Bove et al., 1995; Burg and Gist, 1999; Goldberg et al., 1990; Lagakos et
al., 1986; Rodenbeck et al., 2000; Sonnenfeld et al., 2001; White et al., 1997; see Section
4.9.2.1). TCE in residential water may also be a source of dermal or inhalation exposure during
bathing and showering (Fan, 1988; Franco et al., 2007; Giardino and Andelman, 1996; Lee et al.,
2002; Weisel and Jo, 1996; Wu and Schaum, 2000); it has been estimated that showering and
bathing scenarios in water containing 3 ppm TCE, a child of 22 kg receives a higher dose (about
1.5 times) on a mg/kg basis than a 70 kg adult (Fan, 1988).
Direct childhood inhalation exposure to TCE have been documented in both urban and
rural settings. A study of VOCs measured personal, indoor and outdoor TCE in 284 homes, with
72 children providing personal measures and time-activity diaries (Adgate et al., 2004a). The
intensive-phase of the study found a mean personal level of 0.8 |ig/m3 and mean indoor and
outdoor levels of 0.6 |ig/m3, with urban homes have significantly higher indoor levels of TCE
than non-urban homes (t = 2.3,/? = 0.024) (Adgate et al., 2004a). A similar study of personal,
indoor and outdoor TCE was conducted in two inner-city elementary schools as well as in the
homes of 113 children along with time-activity diaries, and found a median a median personal
level of 0.3 |ig/m3, a median school indoor level of 0.2 |ig/m3, a median home indoor level of 0.3
|ig/m3, a median outdoor level of 0.3 |ig/m3 in the winter, with slightly lower levels in the spring
(Adgate et al., 2004b). Studies from Leipzig, Germany measured the median air level of TCE in
children's bedrooms to be 0.42 |ig/m3 (Lehmann et al., 2001) and 0.6 |ig/m3 (Lehmann et al.,
2002). A study of VOCs in Hong Kong measured air levels in schools, including an 8-hour
average of 1.28 |ig/m3, which was associated with the lowest risk of cancer in the study (Guo et
al., 2004). Another found air TCE levels to be highest in school/work settings, followed by
outside, in home, in other, and in transit settings (Sexton et al., 2007). Children exposed to soil
6/22/2009
919
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
vapor levels ranged from 0.18-140 mg/m3 in indoor air (ATSDR, 2006). Contaminated soil may
be a source of either dermal or ingestion exposure of TCE for children (Wu and Schaum, 2000).
Additional TCE exposure has also been documented to have occurred during medical
procedures. TCE was used in the past as an anesthetic during childbirth (Beppu, 1968; Phillips
and Macdonald, 1971) and surgery during childhood (Jasinka, 1965). These studies are
discussed in more detail in Chapter 4.7.3.1.1. In addition, the TCE metabolite chloral hydrate
has been used as an anesthetic for children for CAT scans (Steinberg, 1993).
Dose received per body weight for 3 ppm TCE via oral, dermal, dermal plus inhalation,
and bathing scenarios was estimated for a 10-kg infant, a 22-kg child, and a 70-kg adult (Fan,
1988; see Table 4.9.1). For the oral route (drinking water), an infant would receive a higher
daily dose than a child, and the child more than the adult. For the dermal and dermal plus
inhalation route, the child would receive more than the adult. For the bathing scenario, the infant
and child would receive comparable amounts, more than the adult.
Table 4.9.1. Estimated lifestage-specific daily doses for TCE in water*
Body Weight
Infant (10 kg)
Child (22 kg)
Adult (70 kg)
Drinking water
0.3 mg/kg
0.204 mg/kg
0.086 mg/kg
Showering - dermal
-
0.1 mg/kg
0.064 mg/kg
Showering - dermal
and inhalation
-
0.129 mg/kg
0.083 mg/kg
Bathing - 15 min
-
0.24 mg/kg
0.154 mg/kg
Bathing - 5 min
0.08 mg/kg
0.08 mg/kg
0.051 mg/kg
*Adapted from Fan (1988).
4.9.1.1.2 Early Lifestage-Specific Toxicokinetics
Chapter 3 describes the toxicokinetics of TCE. However, toxicokinetics in
developmental lifestages are distinct from toxicokinetics in adults (Benedetti et al., 2007;
Ginsberg et al., 2002, 2004a, 2004b; Hattis et al., 2003) due to, for example, altered ventilation
rates, percent adipose tissue, and metabolic enzyme expression. Early lifestage-specific
information is described below for absorption, distribution, metabolism, and excretion, followed
by available early lifestage-specific PBPK models.
Absorption. As discussed in Section 3.1, exposure to TCE may occur via inhalation,
ingestion, and dermal absorption. In addition, prenatal exposure may result in absorption via the
transplacental route. Exposure via inhalation is proportional to the ventilation rate, duration of
exposure, and concentration of expired air, and children have increased ventilation rates per kg
body weight compared to adults, with an increased alveolar surface area per kg body weight for
6/22/2009
920
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
the first two years (NRC, 1993). It is not clear to what extent dermal absorption may be different
for children compared to adults; however, infants have a 2-fold increase in surface area
compared to adults, although similar permeability (except for premature babies) compared to
adults (NRC, 1993).
Distribution. Both human and animal studies provide clear evidence that TCE distributes
widely to all tissues of the body (see Section 3.2). For lipophilic compounds such as TCE,
percentage adipose tissue, which varies with age, will affect absorption and retention of the
absorbed dose. Infants have a lower percentage of adipose tissue per body weight than adults,
resulting in a higher concentration of the lipophilic compound in the fat of the child (NRC,
1993).
During pregnancy of humans and experimental animals, TCE is distributed to the
placenta (Beppu, 1968; Ghantous et al., 1986; Helliwell and Hutton, 1950; Laham, 1970; Withey
and Karpinski, 1985). In humans, TCE has been found in newborn blood after exposure to TCE
during childbirth with ratios of concentrations in fetal:maternal blood ranging from
approximately 0.5 to approximately 2 (Laham, 1970). In childhood, blood levels concentrations
of TCE were found to range from 0.01-0.02 ng/mL (Sexton et al., 2005). Pregnant rats exposed
to TCE vapors on GD 17 resulted in concentrations of TCE in fetal blood approximately one-
third the concentration in corresponding maternal blood, and was altered based upon the position
along the uterine horn (Withey and Karpinski, 1985). TCE has also been found in the organs of
prenatal rabbits including the brain, liver, kidneys and heart (Beppu, 1968). Rats prenatally
exposed to TCE had increased levels measured in the brain at PND10, compared to rats exposed
as adults (Rodriguez et al., 2007). TCE can cross the blood-brain barrier during both prenatal
and postnatal development, and may occur to a greater extent in younger children. It is also
important to note that it has been observed in mice that TCE can cycle from the fetus into the
amniotic fluid and back to the fetus (Ghantous et al., 1986).
Studies have examined the differential distribution by age to a mixture of six VOCs
including TCE to children aged 3-10 years and adults aged 20-82 years old (Mahle et al., 2007)
and in rats at PND10, 2 months (adult), and 2 years (aged) (Mahle et al., 2007; Rodriguez et al.,
2007). In humans, the blood:air partition coefficient for male or female children was
significantly lower compared to adult males (Mahle et al., 2007). In rats, the difference in
tissue:air partition coefficients increased with age (Mahle et al., 2007). Higher peak
concentrations of TCE in the blood were observed in the PND10 rat compared to the adult rat
after inhalation exposure, likely due to the lower metabolic capacity of the young rats (Rodriguez
et al., 2007).
Metabolism. Section 3.3 describes the enzymes involved in the metabolism of TCE,
including CYP and GST. Expression of these enzymes changes during various stages of fetal
6/22/2009
921
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
development (Dome et al., 2005; Hakkola et al., 1996a, b, 1998a, b; Hines and McCarver, 2002;
Shao et al., 2007; van Lieshout et al., 1998) and during postnatal development (Blake et al.,
2005; Dome et al., 2005; Tateishi et al., 1997), and may result in altered susceptibility.
Expression of CYP enzymes have been shown to play a role in decreasing the
metabolism of TCE during pregnancy in rats, although metabolism increased in young rats (3-
week-old) compared to adult rats (18-week-old) (Nakajima et al., 1992a). For TCE, CYP2E1 is
the main metabolic CYP enzyme, and expression of this enzyme has been observed in humans in
prenatal brain tissue at low levels beginning at 8 weeks gestation and increasing throughout
gestation (Brzezinski et al., 1999). Very low levels of CYP2E1 have been detected in some
samples fetal liver during the second trimester (37% of samples) and third trimester (80% of
samples) (Carpenter et al., 1996; Johnsrud et al., 2003), although hepatic expression surges
immediately after birth in most cases (Johnsrud et al., 2003; Vieira et al., 1996) and in most
infants reaches adult values by 3 months of age (Johnsrud et al., 2003; Vieira et al., 1996).
Although there is some uncertainty as to which GST isoforms mediate TCE conjugation,
it should be noted that their expression changes with fetal development (McCarver and Hines,
2002; Raijmakers et al., 2001; van Lieshout et al., 1998).
Excretion. The major processes of excretion of TCE and its metabolites are discussed in
Section 3.4, yet little is know about whether there are age-related differences in excretion of
TCE. The major pathway for elimination of TCE is via exhalation, and its metabolites via urine
and feces, and it is known that renal processes are not mature until about 6 months of age (NRC,
1993). Only one study was identified that measured TCE or its metabolites in exhaled breath
and urine in a 17-year old who ingested a large quantity of TCE (Briining et al., 1998). TCE has
also been measured in the breast milk in lactating women (Fisher et al., 1997; Pellizzari et al.,
1982), goats (Hamada and Tanaka, 1995), and rats (Fisher et al., 1990).
PBPK Models. Early lifestage-specific information regarding absorption, distribution,
metabolism, and excretion needs to be considered for a child-specific and chemical-specific
PBPK model. To adequately address the risk to infants and children, age-specific parameters for
these values should be used in PBPK models that can approximate the internal dose an infant or
child receives based on a specific exposure level (see Section 3.5).
Fisher et al. developed PBPK models to describe the toxicokinetics of TCE in the
pregnant rat (Fisher et al., 1989), lactating rat and nursing pup (Fisher et al., 1990). The prenatal
study demonstrates that approximately two-thirds of maternal exposure to both TCE and TCA
reached the fetus after maternal inhalation, gavage, or drinking water exposure (Fisher et al.,
1989). After birth, only 2% of maternal exposure to TCE reaches the pup; however, 15% and
30%) of maternal TCA reaches the pup after maternal inhalation and drinking water exposure,
respectively (Fisher et al., 1990). One analysis of PBPK models examined the variability in
6/22/2009
922
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
response to VOCs including TCE between adults and children, and concluded that the
intraspecies UF for PK is sufficient to capture variability between adults and children (Pelekis et
al., 2001).
4.9.1.1.3 Early Lifestage-Specific Effects
Although limited data exist on TCE toxicity as it relates to early lifestages, there is
enough information to discuss the qualitative differences. In addition to the evidence described
below, Section 4.7 contains information reproductive and developmental toxicity. In addition,
Sections 4.2 on neurotoxicity and 4.5 on immunotoxicity characterize a wide array of postnatal
developmental effects.
4.9.1.1.3.1 Differential effects in early lifestages.
There are a few adverse health outcomes, in particular birth defects, which are observed
only after early lifestage exposure to TCE.
Birth Defects. A summary of structural developmental outcomes that have been
associated with TCE exposures is presented in Sections 4.7.2.3. In particular, cardiac birth
defects have been observed after exposure to TCE in humans (ATSDR, 2006; Goldberg et al.,
1990; Lagakos et al., 1986; Yauck et al., 2004), rodents (Dawson et al., 1990, 1993; Johnson et
al., 1998a, b, 2003, 2005; Smith et al., 1989, 1992), and chicks (Bross et al., 1983; Loeber et al.,
1988; Boyer et al., 2000; Drake et al., 2006a, b; Mishima et al., 2006; Rufer et al., 2008).
However it is notable that cardiac malformations were not observed in a number of other studies
in humans (Lagakos et al., 1986; Taskinen et al., 1989; Tola et al., 1980), rodents (Carney et al.,
2006; Coberly et al., 1992; Cosby and Dukelow, 1992; Dorfmueller et al., 1979; Fisher et al.,
2001; Hardin et al., 1981; Healy et al., 1982; Narotsky and Kavlock, 1995; Narotsky et al., 1995;
Schwetz et al., 1975), and rabbits (Hardin et al., 1981). See Section 4.7.2.3.2 for further
discussion on cardiac malformations.
Structural CNS birth defects were observed in humans (ATSDR, 2001; Bove, 1996; Bove
et al., 1995; Lagakos et al., 1986). In addition, a number of postnatal non-structural adverse
effects have been observed in humans and experimental animals following prenatal exposure to
TCE. See Sections 4.2.10 and 4.7.2.3.3 for further discussion on developmental neurotoxicity.
A variety of other birth defects have been observed—including eye/ear birth anomalies in
humans and rats (Lagakos et al., 1986; Narotsky et al., 1995; Narotsky and Kavlock, 1995);
lung/respiratory tract disorders in humans and mice (Das and Scott, 1994; Lagakos et al., 1986);
and oral cleft defects (Bove, 1996; Bove et al., 1995; Lagakos et al., 1986), kidney/urinary tract
disorders, musculoskeletal birth anomalies (Lagakos et al., 1986), and anemia/blood disorders
(Burg and Gist, 1999) in humans. See Section 4.7.2.3.5 for further discussion on other structural
6/22/2009
923
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
developmental outcomes. A current follow-up study of the Camp Lejeune cohort will examine
birth defects and may provide additional insight (ATSDR, 2003b; GAO, 2007a, b).
4.9.1.1.3.2 Susceptibility to noncancer outcomes in early lifestages.
There are a number of adverse health outcomes observed after exposure to TCE that are
observed in both children and adults. Below is a discussion of differential exposure, incidence
and/or severity in early lifestages compared to adulthood.
Occupational TCE poisonings via inhalation exposure resulted in an elevated percent of
cases in the adolescents aged 15-19 years old (McCarthy and Jones, 1983). In addition, there is
concern for intentional exposure to TCE during adolescence, including a series of deaths
involving inhaling typewriter correction fluid (King et al., 1985), a case of glue sniffing likely
associated with cerebral infarction in a 12-year-old boy with a 2-year history of exposure (Parker
et al., 1984), and a case of attempted suicide by ingestion of 70 mg TCE in a 17-year-old boy
(Briining et al., 1998).
Neurotoxicity. Adverse CNS effects observed after early lifestage exposure to TCE in
humans include delayed newborn reflexes (Beppu, 1968), impaired learning or memory (Bernad
et al., 1987, abstract; White et al., 1997); aggressive behavior (Bernad et al., 1987; Blossom et
al., 2008); hearing impairment (Burg and Gist, 1999); speech impairment (Burg and Gist, 1995;
White et al., 1997); encephalopathy (White et al., 1997); impaired executive and motor function
(White et al., 1997); attention deficit (Bernad et al., 1987; White et al., 1997), and autism
spectrum disorder (Windham et al., 2006). One analysis observed a trend for increased adversity
during development, with those exposed during childhood demonstrating more deficits than
those exposed during adulthood (White et al., 1997). In experimental animals, observations
include decreased specific gravity of newborn brains until weaning (Westergren et al., 1984),
reductions in myelination in the brains at weaning, significantly decreased uptake of 2-
deoxyglucose in the neonatal rat brain, significant increase in exploratory behavior (Isaacson and
Taylor, 1989; Noland-Gerbec et al., 1986; Taylor et al., 1985), decreased rearing activity
(Fredriksson et al., 1993), and increased time to cross the first grid in open field testing (George
et al., 1986).
Two studies addressed whether or not children are more susceptible to CNS effects (Burg
et al., 1995; White et al., 1997). An analysis of three residential exposures of TCE observed
speech impairments in younger children and not at any other lifestage (White et al., 1997). A
national exposure registry also observed statistically significant speech impairment and hearing
impairment in 0-9 year olds and no other age group (Burg et al., 1995). However, a follow-up
study did not find a continued association with speech and hearing impairment in these children,
6/22/2009
924
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
although the absence of acoustic reflexes remained significant (ATSDR, 2003a). See Section 4.2
for further information on central nervous system toxicity, and Section 4.7.3.3.3 for further
information on developmental neurotoxicity.
Liver Toxicity. No early lifestage-specific effects were observed after TCE exposure.
See Section 4.3 for further information on liver toxicity.
Kidney Toxicity. Residents of Woburn, Massachusetts including 4,978 children were
surveyed on residential and medical history to examine an association with contaminated wells;
an association was observed for higher cumulative exposure measure and history of kidney and
urinary tract disorders (primarily kidney or urinary tract infections) and with lung and respiratory
disorders (asthma, chronic bronchitis, or pneumonia) (Lagakos et al., 1986). See Section 4.4 for
further information on kidney toxicity.
Immunotoxicitv. Several studies in exposure to TCE in early lifestages of humans and
experimental animals were identified that assessed the potential for developmental
immunotoxicity (Adams et al., 2003; Blossom and Doss, 2007; Blossom et al., 2008; Lehmann et
al., 2001, 2002; Peden-Adams et al., 2006, 2008). All noted evidence of immune system
perturbation except one (Lehman et al., 2001). See Section 4.5 for further information on
immunotoxicity, and Section 4.7.2.3.4 for further discussion on developmental immunotoxicity.
Respiratory Toxicity. Residents of Woburn, Massachusetts including 4,978 children
were surveyed on residential and medical history to examine an association with contaminated
wells; an association was observed for lung and respiratory disorders (asthma, chronic bronchitis,
or pneumonia) (Lagakos et al., 1986). See Section 4.6 for further information on respiratory
tract toxicity.
4.9.1.1.3.3 Susceptibility to cancer outcomes in early lifestages.
The epidemiologic and experimental animal evidence is limited regarding susceptibility
to cancer from exposure to TCE during early life stages. The human epidemiological evidence is
summarized above for cancer diagnosed during childhood (see Sections 4.7.2.1 and 4.7.2.3.5),
including a discussion of childhood cancers of the nervous system including neuroblastoma and
the immune system including leukemia (see Section 4.5.1.3). A current follow-up study of the
Camp Lejeune cohort will examine childhood cancers and may provide additional insight
(ATSDR, 2003b; GAO, 2007a, b). No studies of cancers in experimental animals in early
lifestages have been observed.
Total Childhood Cancer. Total childhood cancers have been examined in relationship to
TCE exposure (ATSDR, 2006; Morgan and Cassady, 2002). Two studies examining total
childhood cancer in relation to TCE in drinking water did not observe an association. A study in
6/22/2009
925
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Endicott, NY contaminated by a number of VOCs, including "thousands of gallons" of TCE
observed fewer than 6 cases of cancer diagnosed between 1980 and 2001 in children aged 0-19
years, and did not exceed expected cases or types (ATSDR, 2006). A California community
exposed to TCE in drinking water from contaminated wells was examined for cancer, with a
specific emphasis on childhood cancer (<15 years old); however, the incidence did not exceed
those expected for the community (Morgan and Cassady, 2002). A third study of childhood
cancer in relation to TCE in drinking water in Camp Lejeune, North Carolina is currently
underway (GAO, 2007a, b).
Childhood Leukemia. Childhood leukemia has been examined in relationship to TCE
exposure (Cohn et al., 1994; Lagakos et al., 1986; Lowengart et al., 1987; McKinney et al., 1991;
Costas et al., 2002; Shu et al., 1999). In a study examining drinking water exposure to TCE in
75 New Jersey towns, childhood leukemia, (including ALL) was significantly increased for girls
(n = 6) diagnosed before age 20 years, but this was not observed for boys (Cohn et al., 1994). A
community in Woburn, MA with contaminated well water including TCE experienced 20 cases
of childhood leukemia, significantly more than expected (Lagakos et al., 1986). Further analysis
by Costas et al. (2002) also observed a greater than 2-fold increase over expected cases of
childhood leukemia. Cases were more likely to be male (76%), <9 years old at diagnosis (62%),
breast-fed (OR: 10.17, 95% CI: 1.22-84.50), and exposed during pregnancy (adjusted OR: 8.33,
95%) CI: 0.73-94.67). The highest risk was observed for exposure during pregnancy compared
to preconception or postnatal exposure, and a dose-response was seen for exposure during
pregnancy (Costas et al., 2002). In addition, family members of those diagnosed with childhood
leukemia, including 13 siblings under age 19 at the time of exposure, had altered immune
response, but an analysis looking at only these children was not done (Byers et al., 1988).
Case-control studies examined children diagnosed with ALL for parental occupational
exposures and found a non-significant 2- to 4-fold increase of childhood leukemia risk for
exposure to TCE during preconception, pregnancy, postnatally, or all developmental periods
combined (Lowengart et al., 1987; McKinney et al., 1991; Shu et al., 1999). Some studies
showed an elevated risk for maternal (Shu et al., 1999) or paternal exposure (Lowengart et al.,
1987; McKinney et al., 1991), while others did not show an elevated risk for maternal
(McKinney et al., 1991) or paternal exposure (Shu et al., 1999), possibly due to the small number
of cases. No variability was observed in the developmental stages in Shu et al. (1999), although
Lowengart et al. (1987) observed the highest risk to be paternal exposure to TCE after birth.
CNS Tumors. In a case-control study of parental occupational exposures, paternal self-
reported exposure to TCE was not significantly associated with neuroblastoma in the offspring
(OR = 1.4, 95%>CI = 0.7-2.9) (De Roos et al., 2001). Brain tumors have also been observed in
the offspring of fathers exposed to TCE, but the odds ratio could not be determined (Peters et al.,
6/22/2009
926
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1981, 1985).
Age-Dependent Adjustment Factors (ADAFs). According to U.S. EPA's Supplemental
Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA,
2005b) there may be increased susceptibility to early-life exposures for carcinogens with a
mutagenic MO A. Therefore, because the weight of evidence supports a mutagenic MOA for
TCE carcinogenicity in the kidney (see Section 4.3.7), and in the absence of chemical-specific
data to evaluate differences in susceptibility, early-life susceptibility should be assumed and the
age-dependent adjustment factors (ADAFs) should be applied, in accordance with the
Supplemental Guidance.
4.9.1.2 Later Lifestages
Few studies examine the differential effects of TCE exposure for elderly adults (>65
years old). These limited studies suggest that older adults may experience increased adverse
effects than younger adults. However, there is no further evidence for elderly individuals
exposed to TCE beyond these studies.
Toxicokinetics in later lifestages are distinct from toxicokinetics in younger adults
(Benedetti et al., 2007; Ginsberg et al., 2005). Studies have examined the age differences in TK
after exposure to a mixture of six VOCs including TCE for humans (Mahle et al., 2007) and rats
(Mahle et al., 2007; Rodriguez et al., 2007). In humans, the blood:air partition coefficient for
adult males (20-82 years) was significantly (p < 0.05) higher (11.7 ± 1.9) compared to male
(11.2 ± 1.8) or female (11.0 ± 1.6) children (3-10 years) (Mahle et al., 2007); when the data was
stratified for adults above and below 55 years of age, there was no significant difference
observed between adults (20-55 years) and aged (56-82) (data not reported). In rats, the
difference in tissue:air partition coefficients also increased from PND10 to adult (2 months) to
aged (2 years) rat (Mahle et al., 2007). TCE has also been measured in the brain of rats, with an
increased level observed in older (2 year old) rats compared to adult (2 month old) rats
(Rodriguez et al., 2007). It was also observed that aged rats reached steady state slower with
higher concentrations compared to the adult rat; the authors suggest that the almost 2-fold greater
percentage of body fat in the elderly is responsible for this response (Rodriguez et al., 2007). An
age-related difference in CYP expression (Dome et al., 2005), in particular CYP2E1 activity
were observed in human liver (George et al., 1995), with the lowest activity in those >60 years
and the highest in those <20 years old (Parkinson et al., 2004). Also, GST expression has been
observed to decrease with age in human lymphocytes, with the lowest expression in those aged
60-80 years old (van Lieshout and Peters, 1998).
6/22/2009
927
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
One cohort of TCE exposed metal degreasers found an increase in psychoorganic
syndrome and increased vibration threshold related to increasing age (Rasmussen et al., 1993a, b,
c), although the age groups were <29 years, 30-39 years, and 40+ years, but the age ranged only
from 18-68 years and did not examine >65 years as a separate category.
4.9.2 Other Susceptibility Factors
Aside from age, many other factors may affect susceptibility to TCE toxicity. A partial
list of these factors includes gender, genetic polymorphisms, pre-existing disease status,
nutritional status, diet, and previous or concurrent exposures to other chemicals. The toxicity
that results due to changes in multiple factors may be quite variable, depending on the exposed
population and the type of exposure. Qualitatively, the presence of multiple susceptibility
factors will increase the variability that is seen in a population response to TCE toxicity.
4.9.2.1 Gender
Individuals of different genders are physiologically, anatomically, and biochemically
different. Males and females can differ greatly in many physiological parameters such as body
composition, organ function, and ventilation rate, which can influence the toxicokinetics of
chemicals and their metabolites in the body (Gandhi et al., 2004; Gochfeld, 2007).
4.9.2.1.1 Gender-Specific Toxicokinetics
Chapter 3 describes the toxicokinetics of TCE. Gender-specific information is described
below for absorption, distribution, metabolism, and excretion, followed by available gender-
specific PBPK models.
Absorption. As discussed in Section 3.1, exposure to TCE may occur via inhalation,
ingestion, and skin absorption. Exposure via inhalation is proportional to the ventilation rate,
duration of exposure, and concentration of expired air, and women have increased ventilation
rates during exercise compared to men (Gochfeld, 2007). Percent body fat varies with gender
(Gochfeld, 2007), which for lipophilic compounds such as TCE will affect absorption and
retention of the absorbed dose. After experimental exposure to TCE, women were found to
absorb a lower dose due to lower alveolar intake rates compared to men (Sato, 1993; Sato et al.,
1991b).
Distribution. Both human and animal studies provide clear evidence that TCE distributes
widely to all tissues of the body (see Section 3.2). The distribution of TCE to specific organs
6/22/2009
928
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
will depend on organ blood flow and the lipid and water content of the organ, which may vary
between genders (Gochfeld, 2007). After experimental exposure to humans, higher distribution
of TCE into fat tissue was observed in women leading to a greater blood concentration 16 hours
after exposure compared to men (Sato, 1993; Sato et al., 1991b). In experimental animals, male
rats generally have higher levels of TCE in tissues compared to female rats, likely due to gender
differences in metabolism (Lash et al., 2006). In addition, TCE has been observed in the male
reproductive organs (epididymis, vas deferens, testis, prostate, and seminal vesicle) (Zenick et
al., 1984).
Metabolism. Section 3.3 describes the metabolic processes involved in the metabolism of
TCE, including CYP and GST enzymes. In addition, the role of metabolism in male
reproductive toxicity is discussed in Section 4.7.1.3.2. In general, there is some indication that
TCE metabolism is different between males and females, with females more rapidly
metabolizing TCE after oral exposure to rats (Lash et al., 2006), intraperitoneal injections in rats
(Verma and Rana, 2003), and in mouse, rat and human liver microsomes (Elfarra et al., 1998).
CYP450 expression may differ between genders (Gandhi et al., 2004; Gochfeld, 2007;
Lash et al., 2006; Parkinson et al., 2004). CYP2E1 was detected in the epididymis and testes of
mice (Forkert et al., 2002), and CYP2E1 and GST-a has been detected in the ovaries of rats (Wu
and Berger, 2008), indicating that metabolism of TCE can occur in both the male and female
reproductive tracts. Unrelated to TCE exposure, there is no gender-related difference in
CYP2E1 activity observed in human liver microsomes (Parkinson et al., 2004). One study of
TCE exposure in mice observed induced CYP2E1 expression in the liver of males only
(Nakajima et al., 2000). Male rats have been shown to have higher levels of TCE metabolites in
the liver (Lash et al., 2006), and lower levels of TCE metabolites in the kidney (Lash et al.,
2006) compared to female rats. However, another study did not observe ant sex-related
differences in the metabolism of TCE in rats (Nakajima et al., 1992a).
Unlike P450-mediated oxidation, quantitative differences in the polymorphic distribution
or activity levels of GST isoforms in humans are not presently known. However, the available
data (Lash et al., 1999a, b) do suggest that significant variation in GST-mediated conjugation of
TCE exists in humans. One study observed that GSH conjugation is higher in male rats
compared to female rats (Lash et al., 2000); however it has also been speculated that any gender
difference may be due to a polymorphism in GSH conjugation of TCE rather than a true gender
difference (Lash et al., 1999a). Also, induction of PPARa expression in male mice was greater
than that in females (Nakajima et al., 2000).
Excretion. The major processes of excretion of TCE and its metabolites are discussed in
Section 3.4. Two human voluntary inhalation exposure studies observed the levels of TCE and
6/22/2009
929
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
its metabolites in exhaled breath and urine (Kimmerle and Eben, 1973; Nomiyama and
Nomiyama, 1971). Increased levels of TCE in exhaled breath in males were observed in one
human voluntary inhalation exposure study of 250-380 ppm for 160 minutes (Nomiyama and
Nomiyama, 1971), but no difference was observed in another study of 40 ppm for 4 hours or 50
ppm for 4 hours for 5 days (Kimmerle and Eben, 1973).
After experimental exposure to TCE, women were generally found to excrete higher
levels of TCE and TCA compared to men (Kimmerle and Eben, 1973; Nomiyama and
Nomiyama, 1971). However, other studies observed an increase in TCE in the urine of males
(Inoue et al., 1989), an increase in TCA in the urine of males (Sato et al., 1991b), or no
statistically significant (p > 0.10) gender difference for TCA in the urine (Inoue et al., 1989).
Others found that the urinary elimination half-life of TCE metabolites is longer in women
compared to men (Ikeda, 1977; Ikeda and Imamura, 1973).
In addition to excretion pathways that occur in both genders, excretion occurs uniquely in
men and women. In both humans and experimental animals, it has been observed that females
can excrete TCE and metabolites in breast milk (Fisher et al., 1990, 1997; Hamada and Tanaka,
1995; Pellizzari et al., 1982), while males can excrete TCE and metabolites in seminal fluid
(Forkert et al., 2003; Zenick et al., 1984).
PBPK Models. Gender-specific differences in uptake and metabolism of TCE were
incorporated into a PBPK model using human exposure data (Fisher et al., 1998). The chemical-
specific parameters included cardiac output at rest, ventilation rates, tissue volumes, blood flow,
and fat volume. This model found that gender differences for the toxicokinetics of TCE are
minor.
4.9.2.1.2 Gender-Specific Effects
4.9.2.1.2.1 Gender susceptibility to non-cancer outcomes
Liver Toxicity. No gender susceptibility to non-cancerous outcomes in the liver was
observed. A detailed discussion of the studies examining the effects of TCE on the liver can be
found in Section 4.3.
Kidney Toxicity. A detailed discussion of the studies examining the noncancer effects of
TCE on the kidney can be found in Section 4.4. A residential study found that females aged
55-64 years old had an elevated risk of kidney disease (RR = 4.57, 99% CI = 2.10-9.93),
although an elevated risk of urinary tract disorders was reported for both males and females
(Burg et al., 1995). Additionally, a higher rate of diabetes in females exposed to TCE was
6/22/2009
930
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
reported in two studies (Burg et al., 1995; Davis et al., 2005). In rodents, however, and kidney
weights were increased more in male mice than in females (Kjellstrand et al., 1983a, b), and
male rats have exhibited increased renal toxicity to TCE (Lash et al., 1998, 2001).
Immunotoxicitv. A detailed discussion of the studies examining the immunotoxic effects
of TCE can be found in Section 4.5. Most of the immunotoxicity studies present data stratified
by sex. The prevalence of exposure to TCE is generally lower in women compared with men. In
men, the studies generally reported odds ratios between 2.0 and 8.0, and in women, the odds
ratios were between 1.0 and 2.0. Based on small numbers of cases, an occupational study of
TCE exposure found an increased risk for systemic sclerosis for men (OR: 4.75, 95% CI:
0.99-21.89) compared to women (OR: 2.10; 95% CI: 0.65-6.75) (Diot et al., 2002). Another
study found similar results, with an elevated risk for men with a maximum intensity, cumulative
intensity and maximum probability of exposure to TCE compared to women (Nietert et al.,
1998). These two studies, along with one focused exclusively on the risk of scleroderma to
women (Garabrant et al., 2003), were included in a meta-analysis conducted by the EPA
resulting in a combined estimate for "any" exposure, was OR = 2.5 (95% CI 1.1, 5.4) for men
and OR =1.2 (95% CI 0.58, 2.6) in women.
Respiratory Toxicity. No gender susceptibility to non-cancerous outcomes in the
respiratory tract was observed. A detailed discussion of the studies examining the respiratory
effects of TCE can be found in Section 4.6.
Reproductive Toxicity. A detailed discussion of the studies examining the gender-
specific noncancer reproductive effects of TCE can be found in Section 4.7.1.
Studies examining males after exposure to TCE observed altered sperm morphology and
hyperzoospermia (Chia et al., 1996), altered endocrine function (Chia et al., 1997; Goh et al.,
1998), decreased sexual drive and function (Bardodej and Vyskocil, 1956; El Ghawabi et al.,
1973; Saihan et al., 1978), and altered fertility to TCE exposure. Infertility was not associated
with TCE exposure in other studies (Forkert et al., 2003; Sallmen et al., 1998), and sperm
abnormalities were not observed in another study (Rasmussen et al., 1988).
There is more limited evidence for reproductive toxicity in females. There are
epidemiological indicators of a possible effect of TCE exposure on female fertility (Sallmen et
al., 1995), increased rate of miscarriage (ATSDR, 2001), and menstrual cycle disturbance
(ATSDR, 2001; Bardodej and Vyskocil, 1956; Zielinski, 1973). In experimental animals, the
effects on female reproduction include evidence of reduced in vitro oocyte fertilizability in rats
(Berger and Horner, 2003; Wu and Berger, 2007, 2008). However, in other studies that assessed
reproductive outcome in female rodents (Cosby and Dukelow, 1992; George et al., 1985, 1986;
Manson et al., 1984), there was no evidence of adverse effects of TCE exposure on female
reproductive function.
6/22/2009
931
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Developmental Toxicity. A detailed discussion of the studies examining the gender-
specific noncancer developmental effects of TCE can be found in Section 4.7.3. Only one study
of contaminated drinking water exposure in Camp Lejeune, North Carolina observed a higher
risk of small for gestational age (SGA) in males (ATSDR, 1998; Sonnenfeld et al., 2001).
4.9.2.1.2.2 Gender susceptibility to cancer outcomes
A detailed discussion of the studies examining the carcinogenic effects of TCE can be
found on the liver in Section 4.3, on the kidney in Section 4.4, in the immune system in Section
4.5.4, in the respiratory system in Sections 4.6.1.2 and 4.6.3, and on the reproductive system in
Section 4.7.2.
Liver Cancer. An elevated risk of liver cancer was observed for females in both human
(Raaschou-Nielsen et al., 2003) and rodent (Elfarra et al., 1998) studies. In addition, gallbladder
cancer was significantly elevated for women (Raaschou-Nielsen et al., 2003). A detailed
discussion of the studies examining the gender-specific liver cancer effects of TCE can be found
in Section 4.3.
Kidney Cancer. One study of occupational exposure to TCE observed an increase in
renal cell carcinoma for women compared to men (Dosemeci et al., 1999), but no gender
difference was observed in other studies (Pesch et al., 2000; Raaschou-Nielsen et al., 2003).
Blair et al. (1998) and Hansen et al. (2001) also present some results by sex, but both of these
studies have too few cases to be informative about a sex difference for kidney cancer. Exposure
differences between males and females in Dosemeci et al. (1999) may explain their finding.
These studies, however, provide little information to evaluate susceptibility between sexes
because of their lack of quantitative exposure assessment and lower statistical power. A detailed
discussion of the studies examining the gender-specific kidney cancer effects of TCE can be
found in Section 4.4.
Cancers of the Immune System. Two drinking water studies suggest that there may be an
increase of leukemia (Cohn et al., 1994; Fagliano et al., 1990) and NHL (Cohn et al., 1994)
among females. An occupational study also observed an elevated risk of leukemia in females
(Raaschou-Nielsen et al., 2003), although study of contaminated drinking water in Woburn,
Massachusetts observed an increased risk of childhood leukemia in males (Costas et al., 2002).
A detailed discussion of the studies examining the gender-specific cancers of the immune system
following TCE exposure can be found in Section 4.5.4.
Respiratory Cancers. One study observed significantly elevated risk of lung cancer
following occupational TCE exposure for both men and women, although the risk was found to
be higher for women (Raaschou-Nielsen et al., 2003). This same study observed a non-
6/22/2009
932
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
significant elevated risk in both men and women for laryngeal cancer, again with an increased
risk for women (Raaschou-Nielsen et al., 2003). Conversely, a study of Iowa residents with
TCE-contaminated drinking water observed a 7-fold increased annual age-adjusted incidence for
males compared to females (Isacson et al., 1985). However, other studies did not observe a
gender-related difference (ATSDR, 2003a; Blair et al., 1998; Hansen et al., 2001). A detailed
discussion of the studies examining the gender-specific respiratory cancers following TCE
exposure can be found in Sections 4.6.1.2 and 4.6.3
Reproductive Cancers. Breast cancer in females and prostate cancer in males was
reported after exposure to TCE in drinking water (Isacson et al., 1985). A statistically elevated
risk for cervical cancer, but not breast, ovarian or uterine cancer, was observed in women in
another study (Raaschou-Nielsen et al., 2003). This study also did not observe elevated prostate
or testicular cancer (Raaschou-Nielsen et al., 2003). A detailed discussion of the studies
examining the gender-specific reproductive cancers following TCE exposure can be found in
Section 4.7.2.
Other Cancers. Bladder and rectal cancer was increased in men compared to women
after exposure to TCE in drinking water, but no gender difference was observed for colon cancer
(Isacson et al., 1985). After occupational TCE exposure, bladder, stomach, colon, and
esophageal cancer was non-significantly elevated in women compared to men (Raaschou-
Nielsen et al., 2003).
4.9.2.2 Genetic Variability
Section 3.3 describes the metabolic processes involved in the metabolism of TCE.
Human variation in response to TCE exposure may be associated with genetic variation. TCE is
metabolized by both P450 and GST; therefore, it is likely that polymorphisms will alter the
response to exposure (Garte et al., 2001; Nakajima and Aoyama, 2000), as well as other
chemicals that may alter the metabolism of TCE (Lash et al., 2007). It is important to note that
even with a given genetic polymorphism, metabolic expression is not static, and depends on
lifestage (see Section 4.9.1.1.2), obesity (See Section 4.9.2.4.1), and alcohol intake (see Section
4.9.2.5.1).
4.9.2.2.1 CYP450 Genotypes
Variability in CYP450 expression occurs both within humans (Dome et al., 2005) and
across experimental animal species (Nakajima et al., 1993). In particular, increased CYP2E1
activity may lead to increased susceptibility to TCE (Lipscomb et al., 1997). The CYP2E1*3
allele and the CYP2E1*4 allele were more common among those who developed scleroderma
6/22/2009
933
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
who were exposed to solvents including TCE (Povey et al., 2001). A PBPK model of CYP2E1
expression after TCE exposure has been developed for rats and humans (Yoon et al., 2007).
In experimental animals, toxicokinetics of TCE differed among CYP2E1 knockout and
wild-type mice (Kim and Ghanayem, 2006). This study found that exhalation was more
prevalent among the knockout mice, whereas urinary excretion was more prevalent among the
wild-type mice. In addition, the dose was found to be retained to a greater degree by the
knockout mice compared to the wild-type mice.
4.9.2.2.2 GST Genotype
There is a possibility that GST polymorphisms could play a role in variability in toxic
response (Caldwell and Keshava, 2006), but this has not been sufficiently tested (NRC, 2006).
One study of renal cell cancer in workers exposed to TCE demonstrated a significant increased
for those with GSTM1+ and GSTT1+ polymorphisms, compared to a negative risk for those with
GSTM1- and GSTT1- polymorphisms (Briining et al., 1997). However, another study did not
confirm this hypothesis, observing no clear relationship between GSTM1 and GSTT1
polymorphisms and renal cell carcinoma among TCE exposed individuals, although they did see
a possible association with the homozygous wild-type allele GSTP1*A (Wiesenhiitter et al.,
2007). A third study unrelated to TCE exposure found GSTT1- to be associated with an
increased risk of renal cell carcinoma, but no difference was seen for GSTM1 and GSTP1 alleles
(Sweeney et al., 2000).
4.9.2.2.3 Other Genotypes
Other genetic polymorphisms could play a role in variability in toxic response, in
particular TCE-related skin disorders. Studies have found that many TCE-exposed patients
diagnosed with skin conditions exhibited the slow-acetylator NAT2 genotype (Huang et al.,
2002; Nakajima et al., 2003); whereas there was no difference in NAT2 status for those
diagnosed with renal cell carcinoma (Wiesenhiitter et al., 2007). Other studies have found that
many TCE-exposed patients diagnosed with skin conditions expressed variant HLA alleles (Li et
al., 2007; Yue et al., 2007), in particular HLA-B*1301 which is more common in Asians
compared to whites (Cao et al., 2001; Williams et al., 2001); or TNF a-308 allele (Dai et al.,
2004). Also, an in vitro study of human lung adenocarcinoma cells exposed to TCE varied in
response based on their p53 status, with p53-wild type cells resulting in severe cellular damage,
but not the p53-null cells (Chen et al., 2002).
6/22/2009
934
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.9.2.3 Race/Ethnicity
Different racial or ethnic groups may express metabolic enzymes in different ratios and
proportions due to genetic variability (Garte et al., 2001). In particular, ethnic variability in
CYP450 expression has been reported (Dome et al., 2005; McCarver et al., 1998; Parkinson et
al., 2004; Shimada et al., 1994; Stephens et al., 1994). It has been observed that the metabolic
rate for TCE may differ between the Japanese and Chinese (Inoue et al., 1989). Also, body size
varies among ethnic groups, and increased body size was related to increased absorption of TCE
and urinary excretion of TCE metabolites (Sato et al., 1991b).
4.9.2.4 Pre-Existing Health Status
It is known that kidney and liver diseases can affect the clearance of chemicals from the
body, and therefore poor health may lead to increased half-lives for TCE and its metabolites.
There is some data indicating that obesity/metabolic syndrome, diabetes and hypertension may
increase susceptibility to TCE exposure through altered toxicokinetics. In addition, some of
these conditions lead to increased risk for adverse effects that have also been associated with
TCE exposure, though the possible interaction between TCE and known risk factors for these
effects is not understood.
4.9.2.4.1 Obesity and Metabolic Syndrome
TCE is lipophilic and stored in adipose tissue; therefore, obese individuals may have an
increased body burden of TCE (Clewell et al., 2000). Immediately after exposure, blood
concentrations are higher and urinary excretion of metabolites are faster in thin men than obese
men due to the storage of TCE in the fat. However, the release of TCE from the fat tissue
beginning three hours after exposure reverses this trend and obese men have increased blood
concentrations and urinary excretion of metabolites are compared to thin men (Sato, 1993; Sato
et al., 1991b). This study also reported that increased body size was related to increased
absorption and urinary excretion of TCE metabolites (Sato et al., 1991b). After evaluating the
relationship between mean daily uptake and mean minute volume, body weight, lean body mass,
and amount of adipose tissue, the variation in uptake was more closely correlated with lean body
mass, but not adipose tissue content (Monster et al., 1979). Thus adipose tissue may play an
important role in post-exposure distribution, but is not a primary determinant of TCE uptake.
Increased CYP2E1 expression has been observed in obese individuals (McCarver et al., 1998).
Accumulation into adipose tissue may prolong internal exposures (Davidson and Beliles, 1991;
6/22/2009
935
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Lash et al., 2000), as evidenced by increased durations of elimination in subjects with larger
body mass indices (Monster, 1979).
In addition, individuals with high BMI are at increased risk of some of the same health
effects associated with TCE exposure. For example, renal cell carcinoma, liver cancer, and
prostate cancer may be positively associated with BMI or obesity (Asal et al., 1988a, b;
Benichou et al., 1998; El-Serag and Rudolph, 2007; Wigle et al., 2008). However, whether and
how TCE interacts with known risk factors for such diseases is unknown, as existing
epidemiologic studies have only examined these factors as possible confounders for effects
associated with TCE, or vice versa (Charbotel et al., 2006; Krishnadasan et al., 2008).
4.9.2.4.2 Diabetes
A higher rate of diabetes in females exposed to TCE was reported in two studies (Burg et
al., 1995; Davis et al., 2005). Whether the TCE may have caused the diabetes or the diabetes
may have increased susceptibility to TCE is not clear. However, it has been observed that
CYP2E1 expression is increased in obese Type II diabetics (Wang et al., 2003), and in poorly
controlled Type I diabetics (Song et al., 1990), which may consequently alter the metabolism of
TCE.
4.9.2.4.3 Hypertension
One study found no difference in risk for renal cell carcinoma among those diagnosed
with hypertension among those living in an area with high TCE exposure; however, a slightly
elevated risk was seen for those being treated for hypertension (OR: 1.57, 95% CL0.90-2.72)
(Charbotel et al., 2006). Unrelated to TCE exposure, hypertension has been associated with
increase risk of renal cell carcinoma in women (Benichou et al., 1998).
4.9.2.5 Lifestyle Factors and Nutrition Status
4.9.2.5.1 Alcohol Intake
A number of studies have examined the interaction between TCE and ethanol exposure in
both humans (Bardodej and Vyskocil, 1956; Barret et al., 1984; McCarver et al., 1998; Miiller et
al., 1975; Sato, 1993; Sato et al., 1981, 1991a; Stewart et al., 1974) and experimental animals
(Kaneko et al., 1994; Larson and Bull, 1989; Nakajima et al., 1988, 1990, 1992b; Okino et al.,
1991; Sato et al., 1980, 1983; Sato and Nakajima, 1985; White and Carlson, 1981).
The co-exposure causes metabolic inhibition of TCE in humans (Miiller et al., 1975;
Windemuller and Ettema, 1978), male rats (Kaneko et al., 1994; Larson and Bull, 1989;
6/22/2009
936
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Nakajima et al., 1988, 1990; Nakanishi et al., 1978; Okino et al., 1991; Sato andNakajima, 1985;
Sato et al., 1981), and rabbits (White and Carlson, 1981). Similarly, individuals exposed to TCE
reported an increase in alcohol intolerance (Bardodej and Vyskocil, 1956; Grandjean et al., 1955;
Rasmussen and Sabroe, 1986). Disulfiram, used to treat alcoholism, has also been found to
decrease the elimination of TCE and TCA (Bartonicek and Teisinger, 1962).
A "degreasers flush" has been described, reflecting a reddening of the face of those
working with TCE after drinking alcohol, and measured an elevated level of TCE in exhaled
breath compared to nondrinkers exposed to TCE (Stewart et al., 1974). This may be due to
increased CYP2E1 expression in those that consume alcohol (Caldwell et al., 2008;
Liangpunsakul et al., 2005; Lieber, 2004; McCarver et al., 1998; Parkinson et al., 2004; Perrot et
al., 1989), which has also been observed in male rats fed alcohol (Nakajima et al., 1992b),
although another study of male rats observed that ethanol did not decrease CYP450 activity
(Okino et al., 1991). It is important to note that there a further increased response of TCE and
ethanol has been reported when also combined with low fat diets or low carbohydrate diets in
male rats (Sato et al., 1983).
Since the liver is a target organ for both TCE and alcohol, decreased metabolism of TCE
could be related to cirrhosis of the liver as a result of alcohol abuse (McCarver et al., 1998), and
an in increase in clinical liver impairment along with degreasers flush has been observed (Barret
et al., 1984).
The central nervous system may also be impacted by the co-exposure. Individuals
exposed to TCE and ethanol reported an increase in altered mood states (Reif et al., 2003),
decreased mental capacity as described as small increases in functional load (Windemuller and
Ettema, 1978), and those exposed to TCE and tetrachloroethylene who consumed alcohol had an
elevated color confusion index (Valic et al., 1997).
4.9.2.5.2 Tobacco Smoking
Individuals who smoke tobacco may be at increased risk of the health effects from TCE
exposure. One study examining those living in an area with high TCE exposure found an
increasing trend of risk (p = 0.008) for renal cell carcinoma among smokers, with the highest OR
among those with > 40 pack-years (OR = 3.27, 95% CI = 1.48-7.19) (Charbotel et al., 2006). It
has been shown that renal cell carcinoma is independently associated with smoking in a dose-
response manner (Yuan et al., 1998), particularly in men (Benichou et al., 1998).
A number of factors correlated to smoking (e.g., socioeconomic status, diet, alcohol
consumption) may positively confound results if greater smoking rates were over-represented in
a cohort (Raaschou-Nielsen et al., 2003). Absence of smoking information, on the other hand,
could introduce a negative bias. Morgan and Cassidy (2002) noted the relatively high education
6/22/2009
937
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
high income levels, and high access to health care of subjects in this study compared to the
averages for the county as a whole likely leads to a lower smoking rate. Garabrant et al. (1988)
similarly attributed their observations to negative selection bias introduced when comparison is
made to national mortality rats known as "the healthy worker effect."
4.9.2.5.3 Nutritional Status
Malnutrition may also increase susceptibility to TCE. Bioavailability of TCE after oral
and intravenous exposure increased with fasting from approximately 63% in non-fasted rats to
greater than 90% in fasted rats, with blood levels in fasted rats were elevated 2-3-fold, and
increased half-life in the blood of fasted rats (D'Souza et al., 1985). Food deprivation (Sato and
Nakajima, 1985) and carbohydrate restriction (Nakajima et al., 1982; Sato and Nakajima, 1985)
enhanced metabolism of TCE in male rats, but this was not observed for dietary changes in
protein or fat levels (Nakajima et al., 1982).
Vitamin intake may also alter susceptibility to TCE. An in vitro study of cultured normal
human epidermal keratinocyte demonstrated an increased lipid peroxidation in a dose-dependant
manner after exposure to TCE, which were then attenuated by exposure to Vitamin E (Ding et
al., 2006).
4.9.2.5.4 Physical Activity
Increased inhalation during physical activity leads increases TCE concentrations in the
alveoli when compared to inhalation in a resting state (Astrand, 1975). Studies have examined
the time course of inhaled TCE and metabolites in blood and urine in individuals with different
workloads (Astrand and Ovrum, 1976; Jakubowski and Wieczorek, 1988; Monster et al., 1976;
Vesterberg et al., 1976; Vesterberg and Astrand, 1976). These studies demonstrate that an
increase in pulmonary ventilation increases the amount of TCE taken up during exposure
(Astrand and Ovrum, 1976; Jakubowski and Wieczorek, 1988; Monster et al., 1976; Sato, 1993).
The Rocketdyne aerospace cohort exposed to TCE (and other chemicals) found a
protective effect with high physical activity, but only after controlling for TCE exposure and
socioeconomic status (OR = 0.55, 95% CI = 0.32-0.95, p trend = 0.04) (Krishnadasan et al.,
2008). In general, physical activity may provide a protective effect for prostate cancer (Wigle et
al., 2008) (see Section 4.7.3.1.1).
6/22/2009
938
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.9.2.5.5 Socioeconomic Status
Socioeconomic status (SES) can be an indicator for a number of co-exposures, such as
increased tobacco smoking, poor diet, education, income, and health care access, which may play
a role in the results observed in the health effects of TCE exposure (Morgan and Cassidy, 2002).
Children's exposure to TCE was measured in a low SES community, as characterized by
income, educational level, and receipt of free or reduced cost school meals (Sexton et al., 2005);
however, this study did not compare data to a higher SES community, nor examine health
effects.
An elevated risk of NHL and esophagus/adenocarcinoma after exposure to TCE was
observed for blue-collar workers compared to white collar and unknown SES (Raaschou-Nielsen
et al., 2003). Authors speculate that these results could be confounding due to other related
factors to SES such as smoking.
4.9.3 Uncertainty of Database for Susceptible Populations
There is some evidence that certain subpopulations may be more susceptible to exposure
to TCE. These subpopulations include early and later lifestages, gender, genetic polymorphisms,
race/ethnicity, pre-existing health status, and lifestyle factors and nutrition status. Although
there is more information on early life exposure to TCE than on other potentially susceptible
populations, there remain a number of uncertainties regarding children's susceptibility.
Improved PBPK modeling for using childhood parameters early lifestages as recommended by
the NRC (2006), and validation of these models, will aid in determining how variations in
metabolic enzymes affect TCE metabolism. In particular, the NRC states that it is prudent to
assume children need greater protection than adults—unless sufficient data are available to
justify otherwise (NRC, 2006).
More studies specifically designed to evaluate effects in early and later lifestages are
needed in order to more fully characterize potential life stage-related TCE toxicity. Because the
neurological effects of TCE constitute the most sensitive endpoints of concern for noncancer
effects, it is quite likely that the early lifestages may be more susceptible to these outcomes than
are adults. Lifestage-specific neurotoxic effects, particularly in the developing fetus, need
further evaluation. It is important to consider the use of age-appropriate testing for assessment of
these and other outcomes, both for cancer and noncancer outcomes. Data specific to the
carcinogenic effects of TCE exposure during the critical periods of development of experimental
animals and humans also are sparse.
6/22/2009
939
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
There is a need to better characterize the implications of TCE exposures to susceptible
populations. There is suggestive evidence that there may be greater susceptibility for exposures
to the elderly. Gender and race/ethnic differences in susceptibility are likely due to variation in
physiology and exposure, and genetic variation likely has an effect on the toxicokinetics of TCE.
Diminished health status (e.g., impaired kidney liver or kidney), alcohol consumption, tobacco
smoking, and nutritional status will likely affect an individual's ability to metabolize TCE. In
addition, further evaluation of the effects due to co-exposures to other compounds with similar or
different MO As need to be evaluated. Future research should better characterize possible
susceptibility for certain lifestages or populations.
6/22/2009
940
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
References
Adams, C; Keil, D; Meyers, K; EuDaly, A; Smythe, J; EuDaly, J; Gilkeson, G; Peden-Adams,
MM. (2003) Lifetime exposure to trichloroethylene (TCE) modulates immune function.
Toxicologist 72(S-1):375.
Adgate, JL; Eberly, LE; Stroebel, C; Pellizzari, ED; Sexton K. (2004a) Personal, indoor, and
outdoor VOC exposures in a probability sample of children. J Expo Anal Environ Epidemiol. 14
Suppl 1 :S4—S13.
Adgate, JL; Church, TR; Ryan, AD; Ramachandran, G; Fredrickson, AL; Stock, TH; Morandi,
MT; Sexton, K. (2004b) Outdoor, indoor, and personal exposure to VOCs in children. Environ
Health Perspect 112(14): 1386-1392.
Asal NR, Risser DR, Kadamani S, Geyer JR, Lee ET, Cherng N. (1988a) Risk factors in renal
cell carcinoma: I. Methodology, demographics, tobacco, beverage use, and obesity. Cancer
Detect Prev 11(3—6):359—77.
Asal NR, Geyer JR, Risser DR, Lee ET, Kadamani S, Cherng N. (1988b) Risk factors in renal
cell carcinoma. II. Medical history, occupation, multivariate analysis, and conclusions. Cancer
Detect Prev 13(3-4):263-79.
Astrand I. (1975). Uptake of solvents in the blood and tissues of man. A review. Scand J Work
Environ Health. 1(4): 199-218.
Astrand I, Ovrum P. (1976) Exposure to trichloroethylene I. Uptake and distribution in man.
Scand J Work Environ Health 2(4): 199-211.
ATSDR (Agency for Toxic Substances and Disease Registry). (1998) Volatile organic
compounds in drinking water and adverse pregnancy outcomes. U.S. Department of Health and
Human Services, August 1998, Atlanta, GA.
ATSDR. (2001) Evaluation of priority health conditions in a community with historical
contamination by trichloroethylene. U.S. Department of Health and Human Services, August
2001, Atlanta, GA.
ATSDR. (2003a) Impact of trichloroethylene exposure on oral motor, speech and hearing in
children. U.S. Department of Health and Human Services, July 2003, Atlanta, GA.
6/22/2009
941
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
ATSDR. (2003b) Progress report: survey of specific childhood cancers and birth defects among
children whose mothers were pregnant while living at U.S. Marine Corps Base Camp Lejeune,
North Carolina, 1968-1985. U.S. Department of Health and Human Services, July 2003, Atlanta,
GA.
ATSDR. (2006) Cancer and birth outcome analysis, Endicott Area, Town of Union, Broome
County, New York. U.S. Department of Health and Human Services, May 2006, Atlanta, GA.
Bardodej, Z; Vyskocil, J. (1956) The problem of trichloroethylene in occupational medicine;
trichloroethylene metabolism and its effect on the nervous system evaluated as a means of
hygienic control. AMA Arch Ind Health 13(6):581—592.
Barret L, Faure J, Guilland B, Chomat D, Didier B, Debru JL. (1984) Trichloroethylene
occupational exposure: elements for better prevention. Int Arch Occup Environ Health
53(4):283-9.
Barton HA, Flemming CD, Lipscomb JC. (1986) Evaluating human variability in chemical risk
assessment: hazard identification and dose-response assessment for noncancer oral toxicity of
trichloroethylene. Toxicology 111(1—3):271—87.
Bartonicek V, Teisinger J. (1962) Effect of tetraethyl thiuram disulphide (disulfiram) on
metabolism of trichloroethylene in man. Brit J Industr Med 19:216-221.
Benedetti MS, Whomsley R, Canning M. (2007) Drug metabolism in the paediatric population
and in the elderly. Drug Discov Today 12(15—16):599—610.
Benichou J, Chow WH, McLaughlin JK, Mandel JS, Fraumeni JF Jr. (1998) Population
attributable risk of renal cell cancer in Minnesota. Am J Epidemiol 148(5):424-30.
Beppu, K. (1968) Transmission of the anesthetic agents through the placenta in painless delivery
and their effects on newborn infants. Keio J Med 17(2):81—107.
Berger T, Horner CM. (2003) In vivo exposure of female rats to toxicants may affect oocyte
quality. Reprod Toxicol 17:273-281.
Bernad PG, Newell S, Spyker DA. (1987) Neurotoxicity and behavior abnormalities in a cohort
chronically exposed to trichloroethylene. Abstract. Vet Hum Toxicol 29(6):475.
6/22/2009
942
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Blair A, Hartge P, Stewart PA, McAdams M, Lubin J. (1998) Mortality and cancer incidence of
aircraft maintenance workers exposed to trichloroethylene and other organic solvents and
chemicals: extended follow up. Occup Environ Med 55:161-171.
Blake MJ, Castro L, Leeder JS, Kearns GL. (2005) Ontogeny of drug metabolizing enzymes in
the neonate. Semin Fetal Neonatal Med 10(2): 123-38.
Blossom SJ, Doss JC, Gilbert KM. (2007) Chronic exposure to a trichloroethylene metabolite in
autoimmune-prone MRL+/+ mice promotes immune modulation and alopecia. Toxicol Sci
95(2):401—11.
Blossom SJ, Doss JC, Hennings LJ, Jernigan S, Melnyk S, James SJ. (2008) Developmental
exposure to trichloroethylene promotes CD4(+) T cell differentiation and hyperactivity in
association with oxidative stress and neurobehavioral deficits in MRL+/+ mice. Toxicol Appl
Pharmacol 231(3):344-53.
Bove FJ. (1996) Public drinking water contamination and birthweight, prematurity, fetal deaths,
and birth defects. Toxicology and industrial health 12(2):255-266.
Bove FJ, Fulcomer MC, Klotz JB, Esmart J, Dufficy EM, Savrin JE. (1995) Public drinking
water contamination and birth outcomes. Am J Epidemiol 141(9):850-862.
Boyer, AS; Finch, WT; Runyan, RB. (2000) Trichloroethylene inhibits development of
embryonic heart valve precursors in vitro. Toxicol Sci 53:109-117.
Bross, G; DiFranceisco, D; Desmond, ME. (1983) The effects of low dosages of
trichloroethylene on chick development. Toxicology 28:283-294.
Briining T, Lammert M, Kempkes M, Thier R, Golka K, Bolt HM. (1997) Influence of
polymorphisms of GSTM1 and GSTT1 for risk of renal cell cancer in workers with long-term
high occupational exposure to trichloroethene. Arch Toxicol 71(9):596-9.
Briining T, Vamvakas S, Makropoulos V, Birner G. (1998) Acute intoxication with
trichloroethene: clinical symptoms, toxicokinetics, metabolism, and development of biochemical
parameters for renal damage. Toxicol Sci 41(2): 157-65.
Brzezinski MR, Boutelet-Bochan H, Person RE, Fantel AG, Juchau MR. (1999) Catalytic
activity and quantitation of cytochrome P-450 2E1 in prenatal human brain. J Pharmacol Exp
Ther 289(3): 1648-53.
6/22/2009
943
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Burg, JR; Gist, GL. (1999) Health effects of environmental contaminant exposure: an intrafile
comparison of the Trichloroethylene Subregistry. Arch Environ Health 54(4):231-241.
Burg JR, Gist GL, Allred SL, Radtke TM, Pallos LL, Cusack CD. (1995) The National Exposure
Registry - Morbidity analyses of noncancer outcomes from the trichloroethylene subregistry
baseline data. Int J Occup Med and Toxicol 4(2):237-257.
Byers VS, Levin AS, Ozonoff DM, Baldwin RW. (1988) Association between clinical symptoms
and lymphocyte abnormalities in a population with chronic domestic exposure to industrial
solvent-contaminated domestic water supply and a high incidence of leukaemia. Cancer
Immunol Immunother 27(1):77-81.
Caldwell JC, KeshavaN. (2006) Key issues in the modes of action and effects of
trichloroethylene metabolites for liver and kidney tumorigenesis. Environ Health Perspect
114(9): 1457-63.
Caldwell JC, KeshavaN, Evans MV. (2008) Difficulty of mode of action determination for
trichloroethylene: An example of complex interactions of metabolites and other chemical
exposures. Environ Mol Mutagen 49(2): 142-54.
Cao K, Hollenbach J, Shi X, Shi W, Chopek M, Fernandez-Vina MA. (2001) Analysis of the
frequencies of HLA-A, B, and C alleles and haplotypes in the five major ethnic groups of the
United States reveals high levels of diversity in these loci and contrasting distribution patterns in
these populations. Hum Immunol 62(9): 1009-30.
Carney, EW; Thorsrud, BA; Dugard, PH; Zablotny, CL. (2006) Developmental toxicity studies
in Crl:Cd (SD) rats following inhalation exposure to trichloroethylene and perchloroethylene.
Birth Defects Res B Dev Reprod Toxicol 77:405-412.
Carpenter SP, Lasker JM, Raucy JL. (1996) Expression, induction, and catalytic activity of the
ethanol-inducible cytochrome P450 (CYP2E1) in human fetal liver and hepatocytes. Mol
Pharmacol 49(2):260-8.
Charbotel B, Fevotte J, Hours M, Martin JL, Bergeret A. (2006) Case-control study on renal cell
cancer and occupational exposure to trichloroethylene. Part II: Epidemiological aspects. Ann
Occup Hyg 50(8):777-87.
6/22/2009
944
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Chen SJ, Wang JL, Chen JH, Huang RN. (2002) Possible involvement of glutathione and p53 in
trichloroethylene- and perchloroethylene-induced lipid peroxidation and apoptosis in human lung
cancer cells. Free Radic Biol Med 33(4):464-72.
Chia SE, Ong CN, Tsakok MF, Ho A. (1996) Semen parameters in workers exposed to
trichloroethylene. Reprod Toxicol 10(4):295-299.
Chia SE, Goh VH, Ong CN. (1997) Endocrine profiles of male workers with exposure to
trichloroethylene. Am J Ind Med 32(3):217-222.
Clewell HJ 3rd, Gentry PR, Covington TR, Gearhart JM. (2000) Development of a
physiologically based pharmacokinetic model of trichloroethylene and its metabolites for use in
risk assessment. Environ Health Perspect 108 Suppl 2:283-305.
Coberly, S; Oudiz, DJ; Overstreet, JW; Wiley, LM. (1992) Effects of maternal exposure to
trichloroethylene (TCE) on cell proliferation in the mouse preimplantation embryo. Reprod
Toxicol 6:241-245.
Cohn P, Klotz J, Bove F, Berkowitz M, Fagliano J. (1994) Drinking water contamination and the
incidence of leukemia and non-Hodgkin's lymphoma. Environ Health Perspect
102(6-7):556-61.
Cosby NC, Dukelow WR. (1992) Toxicology of maternally ingested trichloroethylene (TCE) on
embryonal and fetal development in mice and of TCE metabolites on in vitro fertilization.
Fundam Appl Toxicol 19:268-274.
Costas K, Knorr RS, Condon SK. (2002) A case-control study of childhood leukemia in Woburn,
Massachusetts: the relationship between leukemia incidence and exposure to public drinking
water. Sci Total Environ 300(l-3):23-35.
D'Souza RW, Bruckner JV, Feldman S. (1985) Oral and intravenous trichloroethylene
pharmacokinetics in the rat. J Toxicol Environ Health 15(5): 587—601.
Dai Y, Leng S, Li L, Niu Y, Huang H, Cheng J, Zheng Y. (2004) Genetic polymorphisms of
cytokine genes and risk for trichloroethylene-induced severe generalized dermatitis: a case-
control study. Biomarkers 9(6):470-8.
Das, RM; Scott, JE. (1994) Trichloroethylene-induced pneumotoxicity in fetal and neonatal
mice. Toxicology Letters 73:227-239.
6/22/2009
945
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Davidson IW, Beliles RP. (1991) Consideration of the target organ toxicity of trichloroethylene
in terms of metabolite toxicity and pharmacokinetics. Drug Metab Rev 23(5-6):493-599.
Davis SI, Laszlo Pallos L, Wu JQ, Sapp JH 2nd, Cusack C. (2005) ATSDR's trichloroethylene
subregistry methods and results: 1989-2000. Arch Environ Occup Health 60(3): 130-9.
Dawson BV, Johnson PD, Goldberg SJ, Ulreich JB. (1990) Cardiac teratogenesis of
trichloroethylene and dichloroethylene in a mammalian model. J Am Coll Cardiol
16(5): 1304—1309.
Dawson BV, Johnson PD, Goldberg SJ, Ulreich JB. (1993) Cardiac teratogenesis of halogenated
hydrocarbon-contaminated drinking water. J Am Coll Cardiol 21(6): 1466-1472.
De Roos AJ, Olshan AF, Teschke K, Poole C, Savitz DA, Blatt J, Bondy ML, Pollock BH.
(2001) Parental occupational exposures to chemicals and incidence of neuroblastoma in
offspring. Am J Epidemiol 154(2): 106-14.
Ding R, Shen T, Zhu QX. (2006) [Effects of TCE and PCE on cultured human keratinocyte lipid
peroxidation and protective effect of vitamin E on it] Zhonghua Lao Dong Wei Sheng Zhi Ye
Bing Za Zhi 24(11):662-5.
Diot E, Lesire V, Guilmot JL, Metzger MD, Pilore R, Rogier S, Stadler M, Diot P, Lemarie E,
Lasfargues G. (2002) Systemic sclerosis and occupational risk factors: a case-control study.
Occup Environ Med 59(8):545-9.
Dorfmueller, MA; Henne, SP; York,G; Bornschein, RL; Manson, JM. (1979) Evaluation of
teratogenicity and behavioral toxicity with inhalation exposure of maternal rats to
trichloroethylene. Toxicology 14:153-166.
Dome, JL; Walton, K; Renwick, AG. (2005) Human variability in xenobiotic metabolism and
pathway-related uncertainty factors for chemical risk assessment: a review. Food Chem Toxicol
43(2):203-16.
Dosemeci, M; Cocco, P; Chow, WH. (1999) Gender differences in risk of renal cell carcinoma
and occupational exposures to chlorinated aliphatic hydrocarbons. Am J Ind Med 36(l):54-9.
Drake, VJ; Koprowski, SL; Lough, J; Hu, N; Smith, SM. (2006a) Trichloroethylene exposure
during cardiac valvuloseptal morphogenesis alters cushion formation and cardiac hemodynamics
in the avian embryo. Environ Health Perspect 114(6):842-847.
6/22/2009
946
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Drake, VJ; Koprowski, SL; Hu, N; Smith, SM; Lough, J. (2006b) Carcinogenic effects of
trichloroethylene and trichloroacetic acid following exposure during heart specification of avian
development. Toxicol Sci 94(1): 153-162.
Elfarra AA, Krause RJ, Last AR, Lash LH, Parker JC. (1998) Species- and sex-related
differences in metabolism of trichloroethylene to yield chloral and trichloroethanol in mouse, rat,
and human liver microsomes. Drug Metab Dispos 26(8):779-785.
El Ghawabi SM, Mansoor MB, El Gamel MS, El Saharti AA, El Enany FF. 1973. Chronic
trichloroethylene exposure. J Egypt Med Assoc 56(11-12): 715-724.
Fagliano J, Berry M, Bove F, Burke T. (1990) Drinking water contamination and the incidence
of leukemia: an ecologic study. Am J Public Health 80(10): 1209-1212.
Fan, AM. (1988) Trichloroethylene: water contamination and health risk assessment. Rev
Environ Contam Toxicol 101:55-92.
Fisher JW, Whittaker TA, Taylor DH, Clewell HJ 3rd, Andersen ME. (1989) Physiologically
based pharmacokinetic modeling of the pregnant rat: a multiroute exposure model for
trichloroethylene and its metabolite, trichloroacetic acid. Toxicol Appl Pharmacol
99(3):395-414.
Fisher JW, Whittaker TA, Taylor DH, Clewell HJ 3rd, Andersen ME. (1990) Physiologically
based pharmacokinetic modeling of the lactating rat and nursing pup: a multiroute exposure
model for trichloroethylene and its metabolite, trichloroacetic acid. Toxicol Appl Pharmacol
102(3):497-513.
Fisher J, Mahle D, Bankston L, Greene R, Gearhart J. (1997) Lactational transfer of volatile
chemicals in breast milk. Am Ind Hyg Assoc J 58(6):425-431.
Fisher JW, Mahle D, Abbas R. (1998) A human physiologically based pharmacokinetic model
for trichloroethylene and its metabolites, trichloroacetic acid and free trichloroethanol. Toxicol
Appl Pharmacol 152(2):339-359.
Fisher, JW; Channel, SR; Eggers, JS; Johnson, PD; MacMahon, KL; Goodyear, CD; Sudberry,
GL; Warren, DA; Latendesse, JR; Graeter, LJ. (2001) Trichloroethylene, trichloroacetic acid,
and dichloroacetic acid: do they affect fetal rat heart development? Int J Toxicol 20:257-267.
6/22/2009
947
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Forkert, P-G; Lash, LH; Nadeau V; Tardif, R; Simmonds, A. (2002) Metabolism and toxicity of
trichloroethylene in epididymis and testis. Toxicol Appl Pharmacol 182:244-254.
Forkert PG, Lash L, Tardif R, Tanphaichitr N, Vandevoort C, Moussa M. (2003) Identification
of trichloroethylene and its metabolites in human seminal fluid of workers exposed to
trichloroethylene. Drug Metab Dispos 31(3):306—311.
Franco A, Costoya MA, Roca E. (2007) Estimating risk during showering exposure to VOCs of
workers in a metal-degreasing facility. J Toxicol Environ Health A 70(7):627-37.
Fredriksson, A; Danielsson, BRG; Eriksson, P. (1993) Altered behavior in adult mice orally
exposed to tri- and tetrachloroethylene as neonates. Toxicology Letters 66:13-19.
GAO (U.S. Government Accountability Office). (2007a) Defense Health Care: Activities related
to past drinking water contamination at Marine Corps Base Camp Lejeune. Report to
Congressional Committees. GAO-07-276.
GAO. (2007b) Defense Health Care: Issues related to past drinking water contamination at
Marine Corps Base Camp Lejeune. Statement of Marcia Crosse, Director, Health Care.
Testimony before the Subcommittee on Oversight and Investigations, Committee on Energy and
Commerce, House of Representatives. GAO-07-933T.
Gandhi M, Aweeka F, Greenblatt RM, Blaschke TF. (2004) Sex differences in pharmacokinetics
and pharmacodynamics. Annu Rev Pharmacol Toxicol 44:499-523.
Garabrant DH, Held J, Langholz B, Bernstein L. (1988) Mortality of aircraft manufacturing
workers in southern California. Am J Ind Med 13:683-693.
Garabrant DH, Lacey JV Jr, Laing TJ, Gillespie BW, Mayes MD, Cooper BC, Schottenfeld D.
(2003) Scleroderma and solvent exposure among women. Am J Epidemiol 157(6):493-500.
Garte S, Gaspari L, Alexandrie AK, Ambrosone C, et al. Metabolic gene polymorphism
frequencies in control populations. Cancer Epidemiol Biomarkers Prev. 2001
Dec; 10(12): 1239-48.
George, JD; Reel, JR; Myers, CB; Lawton, AD; Lamb, JC. (1985) Trichloroethylene:
reproduction and fertility assessment in CD-I mice when administered in the feed. RTI Project
No. 310-2344, NTP-86-068. National Institute of Environmental Health Sciences, National
Toxicology Program, RTP, NC.
6/22/2009
948
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
George, JD; Reel, JR; Myers, CB; Lawton, AD; Lamb, JC. (1986) Trichloroethylene:
reproduction and fertility assessment in F344 rats when administered in the feed. RTI Project No.
310-2344, NTP-86-085. National Institute of Environmental Health Sciences, National
Toxicology Program, RTP, NC.
George J, Byth K, Farrell GC. (1995) Age but not gender selectively affects expression of
individual cytochrome P450 proteins in human liver. Biochem Pharmacol 50(5):727-30.
Ghantous H, Danielsson BR, Dencker L, Gorczak J, Vesterberg O. (1986) Trichloroacetic acid
accumulates in murine amniotic fluid after tri- and tetrachloroethylene inhalation. Acta
Pharmacol Toxicol (Copenh) 58(2): 105-114.
Giardino NJ, Andelman JB. (1996) Characterization of the emissions of trichloroethylene,
chloroform, and l,2-dibromo-3-chloropropane in a full-size, experimental shower. J Expo Anal
Environ Epidemiol 6(4):413-23.
Ginsberg G, Hattis D, Sonawane B, Russ A, Banati P, Kozlak M, Smolenski S, Goble R. (2002)
Evaluation of child/adult pharmacokinetic differences from a database derived from the
therapeutic drug literature. Toxicol Sci 66(2): 185-200.
Ginsberg G, Hattis D, Miller R, Sonawane B. (2004a) Pediatric pharmacokinetic data:
implications for environmental risk assessment for children. Pediatrics 113(4 Suppl):973-83.
Ginsberg G, Hattis D, Sonawane B. (2004b) Incorporating pharmacokinetic differences between
children and adults in assessing children's risks to environmental toxicants. Toxicol Appl
Pharmacol 198(2): 164-83.
Ginsberg G, Hattis D, Russ A, Sonawane B. (2005) Pharmacokinetic and pharmacodynamic
factors that can affect sensitivity to neurotoxic sequelae in elderly individuals. Environ Health
Perspect 113(9): 1243-9.
Gochfeld M. (2007) Framework for gender differences in human and animal toxicology. Environ
Res 104(1):4—21.
Goh VH, Chia SE, Ong CN. (1998) Effects of chronic exposure to low doses of trichloroethylene
on steroid hormone and insulin levels in normal men. Environ Health Perspect 106(l):41-44.
Goldberg SJ, Lebowitz MD, Graver EJ, Hicks S. 1990. An association of human congenital
cardiac malformations and drinking water contaminants. J Am Coll Cardiol 16(1): 155-164.
6/22/2009
949
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Grandjean E, Miinchinger R, Turrian V, Haas PA, Knoepfel HK, Rosenmund H. (1955)
Investigations into the effects of exposure to trichloroethylene in mechanical engineering. Brit J
IndustrMed 12:131-142.
Guo H, Lee SC, Chan LY, Li WM. (2004) Risk assessment of exposure to volatile organic
compounds in different indoor environments. Environ Res 94(l):57-66.
Hakkola, J; Pasanen, M; Hukkanen, J; et al. (1996a) Expression of xenobiotic-metabolizing
cytochrome P450 forms in human full-term placenta. Biochem Pharmacol 51:403-411.
Hakkola, J; Raunio, H; Purkunen, R; et al. (1996b) Detection of cytochrome P450 gene
expression in human placenta in first trimester of pregnancy. Biochem Pharmacol 52:379-83.
Hakkola, J; Pelkonen, O; Pasanen, M; Raunio, H. (1998a) Xenobiotic-metabolizing cytochrome
P450 enzymes in the human feto- placental unit: role in intrauterine toxicity. Crit Rev Toxicol
28:35-72.
Hakkola J, Tanaka E, Pelkonen O. (1998b) Developmental expression of cytochrome P450
enzymes in human liver. Pharmacol Toxicol. 1998 May;82(5):209-17.
Hamada T, Tanaka H. (1995) Transfer of methyl chloroform, trichloroethylene and
tetrachloroethylene to milk, tissues and expired air following intraruminal or oral administration
in lactating goats and milk-fed kids. Environ Pollut 87(3):313—8.
Hansen J, Raaschou-Nielsen O, Christensen JM, Johansen I, McLaughlin JK, Lipworth L, Blot
WJ, Olsen JH. (2001) Cancer incidence among Danish workers exposed to trichloroethylene. J
Occup Environ Med 43:133-139.
Hardin, BD; Bond, GP; Sikov, MR; Andrew, FD; Beliles, RP; and Niemeier, RW. (1981)
Testing of selected work place chemicals for teratogenic potential. Scand J Work Environ Hlth
7(Suppl 45):66-75.
Hattis D, Ginsberg G, Sonawane B, Smolenski S, Russ A, Kozlak M, Goble R. (2003)
Differences in pharmacokinetics between children and adults—II. Children's variability in drug
elimination half-lives and in some parameters needed for physiologically-based pharmacokinetic
modeling. Risk Anal 23(1): 117-42.
Healy, TEJ; Poole, TR; Hopper, A. (1982) Rat fetal development and maternal exposure to
trichloroethylene 100 p.p.m. British Journal of Anaesthesiology 54:337-340.
6/22/2009
950
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Helliwell PJ, Hutton AM. (1950) Trichloroethylene anaesthesia. Anaesthesia 5(1): 4-13.
Hines RN, McCarver DG. (2002) The ontogeny of human drug-metabolizing enzymes: phase I
oxidative enzymes. J Pharmacol Exp Ther 300(2):355-60.
Huang HL, Li L, Chen B, Huang JX, Kuang SR. (2002) New problems caused by occupational
trichloroethylene exposure. Int J Immunopathol Pharmacol 15:30-32
Ikeda M. (1977) Metabolism of trichloroethylene and tetrachloroethylene in human subjects.
Environ Health Perspect 21:239-45.
Ikeda M, Imamura T. (1973) Biological half-life of trichloroethylene and tetrachloroethylene in
human subjects. Int Arch Arbeitsmed 31(3):209-24.
Ikeda M, Otsuji H, Imamura T, Komoike Y. (1972) Urinary excretion of total trichloro-
compounds, trichloroethanol, and trichloroacetic acid as a measure of exposure to
trichloroethylene and tetrachloroethylene. Br J Ind Med 29(3):328-33.
ILSI (International Life Sciences Institute). (1992) Similarities and differences between children
and adults: implications for risk assessment. New York: ILSI Press.
Inoue O, Seiji K, Kawai T, Jin C, Liu YT, Chen Z, Cai SX, Yin SN, Li GL, Nakatsuka H,
Watanabe T, Ikenda M. (1989) Relationship between vapor exposure and urinary metabolite
excretion among workers exposed to trichloroethylene. Am J Ind Med 15(1): 103-10.
Isacson P, Bean J A, Splinter R, Olson DB, Kohler J. 1985. Drinking water and cancer incidence
in Iowa. III. Association of cancer with indices of contamination. Am J Epidemiol 121:856-869.
Isaacson, LG; Taylor, DH. (1989) Maternal exposure to 1,1,2-trichloroethylene affects myelin in
the hippocampal formation of the developing rat. Brain Research 488:403-407.
Jasinska M. (1965) [General anesthesia with trichloroethylene in children with congenital
developmental defects of the jaw and face], Czas Stomatol 18(8):935-40. Translated.
Jakubowski, M; Wieczorek, H, (1988) The effects of physical effort on pulmonary uptake of
selected organic compound vapours. Pol J Occup Med 1(1):62-71.
Johnson, PD; Dawson, BV; Goldberg, SJ. (1998a) Cardiac teratogenicity of trichloroethylene
metabolites. JACC 72(2):540-545.
6/22/2009
951
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Johnson, PD; Dawson, BV; Goldberg, SJ. (1998b) A review: trichloroethylene metabolites:
potential cardiac teratogens. Environmental Health Perspectives 106(suppl. 4):995-999.
Johnson, PD; Goldberg, SJ; Mays, MZ; Dawson, BV. (2003) Threshold of trichloroethylene
contamination in maternal drinking waters affecting fetal heart development in the rat.
Environmental Health Perspectives lll(3):289-292.
Johnson, PD; Goldberg, SJ; Mays, MZ; Dawson, BV. (2005) Errata. Threshold of
trichloroethylene contamination in maternal drinking waters affecting fetal heart development in
the rat. Environmental Health Perspectives 113(1): A18.
Johnsrud EK, Koukouritaki SB, Divakaran K, Brunengraber LL, Hines RN, McCarver DG.
(2003) Human hepatic CYP2E1 expression during development. J Pharmacol Exp Ther
307(l):402-7.
Kaneko, T; Wang, PY; Sato, A. (1994) Enzymes induced by ethanol differently affect the
pharmacokinetics of trichloroethylene and 1,1,1-trichloroethane. Occup Environ Med
51(2): 113—9.
Kim D, Ghanayem BI. (2006) Comparative metabolism and disposition of trichloroethylene in
Cyp2el-/-and wild-type mice. Drug Metab Dispos 34(12):2020-7.
Kimmerle G, Eben A. (1973) Metabolism, excretion and toxicology of trichloroethylene after
inhalation. 2. Experimental human exposure. Arch Toxikol 30(2): 127-38.
King GS, Smialek JE, Troutman WG. (1985) Sudden death in adolescents resulting from the
inhalation of typewriter correction fluid. JAMA 253(11): 1604-6.
Kjellstrand P, Holmquist B, Aim P, Kanje M, Romare S, Jonsson I, Mansson L, Bjerkemo M.
(1983a) Trichloroethylene: further studies of the effects on body and organ weights and plasma
butyrylcholinesterase activity in mice. Acta Pharmacol Toxicol (Copenh) 53(5):375-84.
Kjellstrand P, Holmquist B, Mandahl N, Bjerkemo M. (1983b) Effects of continuous
trichloroethylene inhalation on different strains of mice. Acta Pharmacol Toxicol (Copenh)
53(5):369-74.
Krishnadasan A, Kennedy N, Zhao Y, Morgenstern H, Ritz B. (2008) Nested case-control study
of occupational physical activity and prostate cancer among workers using a job exposure
matrix. Cancer Causes Control 19(1): 107—14.
6/22/2009
952
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Lagakos SW, Wessen B, Zelen M. (1986) Analysis of contaminated well water and health effect
in Woburn, Massachusetts. J Am Stat Assoc 81:583-596.
Laham, S. (1970) Studies on placental transfer. Trichloroethylene. IMS IndMed Surg. 39(1):
46-49.
Larson JL, Bull RJ. (1989) Effect of ethanol on the metabolism of trichloroethylene. J Toxicol
Environ Health 28(4):395-406.
Lash LH, Qian W, Putt DA, Jacobs K, Elfarra AA, Krause RJ, Parker JC. (1998) Glutathione
conjugation of trichloroethylene in rats and mice: sex-, species-, and tissue-dependent
differences. Drug Metab Dispos 26(1): 12-9.
Lash LH, Fisher JW, Lipscomb JC, Parker JC. (2000) Metabolism of trichloroethylene. Environ
Health Perspect 108 Suppl 2:177-200.
Lash, LH; Qian, W; Putt, DA; Hueni, SE; Elfarra, AA; Krause, RJ; Parker, JC. (2001) Renal and
hepatic toxicity of trichloroethylene and its glutathione-derived metabolites in rats and mice: sex-
, species-, and tissue-dependent differences. J Pharmacol Exp Ther 297(1): 155-64.
Lash LH, Putt DA, Parker JC. (2006) Metabolism and tissue distribution of orally administered
trichloroethylene in male and female rats: identification of glutathione- and cytochrome P-450-
derived metabolites in liver, kidney, blood, and urine. J Toxicol Environ Health A
69(13):1285-309.
Lash LH, Putt DA, Huang P, Hueni SE, Parker JC. (2007) Modulation of hepatic and renal
metabolism and toxicity of trichloroethylene and perchloroethylene by alterations in status of
cytochrome P450 and glutathione. Toxicology 235(1-2): 11-26.
Lee LJ, Chan CC, Chung CW, Ma YC, Wang GS, Wang JD. (2002) Health risk assessment on
residents exposed to chlorinated hydrocarbons contaminated in groundwater of a hazardous
waste site. J Toxicol Environ Health A 65(3-4):219-35.
Lehmann I, Rehwagen M, Diez U, Seiffart A, Rolle-Kampczyk U, Richter M, et al. (2001)
Enhanced in vivo IgE production and T cell polarization toward the type 2 phenotype in
association with indoor exposure to VOC: results of the LARS study. Int J Hyg Environ Health
204(4): 211-221.
Lehmann I, Thoelke A, Rehwagen M, Rolle-Kampczyk U, Schlink U, Schulz R, et al. (2002)
6/22/2009
953
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
The influence of maternal exposure to volatile organic compounds on the cytokine secretion
profile of neonatal T cells. Environ Toxicol 17(3):203-210.
Li H, Dai Y, Huang H, Li L, Leng S, Cheng J, Niu Y, Duan H, Liu Q, Zhang X, Huang X, Xie J,
Feng Z, Wang J, He J, Zheng Y. (2007) HLA-B*1301 as a biomarker for genetic susceptibility to
hypersensitivity dermatitis induced by trichloroethylene among workers in China. Environ
Health Perspect 115(11): 1553-6.
Liangpunsakul S, Kolwankar D, Pinto A, Gorski JC, Hall SD, Chalasani N. (2005) Activity of
CYP2E1 and CYP3A enzymes in adults with moderate alcohol consumption: a comparison with
nonalcoholics. Hepatology 41(5): 1144-50.
Lieber, CS. (2004) The discovery of the microsomal ethanol oxidizing system and its
physiologic and pathologic role. Drug Metab Rev 36(3—4):511-29.
Lipscomb JC, Garrett CM, Snawder JE. (1997) Cytochrome P450-dependent metabolism of
trichloroethylene: interindividual differences in humans. Toxicol Appl Pharmacol 142(2):311-8.
Loeber, CP; Hendrix, MJC; dePinos, SD; Goldberg, SJ. (1988) Trichloroethylene: a cardiac
teratogen in developing chick embryos. Pediatric Research 24(6):740-744.
Lowengart RA, Peters JM, Cicioni C, Buckley J, Bernstein L, Preston-Martin S, Rappaport E.
(1987) Childhood leukemia and parents' occupational and home exposures. J Natl Cancer Inst
79(l):39-46.
Mahle DA, Gearhart JM, Grigsby CC, Mattie DR, Barton HA, Lipscomb JC, Cook RS. (2007)
Age-dependent partition coefficients for a mixture of volatile organic solvents in Sprague-
Dawley rats and humans. J Toxicol Environ Health A 70(20): 1745-51.
Manson, JM; Murphy, M; Richdale, N; Smith, MK. (1984) Effects of oral exposure to
trichloroethylene on female reproductive function. Toxicology 32:229-242.
McCarthy TB, Jones RD. (1983) Industrial gassing poisonings due to trichlorethylene,
perchlorethylene, and 1-1-1 trichloroethane, 1961-80. Br J Ind Med 40(4):450-5.
McCarver DG, Byun R, Hines RN, Hichme M, Wegenek W. (1998) A genetic polymorphism in
the regulatory sequences of human CYP2E1: association with increased chlorzoxazone
hydroxylation in the presence of obesity and ethanol intake. Toxicol Appl Pharmacol
152(1):276—81.
6/22/2009
954
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
McCarver DG, Hines RN. (2002) The ontogeny of human drug-metabolizing enzymes: phase II
conjugation enzymes and regulatory mechanisms. J Pharmacol Exp Ther 300(2):361-6.
McKinney PA, Alexander FE, Cartwright RA, Parker L. (1991) Parental occupations of children
with leukaemia in west Cumbria, north Humberside, and Gateshead. BMJ 302(6778):681-7.
Mishima, N; Hoffman, S; Hill, EG; Krug, EL. (2006) Chick embryos exposed to
trichloroethylene in an ex ovo culture model show selective defects in early endocardial cushion
tissue formation. Birth Defects Res A Clin Mol Teratol 76:517-527.
Monster, AC. (1979) Difference in uptake, elimination, and metabolism in exposure to
trichloroethylene, 1,1,1-trichloroethane and tetrachloroethylene. Int Arch Occup Environ Health
42(3-4):311-7.
Monster AC, Boersma G, Duba WC. (1976) Pharmacokinetics of trichloroethylene in volunteers,
influence of workload and exposure concentration. Int Arch Occup Environ Health
38(2):87-102.
Monster AC, Boersma G, Duba WC. (1979) Kinetics of trichloroethylene in repeated exposure
of volunteers. Int Arch Occup Environ Health 42(3-4):283-92.
Morgan JW, Cassady RE. (2002) Community cancer assessment in response to long-time
exposure to perchlorate and trichloroethylene in drinking water. J Occup Environ Med
44(7):616-21.
Miiller G, Spassowski M, Henschler D. (1975) Metabolism of trichloroethylene in man. III.
Interaction of trichloroethylene and ethanol. Arch Toxicol 33(3): 173-89.
Nakajima T, Koyama Y, Sato A. (1982) Dietary modification of metabolism and toxicity of
chemical sub stances—with special reference to carbohydrate. Biochem Pharmacol
31 (6): 1005—11.
Nakajima T, Okino T, Okuyama S, Kaneko T, Yonekura I, Sato A. (1988) Ethanol-induced
enhancement of trichloroethylene metabolism and hepatotoxicity: difference from the effect of
phenobarbital. Toxicol Appl Pharmacol 94(2):227-37.
Nakajima T, Wang RS, Murayama N, Sato A. (1990) Three forms of trichloroethylene-
metabolizing enzymes in rat liver induced by ethanol, phenobarbital, and 3-methylcholanthrene.
Toxicol Appl Pharmacol 102(3):546-52.
6/22/2009
955
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Nakajima T, Wang RS, Katakura Y, Kishi R, Elovaara E, Park SS, Gelboin HV, Vainio H.
(1992a) Sex-, age- and pregnancy-induced changes in the metabolism of toluene and
trichloroethylene in rat liver in relation to the regulation of cytochrome P450IIE1 and P450IIC11
content. J Pharmacol Exp Ther 261(3):869-74.
Nakajima T, Wang RS, Elovaara E, Park SS, Gelboin HV, Vainio H. (1992b) A comparative
study on the contribution of cytochrome P450 isozymes to metabolism of benzene, toluene and
trichloroethylene in rat liver. Biochem Pharmacol 43(2):251-7.
Nakajima T, Wang RS, Elovaara E, Park SS, Gelboin HV, Vainio H. (1993) Cytochrome P450-
related differences between rats and mice in the metabolism of benzene, toluene and
trichloroethylene in liver microsomes. Biochem Pharmacol 45(5): 1079-85.
Nakajima T, Kamijo Y, Usuda N, Liang Y, Fukushima Y, Kametani K, Gonzalez FJ, Aoyama T.
(2000) Sex-dependent regulation of hepatic peroxisome proliferation in mice by
trichloroethylene via peroxisome proliferator-activated receptor alpha (PPARalpha).
Carcinogenesis 21(4):677-82.
Nakajima T, Aoyama T. (2000) Polymorphism of drug-metabolizing enzymes in relation to
individual susceptibility to industrial chemicals. Ind Health 38(2): 143-52.
Nakajima T, Yamanoshita O, Kamijima M, Kishi R, Ichihara G. (2003) Generalized skin
reactions in relation to trichloroethylene exposure: a review from the viewpoint of drug-
metabolizing enzymes. J Occup Health 45(1):8—14.
Nakanishi S, Shiohara E, Tsukada M, Yamazaki H, Okumura K. (1978) Acetaldehyde level in
the blood and liver aldehyde dehydrogenase activities in trichloroethylene-treated rats. Arch
Toxicol 41(3):207-14.
Narotsky, MG; Kavlock, RJ. (1995) A multidisciplinary approach to toxicological screening: II.
Developmental toxicity. J Toxicol Environ Health 45:145-171.
Narotsky, MG; Weller, EA; Chinchilli, VM; Kavlock, RJ. (1995) Nonadditive developmental
toxicity in mixtures of trichloroethylene, di(2-ethylhexyl) phthalate, and heptachlor in a 5 x 5 x 5
design. Fundam Appl Toxicol 27:203-216.
Nietert PJ, Sutherland SE, Silver RM, Pandey JP, Knapp RG, Hoel DG, Dosemeci M. (1998) Is
occupational organic solvent exposure a risk factor for scleroderma? Arthritis Rheum
41(6): 111 1—8. Erratum in: Arthritis Rheum 1998 Aug;41(8):1512.
6/22/2009
956
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Noland-Gerbec, EA; Pfohl, RJ; Taylor, DH; Bull RJ. (1986) 2-Deoxyglucose uptake in the
developing rat brain upon pre- and postnatal exposure to trichloroethylene. NeuroToxicology
7(3):157-164.
Nomiyama K, Nomiyama H. (1971) Metabolism of trichloroethylene in human. Sex difference
in urinary excretion of trichloroacetic acid and trichloroethanol. Int Arch Arbeitsmed
28(l):37-48.
NRC (National Research Council). (1993) Pesticides in the diets of infants and children.
Washington, DC: National Academy Press.
NRC (National Research Council). (2006) Assessing the human health risks of trichloroethylene
- key scientific issues. The National Academies Press, Washington, DC.
Okino T, Nakajima T, Nakano M. (1991) Morphological and biochemical analyses of
trichloroethylene hepatotoxicity: differences in ethanol- and phenobarbital-pretreated rats.
Toxicol Appl Pharmacol 108(3):379-89.
Parker MJ, Tarlow MJ, Anderson JM. (1984) Glue sniffing and cerebral infarction. Arch Dis
Child 59(7):675-7.
Parkinson A, Mudra DR, Johnson C, Dwyer A, Carroll KM. (2004) The effects of gender, age,
ethnicity, and liver cirrhosis on cytochrome P450 enzyme activity in human liver microsomes
and inducibility in cultured human hepatocytes. Toxicol Appl Pharmacol 199(3): 193-209.
Pastino GM, Yap WY, Carroquino M. (2000) Human variability and susceptibility to
trichloroethylene. Environ Health Perspect 108 Suppl 2: 201-214.
Peden-Adams MM, Eudaly JG, Heesemann LM, Smythe J, Miller J, Gilkeson GS, et al. (2006).
Developmental immunotoxicity of trichloroethylene (TCE): studies in B6C3F1 mice. J Environ
Sci Health A Tox Hazard Subst Environ Eng 41:249-271.
Peden-Adams MM, Eudaly JG, Lee AM, Miller J, Keil DE, Gilkeson GS. (2008). Lifetime
exposure to trichloroethylene (TCE) does not accelerate autoimmune disease in MRL +/+ mice. J
Environ Sci Health A Tox Hazard Subst Environ Eng 43:1402-1409.
Pelekis M, Gephart LA, Lerman SE. (2001) Physiological-model-based derivation of the adult
and child pharmacokinetic intraspecies uncertainty factors for volatile organic compounds. Regul
Toxicol Pharmacol 33(1): 12-20.
6/22/2009
957
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Pellizzari ED, Hartwell TD, Harris BS 3rd, Waddell RD, Whitaker DA, Erickson MD. (1982)
Purgeable organic compounds in mother's milk. Bull Environ Contam Toxicol 28(3):322-8.
Perrot N, Nalpas B, Yang CS, Beaune PH. (1989) Modulation of cytochrome P450 isozymes in
human liver, by ethanol and drug intake. Eur J Clin Invest 19(6): 549-55.
Pesch B, Haerting J, Ranft U, Klimpel A, Oelschlagel B, Schill W. (2000) Occupational risk
factors for renal cell carcinoma: agent-specific results from a case-control study in Germany.
MURC Study Group. Multicenter urothelial and renal cancer study. Int J Epidemiol
29(6): 1014-24.
Peters FM, Preston-Martin S, Yu MC. (1981) Brain tumors in children and occupational
exposure of parents. Science 213(4504):235-7.
Peters JM, Garabrant DH, Wright WE, Bernstein L, Mack TM. (1985) Uses of a cancer registry
in the assessment of occupational cancer risks. Natl Cancer Inst Monogr 69:157-61.
Phillips TJ, Macdonald RR. (1971) Comparative effect of pethidine, trichloroethylene, and
Entonox on fetal and neonatal acid-base and P02. British Medical Journal 3(5774):558-560.
Povey A, Guppy MJ, Wood M, Knight C, Black CM, Silman AJ. (2001) Cytochrome P2
polymorphisms and susceptibility to scleroderma following exposure to organic solvents.
Arthritis Rheum 44(3):662-5.
Raaschou-Nielsen O, Hansen J, McLaughlin JK, Kolstad H, Christensen JM, Tarone RE, Olsen
JH. (2003) Cancer risk among workers at Danish companies using trichloroethylene: a cohort
study. Am JEpidemiol 158(12): 1182—92.
Raijmakers MT, Steegers EA, Peters WH. (2001) Glutathione S-transferases and thiol
concentrations in embryonic and early fetal tissues. Hum Reprod 16(11):2445-50.
Rasmussen, K; Sabroe, S. (1986) Neuropsychological symptoms among metal workers exposed
to halogenated hydrocarbons. Scand J Soc Med 14(3): 161-168.
Rasmussen K, Sabroe S, Wohlert M, Ingerslev HJ, Kappel B, Nielsen J. (1988) A genotoxic
study of metal workers exposed to trichloroethylene. Sperm parameters and chromosome
aberrations in lymphocytes. Int Arch Occup Environ Health 60(6):419-423.
6/22/2009
958
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Rasmussen K, Jeppesen HJ, Sabroe S. (1993a) Solvent-induced chronic toxic encephalopathy.
Am J Industr Med 23:779-792.
Rasmussen K, Jeppesen HJ, Sabroe S. (1993b) Psychometric tests for assessment of brain
function after solvent exposure. Am J Industr Med 24:553-565.
Rasmussen K, Arlien-Soborg P, Sabroe S. (1993c) Clinical neurological findings among metal
degreasers exposed to chlorinated solvents. Acta Neurol Scand 87(3):200-204.
Reif JS, Burch JB, Nuckols JR, Metzger L, Ellington D, Anger WK. (2003) Neurobehavioral
effects of exposure to trichloroethylene through a municipal water supply. Environ Res
93(3):248-58.
Rodenbeck SE, Sanderson LM, Rene A. (2000) Maternal exposure to trichloroethylene in
drinking water and birth-weight outcomes. Arch Environ Health 55(3): 188-194.
Rodriguez CE, Mahle DA, Gearhart JM, Mattie DR, Lipscomb JC, Cook RS, Barton HA. (2007)
Predicting age-appropriate pharmacokinetics of six volatile organic compounds in the rat
utilizing physiologically based pharmacokinetic modeling. Toxicol Sci 98(l):43-56.
Rufer, ES; Hacker, T; Lough, J; Smith, SM. (2008) Low-dose trichloroethylene exposure during
valvuloseptal morphogenesis causes ventricular septal defects in hatched chicks. Toxicologist
102(S-1):316.
Saihan EM, Burton JL, Heaton KW. (1978) A new syndrome with pigmentation, scleroderma,
gynaecomastia, Raynaud's phenomenon and peripheral neuropathy. Br J Dermatol
99(4):437-440.
Sallmen M, Lindbohm ML, Kyyronen P, Nykyri E, Anttila A, Taskinen H, et al. (1995) Reduced
fertility among women exposed to organic solvents. Am J Ind Med 27(5):699-713.
Sallmen M, Lindbohm ML, Anttila A, Kyyronen P, Taskinen H, Nykyri E, Hemminki K. (1998)
Time to pregnancy among the wives of men exposed to organic solvents. Occup Environ Med
55(l):24-30.
Sato A, Nakajima T, Koyama Y. (1980) Effects of chronic ethanol consumption on hepatic
metabolism of aromatic and chlorinated hydrocarbons in rats. Br J Ind Med 37(4):382-6.
6/22/2009
959
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Sato A, Nakajima T, Koyama Y. (1981) Dose-related effects of a single dose of ethanol on the
metabolism in rat liver of some aromatic and chlorinated hydrocarbons. Toxicol Appl Pharmacol
60(1):8—15.
Sato A, Nakajima T, Koyama Y. (1983) Interaction between ethanol and carbohydrate on the
metabolism in rat liver of aromatic and chlorinated hydrocarbons. Toxicol Appl Pharmacol
68(2):242-9.
Sato A, Nakajima T. (1985) Enhanced metabolism of volatile hydrocarbons in rat liver following
food deprivation, restricted carbohydrate intake, and administration of ethanol, phenobarbital,
polychlorinated biphenyl and 3-methylcholanthrene: a comparative study. Xenobiotica
15(1):67—75.
Sato A, Endoh K, Kaneko T, Johanson G. (1991a) Effects of consumption of ethanol on the
biological monitoring of exposure to organic solvent vapours: a simulation study with
trichloroethylene. Br J Ind Med 48(8):548-556.
Sato A, Endoh K, Kaneko T, Johanson G. (1991b) A simulation study of physiological factors
affecting pharmacokinetic behaviour of organic solvent vapours. Br J Ind Med 48(5):342-7.
Sato, A. (1993) Confounding factors in biological monitoring of exposure to organic solvents. Int
Arch Occup Environ Health 65(1 Suppl):S61-7.
Schwetz, BA; Leong, BKJ; Gehring, PJ. (1975) The effect of maternally inhaled
trichloroethylene, perchloroethylene, methyl chloroform, and methylene chloride on embryonal
and fetal development in mice and rats. Toxicol Appl Pharmacol 32:84-96.
Sexton K, Adgate JL, Church TR, Ashley DL, Needham LL, Ramachandran G, Fredrickson AL,
Ryan AD. (2005) Children's exposure to volatile organic compounds as determined by
longitudinal measurements in blood. Environ Health Perspect 113(3):342—9.
Sexton K, Mongin SJ, Adgate JL, Pratt GC, Ramachandran G, Stock TH, Morandi MT. (2007)
Estimating volatile organic compound concentrations in selected microenvironments using time-
activity and personal exposure data. J Toxicol Environ Health A 70(5):465-76.
Shao J, Stapleton PL, Lin YS, Gallagher EP. (2007) Cytochrome p450 and glutathione s-
transferase mRNA expression in human fetal liver hematopoietic stem cells. Drug Metab Dispos
35(1): 168—75.
6/22/2009
960
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Shimada T, Yamazaki H, Mimura M, Inui Y, Guengerich FP. (1994) Interindividual variations in
human liver cytochrome P-450 enzymes involved in the oxidation of drugs, carcinogens and
toxic chemicals: studies with liver microsomes of 30 Japanese and 30 Caucasians. J Pharmacol
Exp Ther 270(1 ):414-23.
Shu XO, Stewart P, Wen WQ, Han D, Potter JD, Buckley JD, et al. (1999) Parental occupational
exposure to hydrocarbons and risk of acute lymphocytic leukemia in offspring. Cancer
Epidemiol Biomarkers Prev 8(9):783-791.
Smith MK; Randall, JL; Read, EJ; Stober, JA. (1989) Teratogenic activity of trichloroacetic acid
in the rat. Teratology 40:445-451.
Smith MK; Randall, JL; Read, EJ; Stober, JA. (1992) Developmental toxicity of dichloroacetate
in the rat. Teratology 46:217-223.
Song BJ, Veech RL, Saenger P. (1990) Cytochrome P450IIE1 is elevated in lymphocytes from
poorly controlled insulin-dependent diabetics. J Clin Endocrinol Metab 71(4): 1036-40.
Sonnenfeld N, Hertz-Picciotto I, Kaye WE. (2001). Tetrachloroethylene in drinking water and
birth outcomes at the US Marine Corps Base at Camp Lejeune, North Carolina. Am J Epidemiol
154(10):902-908.
Steinberg AD. (1993) Should chloral hydrate be banned? Pediatrics 92(3):442-446.
Stephens EA, Taylor JA, Kaplan N, Yang CH, Hsieh LL, Lucier GW, Bell DA. (1994) Ethnic
variation in the CYP2E1 gene: polymorphism analysis of 695 African-Americans, European-
Americans and Taiwanese. Pharmacogenetics 4(4): 185-92.
Stewart RD, Hake CL, Peterson JE. (1974) "Degreaser's flush". Arch Environ Health 29(1): 1-5.
Sweeney C, Farrow DC, Schwartz SM, Eaton DL, Checkoway H, Vaughan TL. (2000)
Glutathione S-transferase Ml, Tl, and PI polymorphisms as risk factors for renal cell
carcinoma: a case-control study. Cancer Epidemiol Biomarkers Prev 9(4):449-54.
Taskinen H, Anttila A, Lindbohm ML, Sallmen M, Hemminki K. (1989) Spontaneous abortions
and congenital malformations among the wives of men occupationally exposed to organic
solvents. Scand J Work Environ Health 15(5):345—352.
6/22/2009
961
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Tateishi, T; Nakura, H; Asoh, M; et al. (1997) A comparison of hepatic cytochrome P450 protein
expression between infancy and postinfancy. Life Sciences 61(21):2567-74.
Taylor, DH; Lagory, KE; Zaccaro, DJ; Pfohl, RJ; Laurie, RD. (1985) Effect of trichloroethylene
on the exploratory and locomotor activity of rats exposed during development. Sci Total Environ
47:415-420.
Tola S, Vilhunen R, Jarvinen E, Korkala ML. (1980) A cohort study on workers exposed to
trichloroethylene. J Occup Med 22(11):737-740.
Valic E, Waldhor T, Konnaris C, Michitsch A, Wolf C. (1997) Acquired dyschromatopsia in
combined exposure to solvents and alcohol. Int Arch Occup Environ Health 70(6):403-6.
van Lieshout EM, Peters WH. (1998) Age and gender dependent levels of glutathione and
glutathione S-transferases in human lymphocytes. Carcinogenesis 19(10): 1873-5.
van Lieshout EM, Knapen MF, Lange WP, Steegers EA, Peters WH. (1998) Localization of
glutathione S-transferases alpha and pi in human embryonic tissues at 8 weeks gestational age.
Hum Reprod 13(5): 1380-6.
Verma Y, Rana SV. (2003) Gender differences in the metabolism of benzene, toluene and
trichloroethylene in rat with special reference to certain biochemical parameters. J Environ Biol
24(2): 135-40.
Vesterberg O, Astrand I. (1976) Exposure to trichloroethylene monitored by analysis of
metabolites in blood and urine. J Occup Med 18(4):224-6.
Vesterberg O, Gorczak J, Krasts M. (1976) Exposure to trichloroethylene II. Metabolites in
blood and urine. Scand J Work Environ Health 2(4):212-9.
Vieira I, Sonnier M, Cresteil T. (1996) Developmental expression of CYP2E1 in the human
liver. Hypermethylation control of gene expression during the neonatal period. Eur J Biochem
238(2):476-83.
Wallace LA, Pellizzari E, Hartwell T, Rosenzweig M, Erickson M, Sparacino C, Zelon H. (1984)
Personal exposure to volatile organic compounds. I. Direct measurements in breathing-zone air,
drinking water, food, and exhaled breath. Environ Res 35(1):293—319.
6/22/2009
962
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Wang Z, Hall SD, Maya JF, Li L, Asghar A, Gorski JC. (2003) Diabetes mellitus increases the in
vivo activity of cytochrome P450 2E1 in humans. Br J Clin Pharmacol 55(1):77—85.
Weisel CP, Jo WK. (1996) Ingestion, inhalation, and dermal exposures to chloroform and
trichloroethene from tap water. Environ Health Perspect 104(1):48—51.
Westergren, I; Kjellstrand, P; Linder, LE; Johansson, BB. (1984) Reduction of brain specific
gravity in mice prenatally exposed to trichloroethylene. Toxicology Letters 23:223-226.
White RF, Feldman RG, Eviator, II, Jabre JF, Niles CA. (1997) Hazardous waste and
neurobehavioral effects: a developmental perspective. Environmental research 73(1-2): 113-124.
White JF, Carlson GP. (1981) Epinephrine-induced cardiac arrhythmias in rabbits exposed to
trichloroethylene: potentiation by ethanol. Toxicol Appl Pharmacol 60(3):466-71.
Wiesenhiitter B, Selinski S, Golka K, Briining T, Bolt HM. (2007) Re-assessment of the
influence of polymorphisms of phase-II metabolic enzymes on renal cell cancer risk of
trichloroethylene-exposed workers. Int Arch Occup Environ Health 81:247-251.
Wigle DT, Turner MC, Gomes J, Parent ME. 2008. Role of hormonal and other factors in
human prostate cancer. J Toxicol Environ Health, Part B 11:242-259.
Williams F, Meenagh A, Darke C, Acosta A, Daar AS, Gorodezky C, Hammond M, Nascimento
E, Middleton D. (2001) Analysis of the distribution of HLA-B alleles in populations from five
continents. Hum Immunol 62(6):645-50.
Windemuller FJ, Ettema JH. (1978) Effects of combined exposure to trichloroethylene and
alcohol on mental capacity. Int Arch Occup Environ Health 41(2):77-85.
Windham GC, Zhang L, Gunier R, Croen LA, Grether JK. (2006) Autism spectrum disorders in
relation to distribution of hazardous air pollutants in the San Francisco Bay area. Environ Health
Perspect 114(9): 1438-1444.
Withey JR, Karpinski K. (1985) The fetal distribution of some aliphatic chlorinated
hydrocarbons in the rat after vapor phase exposure. Biol Res Pregnancy Perinatol. 6(2): 79-88.
Wu, KL; Berger, T. (2007) Trichloroethylene metabolism in the rat ovary reduces oocyte
fertilizibility. Chemico-Biological Interactions 170:20-30.
6/22/2009
963
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Wu, KL; Berger, T. (2008) Ovarian gene expression is stable after exposure to trichloroethylene.
Toxicology Letters 177:59-65.
Wu C, Schaum J. (2000) Exposure assessment of trichloroethylene. Environ Health Perspect 108
Suppl 2:359-363.
Yauck JS, Malloy ME, Blair K, Simpson PM, McCarver DG. (2004) Proximity of residence to
trichloroethylene-emitting sites and increased risk of offspring congenital heart defects among
older women. Birth Defects Res A Clin Mol Teratol 70(10):808-814.
Yoon M, Madden MC, Barton HA. (2007) Extrahepatic metabolism by CYP2E1 in PBPK
modeling of lipophilic volatile organic chemicals: impacts on metabolic parameter estimation
and prediction of dose metrics. J Toxicol Environ Health A 70(18):1527-41.
Yuan JM, Castelao JE, Gago-Dominguez M, Yu MC, Ross RK. (1998) Tobacco use in relation
to renal cell carcinoma. Cancer Epidemiol Biomarkers Prev 7(5):429-33.
Yue F, Huang HL, Huang JX, Liang LY, Huang ZL, Wei QY, Chen XM. (2007) HLA-DM
polymorphism and risk of trichloroethylene induced medicamentosa-like dermatitis. Biomed
Environ Sci 20(6):506-l 1.
Zenick H, Blackburn K, Hope E, Richdale N, Smith MK. (1984) Effects of trichloroethylene
exposure on male reproductive function in rats. Toxicology 31(3-4):237~50.
Zielinski A. (1973) General health state of women professionally exposed to trichloroethylene
vapours. Med Pr 24:263-271.
6/22/2009
964
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.10 Hazard Characterization
4.10.1 Characterization of Non-Cancer Effects
4.10.1.1 Neurotoxicity
Both human and animal studies have associated TCE exposure with effects on several
neurological domains. The strongest neurological evidence of hazard in humans is for changes
in trigeminal nerve function or morphology and impairment of vestibular function. Fewer and
more limited evidence exists in humans on delayed motor function, and changes in auditory,
visual, and cognitive function or performance. Acute and subchronic animal studies show
morphological changes in the trigeminal nerve, disruption of the peripheral auditory system
leading to permanent function impairments and histopathology, changes in visual evoked
responses to patterns or flash stimulus, and neurochemical and molecular changes. Additional
acute studies reported structural or functional changes in hippocampus, such as decreased
myelination or decreased excitability of hippocampal CA1 neurons, although the relationship of
these effects to overall cognitive function is not established. Some evidence exists for motor-
related changes in rats/mice exposed acutely/subchronically to TCE, but these effects have not
been reported consistently across all studies.
Epidemiologic evidence supports a relationship between TCE exposure and trigeminal
nerve function changes, with multiple studies in different populations reporting abnormalities in
trigeminal nerve function in association with TCE exposure (Barret et al., 1982, 1984, 1987;
Feldman et al., 1988, 1992; Kilburn and Warshaw, 1993; Ruitjen et al., 2001; Kilburn, 2002a;
Mhiri et al., 2004). Of these, two well conducted occupational cohort studies, each including
more than 100 TCE-exposed workers without apparent confounding from multiple solvent
exposures, additionally reported statistically significant dose-response trends based on ambient
TCE concentrations, duration of exposure, and/or urinary concentrations of the TCE metabolite
TCA (Barret et al., 1984; Barret et al., 1987). Limited additional support is provided by a
positive relationship between prevalence of abnormal trigeminal nerve or sensory function and
cumulative exposure to TCE (most subjects) or CFC-113 (<25% of subjects) (Rasmussen et al.,
1993c). Test for linear trend in this study was not statistically significant and may reflect
exposure misclassification since some subjects included in this study did not have TCE exposure.
The lack of association between TCE exposure and overall nerve function in three small studies
(trigeminal: El-Ghawabi et al., 1973; ulnar and medial: Triebig et al., 1982, 1983) does not
provide substantial evidence against a causal relationship between TCE exposure and trigeminal
nerve impairment because of limitations in statistical power, the possibility of exposure
6/22/2009
965
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
misclassification, and differences in measurement methods. Laboratory animal studies have also
shown TCE-induced changes in the trigeminal nerve. Although one study reported no significant
changes in trigeminal somatosensory evoked potential in rats exposed to TCE for 13 weeks
(Albee et al., 2006), there is evidence of morphological changes in the trigeminal nerve
following short-term exposures in rats (Barret et al., 1991, 1992).
Human chamber, occupational, geographic based/drinking water, and laboratory animal
studies clearly established TCE exposure causes transient impairment of vestibular function.
Subjective symptoms such as headaches, dizziness, and nausea resulting from occupational
(Granjean et al., 1955; Liu et al., 1988; Rasmussen and Sabroe, 1986; Smith et al., 1970),
environmental (Hirsch et al., 1996), or chamber exposures (Stewart et al., 1970; Smith et al.,
1970) have been reported extensively. A few laboratory animal studies have investigated
vestibular function, either by promoting nystagmus or by evaluating balance (Niklasson et al.,
1993; Tham et al., 1979; Tham et al., 1984; Umezu et al., 1997).
In addition, mood disturbances have been reported in a number of studies, although these
effects also tend to be subjective and difficult to quantify (Gash et al., 2007; Kilburn and
Warshaw, 1993; Kilburn, 2002a, 2002b; McCunney et al., 1988; Mitchell et al., 1969;
Rasmussen and Sabroe, 1986; Troster and Ruff, 1990), and a few studies have reported no
effects from TCE on mood (Reif et al., 2003; Triebig et al., 1976, 1977a). Few comparable
mood studies are available in laboratory animals, although both Moser et al. (2003) and Albee et
al. (2006) report increases in handling reactivity among rats exposed to TCE. Finally,
significantly increased number of sleep hours was reported by Arito et al. (1994) in rats exposed
via inhalation to 50-300 ppm TCE for 8 hr/d for 6 weeks.
Four epidemiologic studies of chronic exposure to TCE observed disruption of auditory
function. One large occupational cohort study showed a statistically significant difference in
auditory function with cumulative exposure to TCE or CFC-113 as compared to control groups
after adjustment for possible confounders, as well as a positive relationship between auditory
function and increasing cumulative exposure (Rasmussen et al., 1993b). Of the three studies
based on populations from ATSDR's TCE Subregistry from the National Exposure Registry,
more limited than Rasmussen et al. (1993b) due to inferior exposure assessment, Burg et al.
(1995) and Burg and Gist (1999) reported a higher prevalence of self-reported hearing
impairments. The third study reported that auditory screening revealed abnormal middle ear
function in children less than 10 years of age, although a dose-response relationship could not be
established and other tests did not reveal differences in auditory function (ATSDR, 2003a).
Further evidence for these effects is provided by numerous laboratory animal studies
demonstrating that high dose subacute and subchronic TCE exposure in rats disrupts the auditory
system leading to permanent functional impairments and histopathology.
6/22/2009
966
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Studies in humans exposed under a variety of conditions, both acutely and chronically,
report impaired visual functions such as color discrimination, visuospatial learning tasks, and
visual depth perception in subjects with TCE exposure. Abnormalities in visual depth perception
were observed with a high acute exposure to TCE under controlled conditions (Vernon and
Ferguson, 1969). Studies of lower TCE exposure concentrations also observed visuofunction
effects. One occupational study (Rasmussen et al., 1993b) reported a statistically significant
positive relationship between cumulative exposure to TCE or CFC-113 and visual gestalts
learning and retention among Danish degreasers. Two studies of populations living in a
community with drinking water containing TCE and other solvents furthermore suggested
changes in visual function (Kilburn et al., 2002a; Reif et al., 2003). These studies used more
direct measures of visual function as compared to Rasmussen et al. (1993b), but their exposure
assessment is more limited because TCE exposure is not assigned to individual subjects (Kilburn
et al., 2002a), or because there are questions regarding control selection (Kilburn et al., 2002a)
and exposure to several solvents (Kilburn et al., 2002a; Reif et al., 2003).
Additional evidence of effects of TCE exposure on visual function is provided by a
number of laboratory animal studies demonstrating that acute or subchronic TCE exposure
causes changes in visual evoked responses to patterns or flash stimulus (Boyes et al., 2003, 2005;
Blain et al., 1994). Animal studies have also reported that the degree of some effects is
correlated with simultaneous brain TCE concentrations (Boyes et al., 2003, 2005) and that, after
a recovery period, visual effects return to control levels (Blain et al., 1994; Rebert et al., 1991).
Overall, the human and laboratory animal data together suggest that TCE exposure can cause
impairment of visual function, and some animal studies suggest that some of these effects may
be reversible with termination of exposure.
Studies of human subjects exposed to TCE either acutely in chamber studies or
chronically in occupational settings have observed deficits in cognition. Five chamber studies
reported statistically significant deficits in cognitive performance measures or outcome measures
suggestive of cognitive effects (Stewart et al., 1970; Gamberale et al., 1976; Triebig et al., 1976,
1977a; Gamberale et al., 1977). Danish degreasers with high cumulative exposure to TCE or
CFC-113 had a high risk [OR = 13.7, 95% CI; 2.0-92.0] for psychoorganic syndrome
characterized by cognitive impairment, personality changes, and reduced motivation, vigilance,
and initiative compared to workers with low cumulative exposure. Studies of populations living
in a community with contaminated groundwater also reported cognitive impairments (Kilburn
and Warshaw, 1993; Kilburn, 2002a), although these studies carry less weight in the analysis
because TCE exposure is not assigned to individual subjects and their methodological design is
weaker.
6/22/2009
967
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Laboratory studies provide some additional evidence for the potential for TCE to affect
cognition, although the predominant effect reported has been changes in the time needed to
complete a task, rather than impairment of actual learning and memory function (Kulig et al.,
1987; Kishi et al., 1993; Umezu et al., 1997). In addition, in laboratory animals, it can be
difficult to distinguish cognitive changes from motor-related changes. However, several studies
have reported structural or functional changes in the hippocampus, such as decreased
myelination (Issacson et al., 1990; Isaacson and Taylor, 1989) or decreased excitability of
hippocampal CA1 neurons (Ohta et al., 2001), although the relationship of these effects to
overall cognitive function is not established.
Two studies of TCE exposure, one chamber study of acute exposure duration and one
occupational study of chronic duration, reported changes in psychomotor responses. The
chamber study of Gamberale et al. (1976) reported a dose-related decrease in performance in a
choice reaction time test in healthy volunteers exposed to 100 and 200 ppm TCE for 70 minutes
as compared to the same subjects without exposure. Rasmussen et al. (1993c) reported a
statistically significant association with cumulative exposure to TCE or CFC-113 and
dyscoordination trend among Danish degreasers. Observations in a third study (Gun et al., 1978)
are difficult to judge given the author's lack of statistical treatment of data. In addition, Gash et
al. (2007) reported that 14 out of 30 TCE-exposed workers exhibited significantly slower fine
motor hand movements as measured through a movement analysis panel test. Studies of
population living in communities with TCE and other solvents detected in groundwater supplies
reported significant delays in simple and choice reaction times in individuals exposed to TCE in
contaminated groundwater as compared to referent groups (Kilburn, 2002a; Kilburn and
Warshaw, 1993; Kilburn and Thornton, 1996). Observations in these studies are more uncertain
given questions of the representativeness of the referent population, lack of exposure assessment
to individual study subjects, and inability to control for possible confounders including alcohol
consumption and motivation. Finally, in a presentation of 2 case reports, decrements in motor
skills as measured by the grooved pegboard and finger tapping tests were observed (Troster and
Ruff, 1990).
Laboratory animal studies of acute or sub-chronic exposure to TCE observed
psychomotor effects, such as loss of righting reflex (Umezu et al., 1997; Shih et al., 2001) and
decrements in activity, sensory-motor function, and neuromuscular function (Kishi et al., 1993;
Moser et al., 1995; Moser et al., 2003). However, two studies also noted an absence of
significant changes in some measures of psychomotor function (Kulig et al., 1987; Albee et al.,
2006). In addition, less consistent results have been reported with respect to locomotor activity
in rodents. Some studies have reported increased locomotor activity after an acute i.p. dosage
(Wolff and Siegmund, 1978) or decreased activity after acute or short term oral gavage dosing
6/22/2009
968
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(Moser et al., 1995, 2003). No change in activity was observed following exposure through
drinking water (Waseem et al., 2001), inhalation (Kulig et al., 1987) or orally during the
neurodevelopment period (Fredriksson et al., 1993).
Several neurochemical and molecular changes have been reported in laboratory
investigations of TCE toxicity. Kjellstrand et al. (1987) reported inhibition of sciatic nerve
regeneration in mice and rats exposed continuously to 150 ppm TCE via inhalation for 24 days.
Two studies have reported changes in GABAergic and glutamatergic neurons in terms of GABA
or glutamate uptake (Briving et al., 1986) or response to GABAergic antagonistic drugs (Shih et
al., 2001) as a result of TCE exposure, with the Briving et al. (1986) conducted at 50 ppm for 12
months. Although the functional consequences of these changes is unclear, Tham et al. (1979,
1984) described central vestibular system impairments as a result of TCE exposure that may be
related to altered GABAergic function. In addition, several in vitro studies have demonstrated
that TCE exposure alters the function of inhibitory ion channels such as receptors for GAB Aa
glycine, and serotonin (Krasowski and Harrison, 2000; Beckstead et al., 2000; Lopreato et al.,
2003) or of voltage-sensitive calcium channels (Shafer et al., 2005).
4.10.1.2 Kidney toxicity
There are few human data pertaining to TCE-related non-cancer kidney toxicity.
Observation of elevated excretion of urinary proteins in the available studies (Rasmussen et al.,
1993a; Briining et al., 1999a, b; Bolt et al., 2004; Green et al., 2004) indicates the occurrence of
a toxic insult among TCE-exposed subjects compared to unexposed controls. Two studies are of
subjects with previously diagnosed kidney cancer (Briining et al., 1999a; Bolt et al., 2004), while
subjects in the other studies are disease free. Urinary proteins are considered nonspecific
markers of nephrotoxicity and include a 1-microglobulin, albumin, and NAG (Price et al., 1996;
Lybarger et al., 1999; Price et al., 1999). Four studies measure al-microglobulin with elevated
excretion observed in the German studies (Briining et al., 1999a, b; Bolt et al., 2004) but not
Green et al. (2004). However, Rasmussen et al. (1993a) reported a positive relationship between
increasing urinary NAG, another nonspecific marker of tubular toxicity, and increasing exposure
duration; and Green et al. (2004) found statistically significant group mean differences in NAG.
Observations in Green et al. (2004) provide evidence of tubular damage among workers exposed
to trichloroethylene at current occupational levels. Elevated excretion of NAG has also been
observed with acute TCE poisoning (Carried et al., 2007). Some support for TCE nephrotoxicity
in humans is provided by a study of end-stage renal disease in a cohort of workers at Hill Air
Force Base (Radican et al., 2006), although subjects in this study were exposed to hydrocarbons,
JP-4 gasoline, and solvents in addition to TCE, including 1,1,1-trichloroethane.
6/22/2009
969
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Laboratory animal and in vitro data provide additional support for TCE nephrotoxicity.
Multiple studies with both gavage and inhalation exposure show that TCE causes renal toxicity
in the form of cytomegaly and karyomegaly of the renal tubules in male and female rats and
mice (summarized in section 4.3.4). Further studies with TCE metabolites have demonstrated a
potential role for DCVC, TCOH, and TCA in TCE-induced nephrotoxicity. Of these, available
data suggest that DCVC induced renal effects most like those of TCE and is formed in sufficient
amounts following TCE exposure to account for these effects. TCE or DCVC have also been
shown to be cytotoxic to primary cultures of rat and human renal tubular cells (Cummings et al.,
2000a, b; Cummings and Lash, 2000).
Overall, multiple lines of evidence support the conclusion that TCE causes nephrotoxicity
in the form of tubular toxicity, mediated predominantly through the TCE GSH conjugation
product DCVC.
4.10.1.3 Liver toxicity
Few studies on liver toxicity and TCE exposure are found in humans. Of these, three
studies reported significant changes in serum liver function tests, widely used in clinical settings
in part to identify patients with liver disease, in metal degreasers whose TCE exposure was
assessed using urinary trichloro-compounds as a biomarker (Nagaya et al., 1993; Rasmussen et
al., 1993; Xu et al., 2009). Two additional studies reported plasma or serum bile acid changes
(Neghab et al., 1997; Driscoll et al., 1992). One study of subjects from the TCE subregistry of
ATSDR's National Exposure Registry is suggestive of liver disorders but limitations preclude
inferences whether TCE caused these conditions is not possible given the study's limitations
(Davis et al., 2005). Furthermore, a number of case reports exist of liver toxicity including
hepatitis accompanying immune-related generalized skin diseases described as a variation of
erythema multiforme, Stevens-Johnson syndrome, toxic epidermal necrolysis patients, and
hypersensitivity syndrome (Kamijima et al., 2007) in addition to jaundice, hepatomegaly,
hepatosplenomegaly, and liver failure TCE-exposed workers (Thiele, 1982; Huang et al., 2002).
Cohort studies have examined cirrhosis mortality and either TCE exposure (Morgan et al., 1998;
Boice et al., 1999, 2006; Garabrant et al., 1988; Blair et al., 1998; Ritz et al., 1999; ATSDR,
2004; Radican et al., 2008) or solvent exposure (Leigh and Jiang, 1993), but are greatly limited
by their use of death certificates where there is a high degree (up to 50%) of underreporting
(Blake et al., 1988), so these null findings do not rule out an effect of TCE on cirrhosis. Overall,
while there some evidence exists of liver toxicity as assessed from liver function tests, the data
are inadequate for making conclusions regarding causality.
In laboratory animals, TCE exposure is associated with a wide array of hepatotoxic
endpoints. Like humans, laboratory animals exposed to TCE have been observed to have
6/22/2009
970
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
increased serum bile acids (Bai et al., 1992b; Neghab et al., 1997), although the toxicologic
importance of this effect is unclear. Most other effects in laboratory animals have not been
studied in humans, but nonetheless provide evidence that TCE exposure leads to hepatotoxicity.
These effects include increased liver weight, small transient increases in DNA synthesis,
cytomegaly in the form of "swollen" or enlarged hepatocytes, increased nuclear size probably
reflecting polyploidization, and proliferation of peroxisomes. Liver weight increases
proportional to TCE dose are consistently reported across numerous studies and appear to be
accompanied by periportal hepatocellular hypertrophy (Nunes et al., 2001; Tao et al., 2000,
Tucker et al., 1982; Goldsworthy and Popp, 1987; Elcombe et al., 1985; Dees and Travis, 1993;
Nakajima et al., 2000; Berman et al., 1995; Melnick et al., 1987; Laughter et al., 2004; Merrick
et al., 1989; Goel et al., 1992; Kjellstand et al., 1981, 1983a, b; Buben and O'Flaherty, 1985).
There is also evidence of increased DNA synthesis in a small portion of hepatocytes at around 10
days in vivo exposure (Mirsalis et al., 1989; Elcombe et al., 1985; Dees and Travis, 1993;
Channel et al., 1998). The lack of correlation of hepatocellular mitotic figures with whole liver
DNA synthesis or DNA synthesis observed in individual hepatocytes (Elcombe et al., 1985;
Dees and Travis, 1993) supports the conclusions that cellular proliferation is not the predominant
cause of increased DNA synthesis and that nonparenchymal cells may also contribute to such
synthesis. Indeed, nonparenchymal cell activation or proliferation has been noted in several
studies (Kjellstrand et al., 1983b; Goel et al., 1992). Moreover, the histological descriptions of
TCE-exposed livers are consistent with and, in some cases, specifically note increased
polyploidy (Buben and O'Flaherty, 1985). Interestingly, changes in TCE-induced hepatocellular
ploidy, as indicated by histological changes in nuclei, have been noted to remain after the
cessation of exposure (Kjellstrand et al., 1983a). In regard to apoptosis, TCE has been reported
either to have no effect or to cause a slight increase at high doses (Dees and Travis, 1993;
Channel et al., 1998). Some studies have also noted effects from dosing vehicle alone (such as
corn oil, in particular) not only on liver pathology, but also on DNA synthesis (Merrick et al.,
1989; Channel et al., 1998). Available data also suggest that TCE does not induce substantial
cytotoxicity, necrosis, or regenerative hyperplasia, as only isolated, focal necroses and mild to
moderate changes in serum and liver enzyme toxicity markers having been reported (Elcombe et
al., 1985; Dees and Travis, 1993; Channel et al., 1998). Data on peroxisome proliferation, along
with increases in a number of associated biochemical markers, show effects in both mice and rats
(Elcombe et al., 1985; Channel et al., 1998; Goldsworthy and Popp, 1987). These effects are
consistently observed across rodent species and strains, although the degree of response at a
given mg/kg/d dose appears to be highly variability across strains, with mice on average
appearing to be more sensitive.
6/22/2009
971
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
While it is likely that oxidative metabolism is necessary for TCE-induced effects in the
liver, the specific metabolite or metabolites responsible is less clear. TCE, TCA and DCA
exposures have all been associated with induction of changes in liver weight, DNA synthesis,
and peroxisomal enzymes. The available data strongly support TCA not being the sole or
predominant active moiety for TCE-induced liver effects, particularly with respect to
hepatomegaly. In particular, TCE and TCA dose-response relationships are quantitatively
inconsistent, for TCE leads to greater increases in liver/body weight ratios that expected from
predicted rates of TCA production (see analysis in section 4.4.6.2.1). In fact, above a certain
dose of TCE, liver/body weight ratios are greater than that observed under any conditions studied
so far for TCA. Histological changes and effects on DNA synthesis are generally consistent with
contributions from either TCA or DCA, with a degree of polyploidization, rather than cell
proliferation, likely to be significant for TCE, TCA, and DCA.
Overall, TCE, likely through its oxidative metabolites, clearly leads to liver toxicity in
laboratory animals, with mice appearing to be more sensitive than other laboratory animal
species, but there is only limited epidemiologic evidence of hepatotoxicity being associated with
TCE exposure.
4.10.1.4 Immunotoxicity
Studies in humans provide evidence of associations between TCE exposure and a number
of immunotoxicological endpoints. The relation between systemic autoimmune diseases, such as
scleroderma, and occupational exposure to TCE has been reported in several recent studies. A
meta-analysis of scleroderma studies (Diot et al., 2002; Garabrant et al., 2003; Nietert et al.,
1998) conducted by the EPA resulted in a statistically significant combined odds ratio for any
exposure in men (OR = 2.5, 95% CI 1.1, 5.4), with a lower relative risk seen in women (OR =
1.2, 95% CI 0.58, 2.6). The incidence of systemic sclerosis among men is very low
(approximately 1 per 100,000 per year), and is approximately 10 times lower than the rate seen
in women (Cooper and Stroehla, 2003). Thus the human data at this time do not allow
determination of whether the difference in effect estimates between men and women reflects the
relatively low background risk of scleroderma in men, gender-related differences in exposure
prevalence or in the reliability of exposure assessment (Messing et al., 2003), a gender-related
difference in susceptibility to the effects of TCE, or chance. Changes in levels of inflammatory
cytokines were reported in an occupational study of degreasers exposed to TCE (Iavicoli et al.,
2005) and a study of infants exposed to TCE via indoor air (Lehmann et al., 2001, 2002).
Experimental studies provide additional support for these effects. Numerous studies have
demonstrated accelerated autoimmune responses in autoimmune-prone mice (Cai et al., 2008;
6/22/2009
972
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Blossom et al., 2007, 2004; Griffin et al., 2000a, b). With shorter exposure periods, effects
include changes in cytokine levels similar to those reported in human studies. More severe
effects, including autoimmune hepatitis, inflammatory skin lesions, and alopecia, were manifest
at longer exposure periods, and interestingly, these effects differ somewhat from the "normal"
expression in these mice. Immunotoxic effects, including increases in anti-ds DNA antibodies in
adult animals, decreased thymus weights, and decreased plaque forming cell response with
prenatal and neonatal exposure, have been also reported in B6C3F1 mice, which do not have a
known particular susceptibility to autoimmune disease (Gilkeson et al., 2004; Keil et al., 2009;
Peden-Adams et al., 2006). Recent mechanistic studies have focused on the roles of various
measures of oxidative stress in the induction of these effects by TCE (Wang et al., 2008, 2007b).
There have been a large number of case reports of a severe hypersensitivity skin disorder,
distinct from contact dermatitis and often accompanied by hepatitis, associated with occupational
exposure to TCE, with prevalences as high as 13% of workers in the same location (Kamijima et
al., 2008, 2007). Evidence of a treatment-related increase in delayed hypersensitivity response
accompanied by hepatic damage has been observed in guinea pigs following intradermal
injection (Tang et al., 2008, 2002), and hypersensitivity response was also seen in mice exposed
via drinking water pre- and post-natally (gestation day 0 through to 8 weeks of age) (Peden-
Adams et al., 2006).
Human data pertaining to TCE-related immunosuppression resulting in an increased risk
of infectious diseases is limited to the report of an association between reported history of
bacteria of viral infections in Woburn, Massachusetts (Lagakos, 1986). Evidence of localized
immunosuppression, as measured by pulmonary response to bacterial challenge (i.e., risk of
Streptococcal pneumonia-related mortality and clearance of Klebsiella bacteria) was seen in an
acute exposure study in CD-I mice (Aranyi et al., 1986). A 4-week inhalation exposure in
Sprague-Dawley rats reported a decrease in plaque forming cell response at exposures of
1,000 ppm (Woolhiser et al., 2006).
Overall, the human and animal studies of TCE and immune-related effects provide strong
evidence for a role of TCE in autoimmune disease and in a specific type of generalized
hypersensitivity syndrome, while there are less data pertaining to immunosuppressive effects.
4.10.1.5 Respiratory tract toxicity
There are very limited human data on pulmonary toxicity and TCE exposure. Two recent
reports of a study of gun manufacturing workers reported asthma-related symptoms and lung
function decrements associated with solvent exposure (Cakmak et al., 2004; Saygun et al., 2007),
but these studies are limited by multiple solvent exposures and the significant effect of smoking
6/22/2009
973
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
on pulmonary function. Laboratory studies in mice and rats have shown toxicity in the bronchial
epithelium, primarily in Clara cells, following acute exposures to TCE by inhalation (Section
4.6.2.1.1). A few studies of longer duration have reported more generalized toxicity, such as
pulmonary fibrosis 90 days after a single 2,000 mg/kg ip dose in mice and pulmonary vasculitis
after 13-week oral gavage exposures to 2,000 mg/kg-d in rats (Forkert and Forkert, 1994. NTP,
1990). However, respiratory tract effects were not reported in other longer-term studies. Acute
pulmonary toxicity appears to be dependent on oxidative metabolism, although the particular
active moiety is not known. While earlier studies implicated chloral produced in situ by P450
enzymes in respiratory tract tissue was responsible for toxicity (reviewed in Green, 2000), the
evidence is inconsistent, and several other possibilities are viable. First, substantial
"accumulation" of chloral is unlikely, as it is likely either to be rapidly converted to TCOH in
respiratory tract tissue or to diffuse rapidly into blood and be converted to TCOH in erythrocytes
or the liver. Conversely, a role for systemically produced oxidative metabolites cannot be
discounted, as CH and TCOH in blood have both been reported following inhalation dosing in
mice. In addition, a recent study reported dichloroacetyl chloride protein adducts in the lungs of
mice to which TCE was administered by ip injection, suggesting dichloroacetyl chloride, which
is not believed to be derived from chloral, may also contribute to TCE respiratory toxicity.
Although humans appear to have lower overall capacity for enzymatic oxidation in the lung
relative to mice, P450 enzymes do reside in human respiratory tract tissue, suggesting that,
qualitatively, the respiratory tract toxicity observed in rodents is biologically plausible in
humans. However, quantitative estimates of differential sensitivity across species due to
respiratory metabolism are highly uncertain due to limited data. Therefore, overall, data are
suggestive of TCE causing respiratory tract toxicity, based primarily on short-term studies in
mice and rats, and no data suggest that such hazards would be biologically precluded in humans.
4.10.1.6 Reproductive toxicity
Reproductive toxicity related to TCE exposure has been evaluated in human and
experimental animal studies for effects in males and females. Only a limited number of studies
have examined whether TCE causes female reproductive toxicity. Epidemiologic studies have
identified possible associations of TCE exposure with effects on female fertility (Sallmen et al.,
1995; ATSDR, 2001) and with menstrual cycle disturbances (ATSDR, 2001; Bardodej and
Vyskocil, 1956; Sagawa et al., 1973; Zielinski, 1973). Reduced in vitro oocyte fertilizability has
been reported as a result of TCE exposure in rats (Berger and Horner, 2003; Wu and Berger,
2007), but a number of other laboratory animal studies did not report adverse effects on female
reproductive function (Cosby and Dukelow, 1992; George et al., 1985, 1986; Manson et al.,
6/22/2009
974
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1984). Overall, there are inadequate data to conclude whether adverse effects on human female
reproduction are caused by TCE.
By contrast, a number of human and laboratory animal studies suggest that TCE exposure
has the potential for male reproductive toxicity. In particular, human studies have reported TCE
exposure to be associated, in several cases statistically-significantly, with increased sperm
density and decreased sperm quality (Chia et al., 1996; Rasmussen et al., 1988), altered sexual
drive or function (El Gawabi et al., 1973; Saihan et al., 1978; Bardodej and Vyskocil, 1956), or
altered serum endocrine levels (Chia et al., 1997; Goh et al., 1998). In addition, three studies
that reported measures of fertility did not or could not report changes associated with TCE
exposure (ATSDR, 2001; Forkert et al., 2003; Sallmen et al., 1998), although the statistical
power of these studies is quite limited. Further evidence of similar effects is provided by several
laboratory animal studies that reported effects on sperm (Kumar et al., 2000a, b, 2001; George et
al., 1985; Land et al., 1981; Veeramachaneni et al., 2001), libido/copulatory behavior (George et
al., 1986; Zenick et al., 1984; Veeramachaneni et al., 2001), and serum hormone levels (Kumar
et al., 2000b; Veeramachaneni et al., 2001). As with the human database, some studies that
assessed sperm measures did not report treatment-related alterations (Cosby and Dukelow, 1992;
Xu et al., 2004; Zenick et al., 1984; George et al., 1986). Additional adverse effects on male
reproduction have also been reported, including histopathological lesions in the testes or
epididymides (George et al., 1986; Kumar et al., 2000a, 2001; Forkert et al., 2002; Kan et al.,
2007) and altered in vitro sperm-oocyte binding or in vivo fertilization due to TCE or metabolites
(Xu et al., 2004; DuTeaux et al., 2004b). While reduced fertility in rodents was only observed in
one study (George et al., 1986), this is not surprising given the redundancy and efficiency of
rodent reproductive capabilities. Furthermore, while George et al. (1986) proposed that the
adverse male reproductive outcomes observed in rats were due to systemic toxicity, the database
as a whole suggests that TCE does induce reproductive toxicity independent of systemic effects.
Therefore, overall, the human and laboratory animal data together support the conclusion that
TCE exposure poses a potential hazard to the male reproductive system.
4.10.1.7 Developmental toxicity
The relationship between TCE exposure (direct or parental) and adverse developmental
outcomes has been investigated in a number of epidemiologic and laboratory animal studies.
Prenatal effects examined include death (spontaneous abortion, perinatal death, pre- or post-
implantation loss, resorptions), decreased growth (low birth weight, small for gestational age,
intrauterine growth restriction, decreased postnatal growth), and congenital malformations, in
particular eye and cardiac defects. Postnatal developmental outcomes examined include growth
6/22/2009
975
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
and survival, developmental neurotoxicity, developmental immunotoxicity, and childhood
cancers.
A few epidemiological studies have reported associations between parental exposure to
TCE and spontaneous abortion or perinatal death (Taskinen et al., 1994; Windham et al., 1991;
ATSDR, 2001), although other studies reported mixed or null findings (ATSDR, 2006, 2008;
Bove, 1996; Bove et al., 1995; Goldberg et al., 1990; Lagakos et al., 1986; Lindbohm et al.,
1990; Taskinen et al., 1989). Studies examining associations between TCE exposure and
decreased birth weight or small for gestational age have reported small, often non-statistically
significant, increases in risk for these effects (ATSDR, 1998, 2006, 2008; Windham et al., 1991).
However, other studies observed mixed or no association (Bove, 1996; Bove et al., 1995;
Lagakos et al., 1986; Rodenbeck et al., 2000). While comprising both occupational and
environmental exposures, these studies are overall not highly informative due to their small
numbers of cases and limited exposure characterization or to the fact that exposures to mixed
solvents were involved. However, a number of laboratory animal studies show analogous effects
of TCE exposure in rodents. In particular, pre- or post-implantation losses, increased
resorptions, perinatal death, and decreased birth weight have been reported in multiple well-
conducted studies in rats and mice (Healy et al., 1982; Kumar et al., 2000a; George et al., 1985,
1986; Narotsky et al., 1995; Narotsky and Kavlock, 1995). Interestingly, the rat studies
reporting these effects used Fischer 344 or Wistar rats, while several other studies, all of which
used Sprague-Dawley rats, reported no increased risk in these developmental measures (Carney
et al., 2006; Hardin et al., 1981; Schwetz et al., 1975). Overall, based on weakly suggestive
epidemiologic data and fairly consistent laboratory animal data, it can be concluded that TCE
exposure poses a potential hazard for prenatal losses and decreased growth or birth weight of
offspring.
Epidemiologic data provide some support for the possible relationship between maternal
TCE exposure and birth defects in offspring, in particular cardiac defects. Other developmental
outcomes observed in epidemiology and experimental animal studies include an increase in total
birth defects (ADHS, 1988; ATSDR, 2001), CNS defects (ATSDR, 2001; Bove, 1996; Bove et
al., 1995; Lagakos et al., 1986), oral cleft defects (Bove, 1996; Bove et al., 1995; Lagakos et al.,
1986; Lorente et al., 2000), eye/ear defects (Lagakos et al., 1986; Narotsky et al., 1995; Narotsky
and Kavlock, 1995), kidney/urinary tract disorders (Lagakos et al., 1986), musculoskeletal birth
anomalies (Lagakos et al., 1986), lung/respiratory tract disorders (Lagakos et al., 1986; Das and
Scott, 1994), and skeletal defects (Healy et al., 1982). Occupational cohort studies, while not
consistently reporting positive results, are generally limited by the small number of observed or
expected cases of birth defects (Lorente et al., 2000; Tola et al., 1980; Taskinen et al., 1989).
6/22/2009
976
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
While only one of the epidemiological studies specifically reported observations of eye
anomalies (Lagakos et al., 1986), studies in rats have identified increases in the incidence of fetal
eye defects following oral exposures during the period of organogenesis with TCE (Narotsky et
al., 1995; Narotsky and Kavlock, 1995) or its oxidative metabolites DCA and TCA (Smith et al.,
1989, 1992; Warren et al., 2006). No other developmental or reproductive toxicity studies
identified abnormalities of eye development following TCE exposures, which may have been
related to the administered dose or other aspects of study design (e.g., level of detail applied to
fetal ocular evaluation). Overall, the study evidence suggests a potential for the disruption of
ocular development by exposure to TCE and its oxidative metabolites.
The epidemiological studies, while individually limited, as a whole show relatively
consistent elevations, some of which were statistically significant, in the incidence of cardiac
effects in TCE-exposed populations compared to reference groups (ATSDR, 2001, 2006, 2008;
Bove et al., 1995; Bove, 1996; Goldberg et al., 1990; Yauck et al., 2004). Interestingly,
Goldberg et al. (1990) noted that the odds ratio for congenital heart disease in offspring declined
from 3-fold to no difference as compared to controls after TCE-contaminated drinking water
wells were closed, suggestive of a causal relationship. However, this study reported no
significant differences in cardiac lesions between exposed and non-exposed groups (Goldberg et
al., 1990). One additional community study reported that, among the 5 cases of cardiovascular
anomalies, there was no significant association with TCE (Lagakos et al., 1986), but due to the
small number of cases this does not support an absence of effect. In laboratory animal models,
avian studies were the first to identify adverse effects of TCE exposure on cardiac development,
and the initial findings have been confirmed multiple times (Bross et al., 1983; Loeber et al.,
1988; Boyer et al., 2000; Drake et al., 2006a, b; Mishima et al., 2006; Rufer et al., 2008).
Additionally, administration of TCE and TCE metabolites TCA and DCA in maternal drinking
water during gestation has been reported to induce cardiac malformations in rat fetuses (Dawson
et al., 1990, 1993; Johnson et al., 1998a, b, 2003, 2005; Smith et al., 1989, 1992; Epstein et al.,
1992). However, it is notable that a number of other studies, several of which were well
conducted, did not report induction of cardiac defects in rats or rabbits from TCE administered
by inhalation (Dorfmueller et al., 1979; Schwetz et al., 1975; Hardin et al., 1981; Healy et al.,
1982; Carney et al., 2006) or in rats and mice by gavage (Cosby and Dukelow, 1992; Narotsky et
al., 1995; Narotsky and Kavlock, 1995; Fisher et al., 2001).
The potential importance of these effects warrants a more detailed discussion of possible
explanations for the apparent inconsistencies in the laboratory animal studies. Many of the
studies that did not identify cardiac anomalies used a traditional free-hand section technique on
fixed fetal specimens (Dorfmueller et al., 1979; Schwetz et al., 1975; Hardin et al., 1981; Healy
et al., 1982). Detection of such anomalies can be enhanced through the use of a fresh dissection
6/22/2009
977
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
technique as described by Staples (1974) and Stuckhardt and Poppe (1984), and this was the
technique used in the study by Dawson et al. (1990), with further refinement of the technique
used in the positive studies by Dawson et al. (1993) and Johnson et al. (2003, 2005). However,
two studies that used the same or similar fresh dissection technique did not report cardiac
anomalies (Fisher et al., 2001; Carney et al., 2006), although it has been suggested that
differences in experimental design (e.g., inhalation versus gavage versus drinking water route of
administration, exposure during organogenesis versus the entire gestational period, or varied
dissection or evaluation procedures) may have been contributing factors to the differences in
observed response. A number of other limitations in the studies by Dawson et al. (1993) and
Johnson et al. (2003, 2005) have been suggested (Hardin et al., 2005; Watson et al., 2006). One
concern is the lack of clear dose-response relationship for the incidence of any specific cardiac
anomaly or combination of anomalies, a disparity for which no reasonable explanation has been
put forth. In addition, analyses on a fetal- rather than litter-basis and the pooling of data
collected over an extended period, including non-concurrent controls, have been criticized. With
respect to the first issue, the study authors provided individual litter incidence data to USEPA for
analysis (see Section 6, dose-response), and, in response to the second issue, the study authors
provided further explanation as to their experimental procedures (Johnson et al., 2004). In sum,
while the studies by Dawson et al. (1993) and Johnson et al. (2003, 2005) have significant
limitations, there is insufficient reason to dismiss their findings.
Finally, mechanistic studies, particularly based on the avian studies mentioned above,
provide additional support for TCE-induced fetal cardiac malformation, particularly with respect
to defects involving septal and valvular morphogenesis. As summarized by NRC (2006), there is
substantial concordance in the stages and events of cardiac valve formation between mammals
and birds. While quantitative extrapolation of findings from avian studies to humans is not
possible without appropriate kinetic data for these experimental systems, the treatment-related
alterations in endothelial cushion development observed in avian in ovo and in vitro studies
(Boyer et al., 2000; Mishima et al., 2006; Ou et al., 2003) provide a plausible mechanistic basis
for defects in septal and valvular morphogenesis observed in rodents, and consequently support
the plausibility of cardiac defects induced by TCE in humans.
Postnatal developmental outcomes examined after TCE prenatal and/or postnatal
exposure in both humans and experimental animals include developmental neurotoxicity,
developmental immunotoxicity, and childhood cancer. Effects on the developing nervous
system included a broad array of structural and behavioral alterations in humans (White et al.,
1997; Windham et al., 2006; Burg et al., 1995; Burg and Gist, 1997; Bernad et al., 1987;
Laslo-Baker et al., 2004; Till et al., 2001; Beppu, 1968; ATSDR, 2003a) and animals
(Fredriksson et al., 1993; George et al., 1986; Isaacson and Taylor, 1989; Narotsky and Kavlock,
6/22/2009
978
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1995; Noland-Gerbec et al., 1986; Taylor et al., 1985; Westergren et al., 1984; Blossom et al.,
2008). Adverse immunological findings in humans following developmental exposures to TCE
were reported by Lehmann et al. (2002) and Byers et al. (1988). In mice, alterations in T-cell
subpopulations, spleen and/or thymic cellularity, cytokine production, autoantibody levels (in an
autoimmune-prone mouse strain), and/or hypersensitivity response were observed after
exposures during development (Blossom and Doss, 2007; Blossom et al., 2008; Peden-Adams et
al., 2006, 2008), Childhood cancers included leukemia and non-Hodgkin's lymphoma (Morgan
and Cassady, 2002; McKinney et al., 1991; Lowengart et al., 1987; Cohn et al., 1994; Cutler et
al., 1986; Lagakos et al., 1986; Costas et al., 2002; MADPH, 1997; Shu et al., 1999; ADHS,
1988, 1990a, b, c, 1997), CNS tumors (Morgan and Cassady, 2002; ADHS, 1998, 1990a, c,
1997; DeRoos et al., 2001; Peters and Preston-Martin, 1984; Peters et al., 1981, 1985), and total
cancers (Morgan and Cassady, 2002; ATSDR, 2006, 2008; ADHS, 1988, 1990a, 1997). These
outcomes are discussed in the other relevant sections for neurotoxicity, immunotoxicity, and
carcinogenesis.
4.10.2 Characterization of Carcinogenicity
In 1995, International Agency for Research on Cancer (IARC) concluded that
trichloroethylene is "probably carcinogenic to humans" (IARC, 1995). In 2000, National
Toxicology Program (NTP) concluded that trichloroethylene is "reasonably anticipated to be a
human carcinogen." (NTP, 2000). In 2001, the draft U.S. EPA health risk assessment of TCE
concluded that TCE was "highly likely" to be carcinogenic in humans. In 2006, a committee of
the National Research Council stated that "findings of experimental, mechanistic, and
epidemiologic studies lead to the conclusion that trichloroethylene can be considered a potential
human carcinogen" (NRC, 2006).
Following U.S. EPA (2005a) Guidelines for Carcinogen Risk Assessment, based on the
available data as of 2009, TCE is characterized as carcinogenic in humans by all routes of
exposure. This conclusion is based on convincing evidence of a causal association between TCE
exposure in humans and kidney cancer. The human evidence of carcinogenicity from
epidemiologic studies of TCE exposure is compelling for Non-Hodgkins Lymphoma (NHL) but
less convincing than for kidney cancer, and more limited for liver and biliary tract cancer.
Additionally, there are several lines of supporting evidence for TCE carcinogenicity in humans.
First, TCE induces site-specific tumors in rodents given TCE by oral gavage and inhalation.
Second, toxicokinetic data indicate that TCE absorption, distribution, metabolism, and excretion
are qualitatively similar in humans and rodents. Finally, with the exception of a mutagenic
MOA for TCE-induced kidney tumors, MOAs have not been established for TCE-induced
6/22/2009
979
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
tumors in rodents, and no mechanistic data indicate that any hypothesized key events are
biologically precluded in humans.
4.10.2.1 Summary evaluation of epidemiologic evidence of TCE and cancer
The available epidemiologic studies provide convincing evidence of a causal association
between TCE exposure and cancer. The strongest epidemiologic evidence consists of reported
increased risks of kidney cancer, with more limited evidence for non-Hodgkin lymphoma and
liver cancer, in several well-designed cohort and case-control studies (discussed below). The
summary evaluation below of the evidence for causality is based on guidelines adapted from Hill
(1965) by U.S. EPA (2005), and focuses on evidence related to kidney cancer, non-Hodgkin
lymphoma, and liver cancer.
(a) Consistency of observed association. Elevated risks for kidney cancer have been
observed across many independent studies. Eighteen studies in which there is a high likelihood
of TCE exposure in individual study subjects (e.g., based on job-exposure matrices or biomarker
monitoring) and which were judged to have met, to a sufficient degree, the standards of
epidemiologic design and analysis, were identified in a systematic review of the epidemiologic
literature. These studies consistently reported increased risks of kidney cancer, with most
estimated relative risks between 1.2 and 1.7 for overall exposure to TCE. Thirteen other cohort,
case-control, and geographic based studies were given less weight because of their lesser
likelihood of TCE exposure and other study design limitations that would decrease statistical
power and study sensitivity.
The consistency of association between TCE exposure and kidney cancer is further
supported by the results of the meta-analyses of the 14 cohort and case-control studies of
sufficient quality and with TCE exposure assigned to individual subjects. These analyses
observed a statistically significant increased pooled relative risk estimate for kidney cancer of
1.26 (95% CI: 1.11, 1.42) for overall TCE. The pooled relative risk did not change appreciably
with the removal of any individual study or with the use of alternate relative risk estimates from
individual studies. In addition, there was no evidence for heterogeneity or publication bias.
The consistency of increased kidney cancer relative risk estimates across a large number
of independent studies of different designs and populations from different countries and
industries argues against chance, bias or confounding as the basis for observed associations.
This consistency, thus, provides substantial support for a causal effect between kidney cancer
and TCE exposure.
Some evidence of consistency is found between TCE exposure and non-Hodgkin
lymphoma and liver cancer. In a systematic review of the non-Hodgkin lymphoma studies, 17
6/22/2009
980
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
studies in which there is a high likelihood of TCE exposure in individual study subjects (e.g.,
based on job-exposure matrices or biomarker monitoring) and which met, to a sufficient degree,
the standards of epidemiologic design and analysis were identified. These studies generally
reported excess relative risk estimates for non-Hodgkin lymphoma between 0.8 and 3.1 for
overall TCE exposure. Statistically significant elevated relative risk estimates were observed in
two cohort (Hansen et al., 2001; Raaschou-Nielsen et al., 2003) and one case-control (Hardell et
al., 1994) studies. The other 12 high-quality studies reported elevated relative risk estimates
with overall TCE exposure that were not statistically significant. Fifteen additional studies were
given less weight because of their lesser likelihood of TCE exposure and other design limitations
that would decrease study power and sensitivity. The observed lack of association with
lymphoma in these studies likely reflects study design and exposure assessment limitations and
is not considered inconsistent with the overall evidence on TCE and lymphoma.
Consistency of the association between TCE exposure and lymphoma is further
supported by the results of meta-analyses. These meta-analyses found a statistically significant
increased pooled relative risk estimate for lymphoma of 1.27 (95% CI: 1.04, 1.53) for overall
TCE exposure. This result and its statistical significance were not overly influenced by most
individual studies, although the removal of Hansen et al. (2001) resulted in the RRp just missing
statistical significance, with a RRp of 1.17 (95% CI: 1.00, 1.38). The result is similarly not
sensitive to most individual risk ratio estimate selections, except that the RRp is no longer
statistically significant when the Zhao et al. (2005) mortality results are substituted by either the
study's incidence results [RRp of 1.22 (95% CI: 0.99, 1.49)] or the Boice et al. (2006) results
[RRp of 1.24 (95%) CI: 1.00, 1.54). However, some heterogeneity was observed, particularly
between cohort and case-control studies, and, in addition, there was some evidence of potential
publication bias. Analyzing the cohort and case-control studies separately resolved most of the
heterogeneity, but the result for the pooled case-control studies was only a 5%> increased relative
risk estimate and was not statistically significant. The sources of heterogeneity are uncertain but
may be the result of some bias associated with exposure assessment and/or disease classification,
or from differences between cohort and case-control studies in average TCE exposure.
There are fewer studies on liver cancer than for kidney cancer and non-Hodgkin
lymphoma. Of nine studies, all of them cohort studies, in which there is a high likelihood of
TCE exposure in individual study subjects (e.g., based on job-exposure matrices or biomarker
monitoring) and which met, to a sufficient degree, the standards of epidemiologic design and
analysis in a systematic review, most reported relative risk estimates for liver and gallbladder
cancer between 0.5 and 2.0 for overall exposure to TCE. Relative risk estimates were generally
based on small numbers of cases or deaths, with the result of wide confidence intervals on the
estimates, except for one study (Raaschou-Nielsen et al., 2003). This study has almost 6 times
6/22/2009
981
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
more cancer cases than the next largest study and observed a statistically significant elevated
liver and gallbladder cancer risk with overall TCE exposure (95% CI: 1.0, 1.6). Two studies
reported a non-statistically significant reduced relative risk estimate for liver cancer and overall
TCE exposure (Boice et al., 1999; Greenland et al., 1994). Ten additional studies were given
less weight because of their lesser likelihood of TCE exposure and other design limitations that
would decrease statistical power and study sensitivity.
Consistency of the association between TCE exposure and liver cancer is further
supported by the results of meta-analyses. These meta-analyses found a statistically significant
increased pooled relative risk estimate for liver and biliary tract cancer of 1.34 (95% CI: 1.09,
1.65) with overall TCE exposure. Although there was no evidence of heterogeneity or
publication bias and the pooled estimate was fairly insensitive to the use of alternative relative
risk estimates, the statistical significance of the pooled estimate depends heavily on the one large
study by Raaschou-Nielsen et al. (2003). However, there were fewer adequate studies available
for meta-analysis of liver cancer (9 versus 15 for non-Hodgkin lymphoma and 14 for kidney),
leading to lower statistical power, even with pooling. Moreover, liver cancer is comparatively
rarer, with age-adjusted incidences roughly half or less those for kidney cancer or non-Hodgkin
lymphoma; thus, fewer liver cancer cases are generally observed in individual cohort studies.
(b) Strength of the observed association. In general, the observed associations between
TCE exposure and cancer are modest, with relative risks or odds ratios for overall TCE exposure
generally less than 2.0, and higher relative risks or odds ratios for high exposure categories.
Among the highest statistically significant relative risks were those reported for kidney cancer in
the studies by Henschler et al. (1995) [7.97 (95% CI: 2.59, 8.59)] and Vamvakas et al. (1998)
[10.80 (95% CI: 3.36, 34.75)]. As discussed in Section 4.4.3., risk magnitude in both studies is
highly uncertain due, in part, to possible selection biases, and neither was included in the meta-
analyses. However, the findings of these studies were corroborated, though with lower reported
relative risks, by later studies which overcame many of their deficiencies, such as Briining et al.
(2003) [2.47 (95% CI: 1.36, 4.49)] and Charbotel et al. (2006, 2009) [2.16 (95% CI: 1.02, 4.60)
for the high cumulative exposure group]. In addition, the very high apparent exposure in the
subjects of Henschler et al. (1995) and Vamvakas et al. (1998) may have contributed to their
reported relative risks being higher than those in other studies. Exposures in most population
case-control studies are of lower overall TCE intensity compared to exposures in Briining et al.
(2003) and Charbotel et al. (2006, 2009), and, as would be expected, observed relative risk
estimates are lower [1.24 (95% CI: 1.03, 1.49), Pesch et al., 2000a; 1.30 (95% CI: 0.9, 1.9),
Dosemeci et al., 1999], A few high-quality cohort studies reported statistically significant
relative risks of approximately 2.0 with highest exposure, including Zhao et al. (2005) [4.9 (95%
6/22/2009
982
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
CI: 1.23, 19.6) for high TCE score] and Raaschou-Nielsen et al. (2003) [1.7 (95% CI: 1.1, 2.4 for
>5 year exposure duration, subcohort with higher exposure].
Among the highest statistically significant relative risks reported for non-Hodgkin
lymphoma were those of Hansen et al. (2001) [3.1 (95% CI: 1.3, 6.1)] and Hardell et al. (1994)
[7.2 (95%) CI: 1.3, 42], the latter a case-control study whose magnitude of risk is uncertain
because of self-reported occupational TCE exposure. However, these findings are corroborated
in Seidler et al. (2007) [2.1 (95%> CI: 1.0, 4.88) for high cumulative exposure], a population case-
control study with a higher quality exposure assessment approach. Observed relative risk
estimates for liver cancer and overall TCE exposure are generally more modest.
Overall, the strength of association between TCE exposure and cancer is not large. Large
relative risk estimates are considered strong evidence of causality; however, a modest risk does
not preclude a causal association and may reflect a lower level of exposure, an agent of lower
potency, or a common disease with a high background level (U.S. EPA, 2005). Modest relative
risk estimates have been observed with several well-established human carcinogens such as
benzene and secondhand smoke. Chance cannot explain the observed association between TCE
and cancer; statistically significant associations are found in a number of the studies that
contribute greater weight to the overall evidence, given their design and statistical analysis
approaches. In addition, other known or suspected risk factors can not fully explain the observed
elevations in kidney cancer relative risks. All kidney cancer case-control studies included
adjustment for possible confounding effects of smoking, and some studies included body mass
index and hypertension. The associations between kidney cancer and TCE exposure remained in
these studies after adjustment for possible known and suspected confounders. Charbotel et al.
(2009) observed a nonstatistically significantly kidney cancer risk with exposure to only TCE
with cutting fluids [1.11 (95%> CI: 0.11, 10.71)] or to only cutting fluids without TCE [1.24 (95%>
CI: 0.39, 3.93)]; however, the finding of a 4-fold higher risk with both cutting fluid and time-
weight-average TCE exposure >50 ppm [3.74 (95%> CI: 1.32, 10.57) supports association with
TCE. Although direct examination of smoking and other suspected kidney cancer risk factors is
usually not possible in cohort studies, confounding is less likely in Zhao et al. (2005), given their
use of an internal referent group and adjustment for socioeconomic status, an indirect surrogate
for smoking, and other occupational exposures. In addition, the magnitude of the lung cancer
risk in Raaschou-Nielsen et al. (2003) suggests a high smoking rate is unlikely and cannot
explain their finding on kidney cancer.
Few risk factors are recognized for non-Hodgkin lymphoma, with the exception of
viruses and suspected factors such as immunosuppression or smoking, which are associated with
specific lymphoma subtypes. Associations between non-Hodgkin lymphoma and TCE exposure
are based on groupings of several non-Hodgkin lymphoma subtypes. Three of the six non-
6/22/2009
983
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Hodgkin lymphoma case-control studies adjusted for age, sex and smoking in statistical analyses
(Miligi et al., 2006; Seidler et al., 2007; Wang et al., 2009), the other three case-control studies
presented only unadjusted estimates of the odds ratio. Like for kidney cancer, direct examination
of possible confounding in cohort studies is not possible. The use of internal controls in some of
the higher quality cohort studies is intended to reduce possible confounding related to lifestyle
differences, including smoking habits, between exposed and referent subjects.
Heavy alcohol use and viral hepatitis are established risk factors for liver cancer, with
severe obesity and diabetes characterized as a metabolic syndrome associated with liver cancer.
Only cohort studies for liver cancer are available, and they were not able to consider these
possible risk factors.
(c) Specificity of the observed association. Specificity is generally not as relevant as
other aspects forjudging causality. As stated in the U.S. EPA Cancer Guidelines (2005), based
on our current understanding that many agents cause cancer at multiple sites, and cancers have
multiple causes, the absence of specificity does not detract from evidence for a causal effect.
Evidence for specificity could be provided by a biological marker in tumors that was specific to
TCE exposure. There is some evidence suggesting particular VHL mutations in kidney tumors
may be caused by TCE, but uncertainties in these data preclude a definitive conclusion.
(d) Temporal relationship of the observed association. Each cohort study was evaluated
for the adequacy of the follow-up period to account for the latency of cancer development. The
studies with the greatest weight based on study design characteristics (e.g., those used in the
meta-analysis) all had adequate follow-up to assess associations between TCE exposure and
cancer. Therefore, the findings of those studies are consistent with a temporal relationship.
(e) Biological gradient (exposure-response relationship). Exposure-response
relationships are examined in the TCE epidemiologic studies only to a limited extent. Many
studies examined only overall "exposed" versus "unexposed" groups and did not provide
exposure information by level of exposure. Others do not have adequate exposure assessments
to confidently distinguish between levels of exposure. For example, many studies used duration
of employment as an exposure surrogate; however, this is a poor exposure metric given subjects
may have differing exposure intensity with similar exposure duration (NRC, 2006).
Two studies of kidney cancer reported a statistically significant trend of increasing risk
with increasing TCE exposure, Zhao et al. (2005) \p = 0.023 for trend with TCE score] and
Charbotel et al. (2005, 2007) \p = 0.04 for trend with cumulative TCE exposure], Charbotel et
al. (2007) was specifically designed to examine TCE exposure and had a high-quality exposure
assessment. Zhao et al. (2005) also had a relatively well-designed exposure assessment. A
positive trend was also observed in one other study (Raaschou-Nielsen et al., 2003, with
employment duration).
6/22/2009
984
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Biological gradient is further supported by meta-analyses for kidney cancer using only
the highest exposure groups, which yielded a higher pooled relative risk estimate [1.61 (95% CI:
1.27, 2.03)] than for overall TCE exposure. Although this analysis uses a subset of studies in the
overall TCE exposure analysis, the finding of higher risk in the highest exposure groups, where
such groups were available, is consistent with a trend of increased risk with increased exposure.
The non-Hodgkin lymphoma case-control study of Seidler et al. (2007) reported a
statistically significant trend with TCE exposure \p = 0.03 for Diffuse B-cell lymphoma trend
with cumulative TCE exposure], and non-Hodgkin lymphoma risk in Boice et al. (1999)
appeared to increase with increasing exposure duration \p = 0.20 for routine-intermittent exposed
subjects]. The borderline trend with TCE intensity in the case-control study of Wang et al.
(2009) \p = 0.06] is consistent with Seidler et al. (2007). As with kidney cancer, further support
was provided by meta-analyses using only the highest exposure groups, which yielded a higher
pooled relative risk estimate [1.50 (95% CI: 1.20, 1.88)] than for overall TCE exposure. For
liver cancer, the meta-analyses using only the highest exposure groups yielded a lower, and non-
statistically significant, pooled estimate for primary liver cancer [1.25 (95% CI: 0.87, 1.79)] than
overall TCE exposure. There were no case-control studies on liver cancer and TCE, and the
cohort studies generally had few liver cancer cases, making it difficult to assess exposure-
response relationships. The one large study (Raaschou-Nielsen et al., 2003) used only duration
of employment, which is an inferior exposure metric.
(f) Biological plausibility. TCE metabolism is similar in humans, rats, and mice and
results in reactive metabolites. TCE is metabolized in multiple organs and metabolites are
systemically distributed. Several oxidative metabolites produced primarily in the liver, including
CH, TCA and DCA, are rodent hepatocarcinogens. Two other metabolites, DCVC and DCVG,
which can be produced and cleared by the kidney, have shown genotoxic activity, suggesting the
potential for carcinogenicity. Kidney cancer, lymphomas, and liver cancer have all been
observed in rodent bioassays (see below). The laboratory animal data for liver and kidney cancer
are the most robust, corroborated in multiple studies, sexes, and strains, although each has only
been reported in a single species and the incidences of kidney cancer are quite low. Lymphomas
were only reported to be statistically significantly elevated in a single study in mice, but one
additional mouse study reported elevated lymphoma incidence and one rat study reported
elevated leukemia incidence. In addition, there is some evidence both in humans and laboratory
animals for kidney, liver and immune system non-cancer toxicity from TCE exposure. Several
hypothesized modes of action have been presented for the rodent tumor findings, although there
are insufficient data to support any one mode of action, and the available evidence does not
preclude the relevance of the hypothesized modes of action to humans.
6/22/2009
985
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(g) Coherence. Coherence is defined as consistency with the known biology. As
discussed under biological plausibility, the observance of kidney and liver cancer, and
lymphomas in humans is consistent with the biological processing and toxicity of TCE.
(h) Experimental evidence (from human populations). Few experimental data from
human populations are available on the relationship between TCE exposure and cancer. The only
study of a "natural experiment" (i.e., observations of a temporal change in cancer incidence in
relation to a specific event) notes that childhood leukemia cases appeared to be more evenly
distributed throughout Woburn, MA, after closure of the two wells contaminated with
trichloroethylene and other organic solvents (MA DPH, 1997).
(i) Analogy. Exposure to structurally related chlorinated solvents such as
tetrachloroethylene and dichloromethane have also been associated with kidney, lymphoid, and
liver tumors in human, although the evidence for TCE is considered stronger.
In conclusion, based on the weight-of-evidence analysis for kidney cancer and in
accordance with EPA guidelines, TCE is characterized as "carcinogenic to humans." This
hazard descriptor is used when there is convincing epidemiologic evidence of a causal
association between human exposure and cancer. Convincing evidence is found in the
consistency of the kidney cancer findings. The strong consistency argues against chance, bias,
and confounding as explanations for the elevated kidney cancer risks. In addition, statistically
significant exposure-response trends are observed in high-quality studies. These studies were
designed to examine kidney cancer in populations with high TCE exposure intensity. These
studies addressed important potential confounders and biases, further supporting the observed
associations with kidney cancer as causal. Meta-analyses of 14 high-quality studies show
estimated relative risks or odds ratios in cohort and case-control studies are consistent, robust,
and insensitive to individual study inclusion, with no indication of publication bias or significant
heterogeneity. A statistically significant pooled relative risk estimate was observed for overall
TCE exposure [pRR = 1.27 (95% CI: 1.11, 1.42)], and the pooled relative risk estimate was
greater for the highest TCE exposure groups [pRR = 1.55 (95% CI: 1.24, 1.94)]. Given the
modest relative risk estimates and the relative rarity of the cancers observed, and therefore the
limited statistical power of individual studies, the consistency of the database is compelling. It
would require a substantial amount of high-quality negative data in order to rule out this
observed association.
The evidence is less convincing for non-Hodgkin lymphoma and liver cancer. While the
evidence is strong for NHL, issues of study heterogeneity, potential publication bias, and weaker
exposure-response results contribute greater uncertainty. The evidence is more limited for liver
cancer mainly because only cohort studies are available and most of these studies have small
numbers of cases.
6/22/2009
986
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
4.10.2.2 Summary of evidence for TCE carcinogenicity in rodents
Additional evidence of TCE carcinogenicity consists of increased incidences of tumors
reported in multiple chronic bioassays in rats and mice. In total, this database identifies some of
the same target tissues of TCE carcinogenicity also seen in epidemiological studies, including the
kidney, liver, and lymphoid tissues.
Of particular note is the site-concordant finding of TCE-induced kidney cancer in rats. In
particular, low, but biologically and sometimes statistically significant, increases in the incidence
of kidney tumors were observed in multiple strains of rats treated with TCE by either inhalation
or corn oil gavage (Maltoni et al., 1986; NTP, 1988, 1990). In addition, the gavage study by NCI
(1976) and two inhalation studies by Henschler et al. (1980), and Fukuda et al. (1983) each
observed one renal adenoma or adenocarcinoma in some dose groups and none in controls. The
largest (but still small) incidences were observed in treated male rats, although given the small
numbers, an effect in females cannot be ruled out. In fact, when results for the five rat strains
from NTP (1988) and NTP (1990) are pooled, a statistically significant trend for increased
incidence of kidney tumors is observed in females. While individual studies provide only
suggestive evidence of renal carcinogenicity, these studies taken together support the conclusion
that TCE is a kidney carcinogen in rats, with males being more sensitive than females. No other
tested laboratory species (i.e., mice and hamsters) have exhibited increased kidney tumors,
although high incidences of kidney toxicity have been reported in mice (NCI, 1976; Maltoni et
al., 1986; NTP, 1990). The GSH-conjugation-derived metabolites suspected of mediating TCE-
induced kidney carcinogenesis have not been tested in a standard 2-year bioassay, so their role
cannot be confirmed definitively. However, it is clear that GSH conjugation of TCE occurs in
humans and that the human kidney contains the appropriate enzymes for bioactivation of GSH
conjugates. Therefore, the production of the active metabolites thought to be responsible for
kidney tumor induction in rats likely occurs in humans.
Statistically significant increases in TCE-induced liver tumors have been reported in
multiple inhalation and gavage studies with male Swiss mice and B6C3F1 mice of both sexes
(NCI, 1976; Maltoni et al., 1986; NTP, 1990; Anna et al., 1994; Herren-Freund et al., 1987; Bull
et al., 2002). In female Swiss mice, on the other hand, Fukuda et al. (1983), in CD-I (ICR,
Swiss-derived) mice, and Maltoni et al. (1986) both reported small, non-significant increases at
the highest dose by inhalation. Henschler et al. (1980, 1984) reported no increases in either sex
of Han:NMRI (also Swiss-derived) mice exposed by inhalation and ICR/HA (Swiss) mice
exposed by gavage. However, the inhalation study (Henschler et al., 1980) had only 30 mice per
dose group and the gavage study (Henschler et al., 1984) had dosing interrupted due to toxicity.
Studies in rats (NCI, 1976; Henschler et al., 1980; Maltoni et al., 1986; NTP, 1988, 1990) and
hamsters (Henschler et al., 1980) did not report statistically significant increases in liver tumor
6/22/2009
987
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
induction with TCE treatment. However, several studies in rats were limited by excessive
toxicity or accidental deaths (NCI, 1976; NTP, 1988, 1990), and the study in hamsters only had
30 animals per dose group. These data are inadequate for concluding that TCE lacks
hepatocarcinogenicity in rats and hamsters, but are indicative of a lower potency in these species.
Moreover, it is notable that a few studies in rats reported low incidences (too few for statistical
significance) of very rare biliary- or endothelial-derived tumors in the livers of some treated
animals (Fukuda et al., 1983; Henschler et al., 1980; Maltoni et al.,1986). Further evidence for
the hepatocarcinogenicity of TCE is derived from chronic bioassays of the TCE oxidative
metabolites CH, TCA, and DCA in mice (e.g., George et al., 2000; Leakey et al., 2003a; Bull et
al., 1990; DeAngelo et al., 1996, 1999, 2008), all of which reported hepatocarcinogenicity. Very
limited testing of these TCE metabolites has been done in rats, with a single experiment reported
in both Richmond et al. (1995) and DeAngelo et al. (1996) finding statistically significant
DCA-induced hepatocarcinogenicity. With respect to TCA, DeAngelo et al. (1997), often cited
as demonstrating lack of hepatocarcinogenicity in rats, actually reported elevated adenoma
multiplicity and carcinoma incidence from TCA treatment. However, statistically, the role of
chance could not be confidently excluded because of the low number of animals per dose group
(20-24 per treatment group at final sacrifice). Overall, TCE and its oxidative metabolites are
clearly carcinogenic in mice, with males more sensitive than females and the B6C3F1 strain
appearing to be more sensitive than the Swiss strain. Such strain and sex differences are not
unexpected, as they appear to parallel, qualitatively, differences in background tumor incidence.
Data in other laboratory animal species are limited. Thus, except for DCA, which is
carcinogenic in rats, inadequate evidence exists to evaluate the hepatocarcinogenicity of these
compounds in rats or hamsters. However, to the extent that there is hepatocarcinogenic potential
in rats, TCE is clearly less potent in the strains tested in this species than in B6C3F1 and Swiss
mice.
Additionally, there is more limited evidence for TCE-induced lymphatic cancers in rats
and mice, lung tumors in mice, and testicular tumors in rats. With respect to the lymphomas,
Henschler et al. (1980) reported statistically significant increases in lymphomas in female
Han:NMRI mice treated via inhalation. While Henschler et al. (1980) suggested these
lymphomas were of viral origin specific to this strain, subsequent studies reported increased
lymphomas in female B6C3F1 mice treated via corn oil gavage (NTP, 1990) and leukemias in
male Sprague-Dawley and female August rats (Maltoni et al., 1986; NTP, 1988). However,
these tumors had relatively modest increases in incidence with treatment, and were not reported
to be increased in other studies. With respect to lung tumors, rodent bioassays have
demonstrated a statistically significant increase in pulmonary tumors in mice following chronic
inhalation exposure to TCE (Fukuda et al., 1983; Maltoni et al., 1988, 1986). Pulmonary tumors
6/22/2009
988
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
were not reported in other species tested (i.e., rats and hamsters; Maltoni et al., 1986, 1988;
Fukuda et al., 1983; Henschler et al., 1980). Chronic oral exposure to TCE led to a non-
statistically significant increase in pulmonary tumors in mice but, again, not in rats or hamsters
(Henschler et al., 1984; Van Duuren et al., 1979; NCI, 1976; NTP, 1988, 1990; Maltoni et al.,
1986). A lower response via oral exposure would be consistent with a role of respiratory
metabolism in pulmonary carcinogenicity. Finally, increased testicular (interstitial cell and
Ley dig cell) tumors have been observed in rats exposed by inhalation and gavage (NTP, 1988,
1990; Maltoni et al., 1986). Statistically significant increases were reported in Sprague-Dawley
rats exposed via inhalation (Maltoni et al., 1986) and Marshall rats exposed via gavage (NTP,
1988). In three rat strains, ACI, August, and F344/N, a high (>75%) control rate of testicular
tumors was observed, limiting the ability to detect a treatment effect (NTP, 1988, 1990).
In summary, there is clear evidence for TCE carcinogenicity in rats and mice, with
multiple studies showing TCE to cause tumors at multiple sites. The apparent lack of site
concordance across laboratory animal species may be due to limitations in design or conduct in a
number of rat bioassays and/or genuine inter-species differences in sensitivity. Nonetheless,
these studies have shown carcinogenic effects across different strains, sexes, and routes of
exposure, and site-concordance is not necessarily expected for carcinogens.
4.10.2.3 Summary of additional evidence on biological plausibility
Additional evidence from toxicokinetic, toxicity, and mechanistic studies supports the
biological plausibility of TCE carcinogenicity in humans.
4.10.2.3.1 Toxicokinetics
As described in Chapter 3, there is no evidence of major qualitative differences across
species in TCE absorption, distribution, metabolism, and excretion. In particular, available
evidence is consistent with TCE being readily absorbed via oral, dermal, and inhalation
exposures, and rapidly distributed to tissues via systemic circulation. Extensive in vivo and in
vitro data show that mice, rats, and humans all metabolize TCE via two primary pathways:
oxidation by CYP450s and conjugation with glutathione via GSTs. Several metabolites and
excretion products from both pathways, including TCA, DCA, TCOH, TCOG, NAcDCVC, and
DCVG, have been detected in blood and urine from exposed humans was well as from at least
one rodent species. In addition, the subsequent distribution, metabolism, and excretion of TCE
metabolites are qualitatively similar among species. Therefore, humans possess the metabolic
pathways that produce the TCE metabolites thought to be involved in the induction of rat kidney
and mouse liver tumors, and internal target tissues of both humans and rodents experience a
similar mix of TCE and metabolites.
6/22/2009
989
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
As addressed in further detail elsewhere (Chapters 3 and 5), examples of quantitative
inter-species differences in toxicokinetics include differences in partition coefficients, metabolic
capacity and affinity in various tissues, and plasma binding of the metabolite TCA. These and
other differences are addressed through PBPK modeling, which also incorporates physiological
differences among species (Section 3.5), and are accounted for in the PBPK model-based dose-
response analyses (Chapter 5). Importantly, these quantitative differences affect only inter-
species extrapolations of carcinogenic potency, and do not affect inferences as to the
carcinogenic hazard for TCE. In addition, available data on toxicokinetic differences do not
appear sufficient to explain inter-species differences in target sites of TCE carcinogenicity
(discussed further in Section 5: Dose-Response).
4.10.2.3.2 Toxicity and mode of action
Many different MO As have been proposed for TCE-induced carcinogenesis. With
respect to genotoxicity, although it appears unlikely that TCE, as a pure compound, causes point
mutations, there is evidence for TCE genotoxicity with respect to other genetic endpoints, such
as micronucleus formation (Section 4.1.1.4.4). In addition, as discussed further below, several
TCE metabolites have tested positive in genotoxicity assays. The MOA conclusions for specific
target organs in laboratory animals are summarized below. Only in the case of the kidney is it
concluded that the data are sufficient to support a particular MOA being operative. However, the
available evidence do not indicate that qualitative differences between humans and test animals
would preclude any of the hypothesized key events in rodents from occurring in humans.
For the kidney, the predominance of positive genotoxicity data in the database of
available studies of TCE metabolites derived from GSH conjugation (in particular DCVC, see
Section 4.1.5), together with toxicokinetic data consistent with their systemic delivery to and in
situ formation in the kidney, supports the conclusion that a mutagenic MOA is operative in TCE-
induced kidney tumors (see Section 4.3.7.1). Relevant data include demonstration of
genotoxicity in available in vitro assays of GSH conjugation metabolites and reported kidney-
specific genotoxicity after in vivo administration of TCE or DCVC. Mutagenicity is a well-
established cause of carcinogenicity. While supporting the biological plausibility of this
hypothesized MOA, available data on the VHL gene in humans or transgenic animals do not
conclusively elucidate the role of VHL mutation in TCE-induced renal carcinogenesis.
Cytotoxicity and compensatory cell proliferation, also presumed to be mediated through
metabolites formed after GSH-conjugation of TCE, have also been suggested to play a role in the
MOA for renal carcinogenesis, as high incidences of nephrotoxicity have been observed in
animals at doses that also induce kidney tumors. Human studies have reported markers for
nephrotoxicity at current occupational exposures, although data are lacking at lower exposures.
Nephrotoxicity alone appears to be insufficient, or at least not rate-limiting, for rodent renal
6/22/2009
990
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
carcinogenesis, since, although very high incidences of toxicity are observed in both mice and
rats, kidney tumors are only observed at low incidences in rats. In addition, nephrotoxicity has
not been shown to be necessary for kidney tumor induction by TCE in rodents. In particular,
there is a lack of experimental support for causal links, such as compensatory cellular
proliferation or clonal expansion of initiated cells, between nephrotoxicity and kidney tumors
induced by TCE. Furthermore, it is not clear if nephrotoxicity is one of several key events in a
MO A, if it is a marker for an "upstream" key event (such as oxidative stress) that may contribute
independently to both nephrotoxicity and renal carcinogenesis, or if it is incidental to kidney
tumor induction. Moreover, while toxicokinetic differences in the GSH conjugation pathway,
along with their uncertainty, are addressed through PBPK modeling, no data suggest that any of
the proposed key events for TCE-induced kidney tumors rats are precluded in humans.
Therefore, TCE-induced rat kidney tumors provide additional support for the convincing human
evidence of TCE-induced kidney cancer, with mechanistic data supportive of a mutagenic MOA.
The strongest data supporting the hypothesis of a mutagenic MOA in either the lung or
the liver are those demonstrating the genotoxicity of CH (Section 4.1.4), which is produced in
these target organs as a result of oxidative metabolism of TCE. It has been suggested that CH
mutagenicity is unlikely to be the cause of TCE hepatocarcinogenicity because the
concentrations required to elicit these responses are several orders of magnitude higher that
achieved in vivo (Moore and Harrington-Brock, 2000). However, it is not clear how much of a
correspondence is to be expected from concentrations in genotoxicity assays in vitro and
concentrations in vivo, as reported in vivo CH concentrations are in whole liver homogenate
while in vitro concentrations are in culture media. The use of ip administration, which leads to
an inflammatory response, in many other in vivo genotoxicity assays in the liver and lung
complicates the comparison with carcinogenicity data. Also, it is difficult with the available data
to assess the contributions from genotoxic effects of CH along with those from the genotoxic and
non-genotoxic effects of other oxidative metabolites (e.g., DCA and TCA). Therefore, while
data are insufficient to conclude that a mutagenic MOA mediated by CH is operant, a mutagenic
MOA in the liver or lung, either mediated by CH or by some other oxidative metabolite of TCE,
cannot be ruled out.
A second MOA hypothesis for TCE-induced liver tumors involves activation of the
PPARa receptor. Clearly, in vivo administration of TCE leads to activation of PPARa in rodents
and likely does so in humans as well (based on in vitro data for TCE and its oxidative
metabolites). However, the evidence as a whole does not support the view that PPAR-a is the
sole operant MOA mediating TCE hepatocarcinogenesis. Although metabolites of TCE activate
PPARa, the data on the subsequent elements in the hypothesized MOA (e.g., gene regulation,
cell proliferation, apoptosis, and selective clonal expansion), while limited, indicate significant
6/22/2009
991
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
differences between PPAR-a agonists such as Wy-14643 and TCE or its metabolites. For
example, compared with other agonists, TCE induces transient as opposed to persistent increases
in DNA synthesis; increases (or is without effect on), as opposed to decreases, apoptosis; and
induces a different H-ras mutation frequency or spectrum. These data support the view that
mechanisms other than PPARa activation may contribute to these effects; besides PPARa
activation, the other hypothesized key events are non-specific, and available data (e.g., using
knockout mice) do not indicate that they are solely or predominantly dependent on PPARa. A
second consideration is whether certain TCE metabolites (e.g., TCA) that activate PPAR-a are
the sole contributors to its carcinogenicity. As summarized above (see Section 4.10.1.3), TCA is
not the only metabolite contributing to the observed non-cancer effects of TCE in the liver.
Other data also suggest that multiple metabolites may also contribute to the hepatic
carcinogenicity of TCE. Liver phenotype experiments, particularly those utilizing
immunostaining for c-Jun, support a role for both DCA and TCA in TCE-induced tumors, with
strong evidence that TCA cannot solely account for the characteristics of TCE-induced tumors
(e.g., Bull et al., 2002). In addition, H-ras mutation frequency and spectrum of TCE-induced
tumors more closely resembles that of spontaneous tumors or of those induced by DCA, and
were less similar in comparison to that of TCA-induced tumors. The heterogeneity of TCE-
induced tumors is similar to that observed to be induced by a diversity carcinogens including
those that do not activate PPAR-a, and to that observed in human liver cancer. Taken together,
the available data indicate that, rather than being solely dependent on a single metabolite (TCA)
and/or molecular target (PPAR-a) multiple TCE metabolites and multiple toxicity pathways
contribute to TCE-induced liver tumors.
Other considerations as well as new data published since the NRC (2006) review are also
pertinent to the liver tumor MOA conclusions. It is generally acknowledged that, qualitatively,
there are no data to support the conclusion that effects mediated by the PPAR-a receptor that
contribute to hepatocarcinogenesis would be biologically precluded in humans (Klaunig et al.,
2003; NRC, 2006). It has, on the other hand, been argued that due to quantitative toxicokinetic
and toxicodynamic differences, the hepatocarcinogenic effects of chemicals activating this
receptor are "unlikely" to occur in humans (Klaunig et al., 2003; NRC, 2006); however, several
lines of evidence strongly undermine the confidence in this assertion. With respect to
toxicokinetics, as discussed above, quantitative differences in oxidative metabolism are
accounted for in PBPK modeling of available in vivo data, and do not support inter-species
differences of a magnitude that would preclude hepatocarcinogenic effects based on
toxicokinetics alone. With respect to the MOA proposed by Klaunig et al. (2003), recent
experiments have demonstrated that PPAR-a activation and the sequence of key events in the
hypothesized MOA are not sufficient to induce hepatocarcinogenesis (Yang et al., 2007).
6/22/2009
992
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Moreover, the demonstration that the PPAR-a agonist DEHP induces tumors in PPAR-a-null
mice supports the view that the events comprising the hypothesized MOA are not necessary for
liver tumor induction in mice by this PPARa agonist (Ito et al., 2007). Therefore several lines of
evidence, including experiments published since the NRC (2006) review, call into question the
scientific validity of using the PPAR-a MOA hypothesis as the basis for evaluating the relevance
to human carcinogenesis of rodent liver tumors (Guyton et al., 2009).
In summary, available data support the conclusion that the MOA for TCE-induced liver
tumors in laboratory animals is not known. However, a number of qualitative similarities exist
between observations in TCE-exposed mice and what is known about the etiology and induction
of human hepatocellular carcinomas. Polyploidization, changes in glycogen storage, inhibition
of GST-zeta, and aberrant DNA methylation status, which have been observed in studies of mice
exposed to TCE or its oxidative metabolites, are all either clearly related to human
carcinogenesis or are areas of active research as to their potential roles (PPARa activation is
discussed below). The mechanisms by which TCE exposure may interact with known risk
factors for human hepatocellular carcinomas are not known. However, available data do not
suggest that TCE exposure to mice results in liver tumors that are substantially different in terms
of their phenotypic characteristics either from human hepatocellular carcinomas or from rodent
liver tumors induced by other chemicals.
Comparing various other, albeit relatively non-specific, tumor characteristics between
rodent species and humans provides additional support to the biologic plausibility of TCE
carcinogenicity. For example, in the kidney and the liver, the higher incidences of background
and TCE-induced tumors in male rats and mice, respectively, as compared to females parallels
the observed higher human incidences in males for these cancers (Ries et al., 2008). For the
liver, while there is a lower background incidence of liver tumors in humans than in rodents, in
the United States there is an increasing occurrence of liver cancer associated with several factors,
including viral hepatitis, higher survival rates for cirrhosis, and possibly diabetes (reviewed in
El-Serag, 2007). In addition, Leakey et al. (2003) reported that increased body weight in
B6C3F1 mice is strongly associated with increased background liver tumor incidences, although
the mechanistic basis for this risk factor in mice has not been established. Nonetheless, it is
interesting that recent epidemiologic studies have suggested obesity, in addition to associated
disorders such as diabetes and metabolic syndrome, as a risk factor for human liver cancer
(El-Serag, 2007; El-Serag and Rudolph, 2007). Furthermore, the phenotypic and morphologic
heterogeneity of tumors seen in the human liver is qualitatively similar to descriptions of mouse
liver tumors induced by TCE exposure, as well as those observed from exposure to a variety of
other chemical carcinogens. These parallels suggest similar pathways (e.g., for cell signaling) of
6/22/2009
993
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
carcinogenesis may be active in mice and humans and support the qualitative relevance of mouse
models of liver to human liver cancer.
For mouse lung tumors, MO A hypotheses have centered on TCE metabolites produced
via oxidative metabolism in situ. As discussed above, the hypothesis that the mutagenicity of
reactive intermediates or metabolites (e.g., CH) generated during P450 metabolism contributes to
lung tumors cannot be ruled out, although available data are inadequate to conclusively support
this MOA. An alternative MOA has been posited involving other effects of such oxidative
metabolites, particularly CH, including cytotoxicity and regenerative cell proliferation.
Experimental support for this alternative hypothesis remains limited, with no data on proposed
key events in experiments of duration 2 weeks or longer. While the data are inadequate to
support this MOA hypothesis, the data also do not suggest that any proposed key events would
be biologically plausible in humans. Furthermore, the focus of the existing MOA hypothesis
involving cytotoxicity has been CH, and, as summarized above (Section 4.10.1.5), other
metabolites may contribute to respiratory tract non-cancer toxicity or carcinogenicity. In sum,
the MOA for mouse lung tumors induced by TCE is not known.
A MOA subsequent to in situ oxidative metabolism, whether involving mutagenicity,
cytotoxicity, or other key events, may also be relevant to other tissues where TCE would
undergo P450 metabolism. For instance, CYP2E1, oxidative metabolites, and protein adducts
have been reported in the testes of rats exposed to TCE, and, in some rat bioassays, TCE
exposure increased the incidence of rat testicular tumors. However, inadequate data exist to
adequately define a MOA hypothesis for this tumor site.
4.10.3 Characterization of Factors Impacting Susceptibility
As discussed in more detail in Section 4.9, there is some evidence that certain
subpopulations may be more susceptible to exposure to TCE. Factors affecting susceptibility
examined include lifestage, gender, genetic polymorphisms, race/ethnicity, pre-existing health
status, and lifestyle factors and nutrition status.
Examination of early lifestages includes exposures such as transplacental transfer
(Beppu, 1968; Laham, 1970; Withey and Karpinski, 1985; Ghantous et al., 1986; Helliwell and
Hutton, 1950) and breast milk ingestion (Fisher et al., 1990, 1997; Pellizzari et al., 1982;
Hamada and Tanaka, 1995), early lifestage-specific toxicokinetics, PBPK models (Fisher et al.,
1989, 1990), and differential outcomes in early lifestages such as developmental cardiac defects.
Although there is more information on susceptibility to TCE during early lifestages than on
susceptibility during later lifestages or for other populations with potentially increased
susceptibility, there remain a number of uncertainties regarding children's susceptibility.
6/22/2009
994
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Improved PBPK modeling for using childhood parameters for early lifestages as recommended
by the NRC (2006), and validation of these models will aid in determining how variations in
metabolic enzymes affect TCE metabolism. In particular, the NRC states that it is prudent to
assume children need greater protection than adults, unless sufficient data are available to justify
otherwise (NRC, 2006). Because the weight of evidence supports a mutagenic MOA for TCE
carcinogenicity in the kidney (see Section 4.3.7), and there is an absence of chemical-specific
data to evaluate differences in carcinogenic susceptibility, early-life susceptibility should be
assumed and the age-dependent adjustment factors (ADAFs) should be applied, in accordance
with the Supplemental Guidance (discussed further in Chapter 5).
Fewer data are available on later lifestages, although there is suggestive evidence to
indicate that older adults may experience increased adverse effects than younger adults (Mahle et
al., 2007; Rodriguez et al., 2007). In general, more studies specifically designed to evaluate
effects in early and later lifestages are needed in order to more fully characterize potential life
stage-related TCE toxicity.
Examination of gender-specific susceptibility includes toxicokinetics, PBPK models
(Fisher et al., 1998), and differential outcomes. Gender differences observed are likely due to
variation in physiology and exposure.
Genetic variation likely has an effect on the toxicokinetics of TCE. In particular,
differences in CYP2E1 activity may affect susceptibility of TCE due to effects on production of
toxic metabolites (Kim and Ghanayem, 2006; Lipscomb et al., 1997; Povey et al., 2001; Yoon et
al., 2007). GST polymorphisms could also play a role in variability in toxic response (Briining et
al., 1997; Wiesenhiitter et al., 2007), as well as other genotypes, but these have not been
sufficiently tested. Differences in genetic polymorphisms related to the metabolism of TCE have
also been observed among various race/ethnic groups (Inoue et al., 1989; Sato et al., 1991b).
Pre-existing diminished health status may alter the response to TCE exposure.
Individuals with increased body mass may have an altered toxicokinetic response (Clewell et al.,
2000; Sato, 1993; Sato et al., 1991b; Monster et al., 1979; McCarver et al., 1998; Davidson and
Beliles, 1991; Lash et al., 2000) resulting in changes the internal concentrations of TCE or in the
production of toxic metabolites. Other conditions, including diabetes and hypertension, are risk
factors for some of the same health effects that have been associated with TCE exposure, such as
renal cell carcinoma. However, the interaction between TCE and known risk factors for human
diseases is not known, and further evaluation of the effects due to these factors is needed.
Lifestyle and nutrition factors examined include alcohol consumption, tobacco smoking,
nutritional status, physical activity, and socioeconomic status. In particular, alcohol intake has
been associated with metabolic inhibition (altered CYP2E1 expression) of TCE in both humans
and experimental animals (Bardodej and Vyskocil, 1956; Barret et al., 1984; McCarver et al.,
6/22/2009
995
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1998; Miiller et al., 1975; Sato, 1993; Sato et al., 1980, 1981, 1983, 1991a; Stewart et al., 1974;
Kaneko et al., 1994; Larson and Bull, 1989; Nakajima et al., 1988, 1990, 1992b; Okino et al.,
1991; Sato and Nakajima, 1985; White and Carlson, 1981). In addition, such factors have been
associated with increased baseline risks for health effects associated with TCE, such as kidney
cancer (e.g., smoking) and liver cancer (e.g., alcohol consumption). However, the interaction
between TCE and known risk factors for human diseases is not known, and further evaluation of
the effects due to these factors is needed.
In sum, there is some evidence that certain subpopulations may be more susceptible to
exposure to TCE. Factors affecting susceptibility examined include lifestage, gender, genetic
polymorphisms, race/ethnicity, pre-existing health status, and lifestyle factors and nutrition
status. However, except in the case of toxicokinetic variability characterized using the PBPK
model described in Section 3.5, there are inadequate chemical-specific data to quantify the
degree of differential susceptibility due to such factors.
6/22/2009
996
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
References
ADHS (Arizona Department of Health Services). (1988) Report on mortality in Maricopa
County 1966-1986.
ADHS. (1990a) Incidence study of childhood cancer in Maricopa County: 1965-1986.
ADHS. (1990b) Cancer incidence and mortality in an East Phoenix area overlaying groundwater
contaminated with volatile organic compounds.
ADHS. (1990c) The incidence of childhood leukemia and testicular cancer in Pima County:
1970-1986.
ADHS. (1997) Case-referent study of childhood leukemia in Maricopa County, Arizona 1965-
1990.
Albee, RR; Spencer, PJ; Johnson, KA; et al. (2006) Lack of trigeminal nerve toxicity in rats
exposed to trichloroethylene vapor for 13 weeks. Int J Toxicol 25(6):531-540.
Anna CH, Maronpot RR, Pereira MA, Foley JF, Malarkey DE, Anderson MW. 1994. ras proto-
oncogene activation in dichloroacetic-, trichloroethylene- and tetrachloroethylene-
induced liver tumors in B6C3F1 mice. Carcinogenesis 15:2255-2261.
Aranyi C, O'Shea WJ, Graham JA, Miller FJ. 1986. The effects of inhalation of organic chemical
air contaminants on murine lung host defenses. Fundam Appl Toxicol 6:713-720.
Arito, H; Takahashi, M; Ishikawa, T. (1994) Effect of subchronic inhalation exposure to low-
level Trichloroethylene on heart rate and wakefulness-sleep in freely moving rats. Jpn J
Ind Health (Sangyo Igaku) 36(1): 1-8.
Asal NR, Risser DR, Kadamani S, Geyer JR, Lee ET, Cherng N. (1988a) Risk factors in renal
cell carcinoma: I. Methodology, demographics, tobacco, beverage use, and obesity.
Cancer Detect Prev 11 (3-6):359-77.
Asal NR, Geyer JR, Risser DR, Lee ET, Kadamani S, Cherng N. (1988b) Risk factors in renal
cell carcinoma. II. Medical history, occupation, multivariate analysis, and conclusions.
Cancer Detect Prev 13(3-4):263-79.
6/22/2009
997
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
ATSDR. (1998) Volatile organic compounds in drinking water and adverse pregnancy outcomes.
U.S. Department of Health and Human Services, August 1998, Atlanta, GA.
ATSDR. (2001) Final Report: evaluation of priority health conditions in a community with
historical contamination by trichloroethylene. U.S. Department of Health and Human
Services, August 2001, Atlanta, GA.
ATSDR. (2003a) Impact of trichloroethylene exposure on oral motor, speech and hearing in
children. U.S. Department of Health and Human Services, July 2003, Atlanta, GA.
ATSDR (Agency for Toxic Substance and Disease Registry). 2004. Feasibility Investigation of
Worker Exposure to Trichloroethylene at the View-Master Factory in Beaverton, Oregon.
Prepared by the Oregon Department of Human Services. Report No. PB2005-102173.
ATSDR (Agency for Toxic Substances and Disease Registry). 2006. Health Consultation.
Cancer and Birth Outcome Analysis, Endicott Area, Town of Union, Broome County,
New York. Health Statistics Review. Atlanta, GA: U.S. Department of Health and
Human Services, Public Health Service, Agency for Toxic Substances and Disease
Registry. May 26, 2006.
ATSDR. (2008) Health statistics review follow-up: cancer and birth outcome analysis. Endicott
Area Investigation, Endicott Area, Town of Union, Broome County, New York. U.S.
Department of Health and Human Services, May 2008, Atlanta, GA.
Bai CL, Canfield PJ, Stacey NH. 1992b. Individual serum bile acids as early indicators of
carbon tetrachloride- and chloroform-induced liver injury. Toxicology 75:221-234.
Bardodej Z, Vyskocil J. (1956) The problem of trichloroethylene in occupational medicine;
trichloroethylene metabolism and its effect on the nervous system evaluated as a means
of hygienic control. AM A Arch Ind Health 13(6): 581-592.
Barret, L; Arsac, P; Vincent, M; et al. (1982) Evoked trigeminal nerve potential in chronic
trichloroethylene intoxication. J Toxicol Clin Toxicol 19(4):419-423.
Barret L, Faure J, Guilland B, Chomat D, Didier B, Debru JL. (1984) Trichloroethylene
occupational exposure: elements for better prevention. Int Arch Occup Environ Health
53(4):283-9.
6/22/2009
998
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Barret, L; Garrel, S; Danel, V; et al. (1987) Chronic trichloroethylene intoxication: a new
approach by trigeminal-evoked potentials? Arch Environ Health 42(5):297-302.
Barret, L; Torch, S; Leray, CL; Sarlieve, L; Saxod, R (1992) Morphometric and Biochemical
Studies in Trigeminal Nerves of Rat after Trichl or ethylene or Dichloroacetylene Oral
Administration. Neurotoxicology 13:601-614.
Barret, L; Torch, S; Usson, Y Barret, L; Torch, S; Usson, Y; et al. (1991) A morphometric
evaluation of the effects of trichloroethylene and dichloroacetylene on the rat mental
nerve. Preliminary results. Neurosci Lett 131(2): 141-144.
Beckstead MJ, Weiner JL, Eger EI 2nd, Gong DH, Mihic SJ. (2000). Glycine and gamma-
aminobutyric acid(A) receptor function is enhanced by inhaled drugs of abuse. Mol
Pharmacol. 57:1199-205.
Beppu, K. (1968) Transmission of the anesthetic agents through the placenta in painless delivery
and their effects on newborn infants. Keio J Med 17(2):81—107.
Berger T, Horner CM. (2003) In vivo exposure of female rats to toxicants may affect oocyte
quality. Reprod Toxicol 17:273-281.
Berman, E; Schlicht, M; Moser, VC; et al. (1995) A Multidisciplinary Approach to
Toxicological Screening: I. Systemic Toxicity. Journal of Toxicology and Environmental
Health 45:127-143.
Bernad PG, Newell S, Spyker DA. (1987) Neurotoxicity and behavior abnormalities in a cohort
chronically exposed to trichloroethylene. Abstract. Vet Hum Toxicol 29(6):475.
Blain, L; Lachapelle, P; Molotchnikoff, S. (1994) Electroretinal responses are modified by
chronic exposure to trichloroethylene. Neurotoxicology 15(3):627—631.
Blair A, Hartge P, Stewart PA, McAdams M, Lubin J. 1998. Mortality and cancer incidence of
aircraft maintenance workers exposed to trichloroethylene and other organic solvents and
chemicals: extended follow up. Occup Environ Med 55:161-171.
Blake JE, Compton KV, Schmidt W, Orrego H. 1988. Accuracy of death certificates in the
diagnosis of alcoholic liver cirrhosis. Alcoholism: Clinical and Experimental Research
12:168-172.
6/22/2009
999
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Blossom SJ, Doss JC. (2007) Trichloroethylene alters central and peripheral immune function in
autoimmune-prone MRL+/+ mice following continuous developmental and early life
exposure. Journal of Immunotoxicology 4:129-141.
Blossom SJ, Doss JC, Gilbert KM. (2007) Chronic exposure to a trichloroethylene metabolite in
autoimmune-prone MRL+/+ mice promotes immune modulation and alopecia. Toxicol
Sci 95(2):401—11.
Blossom SJ, Pumford NR, Gilbert KM. 2004. Activation and attenuation of apoptosis of CD4+ T
cells following in vivo exposure to two common environmental toxicants,
trichloroacetaldehyde hydrate and trichloroacetic acid. J Autoimmun 23:211-220.
Blossom SJ, Doss JC, Hennings LJ, Jernigan S, Melnyk S, James SJ. 2008. Developmental
exposure to trichloroethylene promotes CD4+ T cell differentiation and hyperactivity in
association with oxidative stress and neurobehavioral deficits in MRL+/+ mice. Toxicol
Appl Pharmacol 231:344-353.
Boice JD, Marano DE, Fryzek JP, Sadler CJ, McLaughlin JK. (1999) Mortality among aircraft
manufacturing workers. Occup Environ Med 56:581-597.
Boice JD, MaranoDE, Cohen SS, Mumma MR, Blott WJ, Brill AB, Fryzek JP, Henderson BE,
McLaughlin JK. 2006. Mortality among Rocketdyne worker who tested rocket engines,
1948-1999. J Occup Environ Med 48:1070-1092.
Bolt H, Lammert M, Selinski S, Briining T. 2004. Urinary al-microglobulin excretion as
biomarker of renal toxicity in trichloroethylene-exposed persons. Int Arch Occup
Environ Health 77:186-190.
Bove FJ, Fulcomer MC, Klotz JB, Esmart J, Dufficy EM, Savrin JE. (1995) Public drinking
water contamination and birth outcomes. Am J Epidemiol 141(9):850-862.
Bove FJ. (1996) Public drinking water contamination and birthweight, prematurity, fetal deaths,
and birth defects. Toxicol Ind Health 12(2):255-266.
Boyer, AS; Finch, WT; Runyan, RB. (2000) Trichloroethylene inhibits development of
embryonic heart valve precursors in vitro. Toxicol Sci 53:109-117.
Boyes, WK; Bercegeay, M; Ali, JS; et al. (2003) Dose-based duration adjustments for the effects
of inhaled trichloroethylene on rat visual function. Toxicol Sci 76(1): 121-130.
6/22/2009
1000
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Boyes, WK; Bercegeay, M; Krantz, T; et al. (2005) Momentary brain concentration of
trichloroethylene predicts the effects on rat visual function. Toxicol Sci 87(1): 187-196.
Briving, C; Jacobson, I; Hamberger, A; et al. (1986) Chronic effects of perchloroethylene and
trichloroethylene on the gerbil brain amino acids and glutathione. Neurotoxicology
7(1): 101-108.
Bross, G; DiFranceisco, D; Desmond, ME. (1983) The effects of low dosages of
trichloroethylene on chick development. Toxicology 28:283-294.
Briming T, Lammert M, Kempkes M, Thier R, Golka K, Bolt HM. (1997) Influence of
polymorphisms of GSTM1 and GSTT1 for risk of renal cell cancer in workers with long-
term high occupational exposure to trichloroethene. Arch Toxicol 71(9):596-9.
Briining T, Mann H, Melzer H, Sundberg AGM, Bolt HM. 1999a. Pathological excretion
patterns of urinary proteins in renal cell cancer patients exposed to trichloroethylene.
Occup Med. 29:299-305.
Briining T, Sundberg AGM, Birner G, Lammert M, Bolt HM, Appelkvist E-L, Nilsson R,
Dallner G. 1999b. Glutathione transferase alpha as a marker for tubular damage after
trichloroethylene exposure. Arch Toxicol 73:246-254.
Briining T, Pesch B, Wiesenhiitter B, Rabstein S, Lammert M, Baumiiller A, Bolt HM. 2003.
Renal cell cancer risk and occupational exposure to trichloroethylene: results of a
consecutive case-control study in Arnsberg, Germany. Am J Ind Med. 43:274-85.
Buben, JA; O'Flaherty, EJ. (1985) Delineation of the role of metabolism in the hepatotoxicity of
trichloroethylene and perchloroethylene: a dose-effect study. Toxicol Appl Pharmacol
78:105-122.
Bull RJ, Sanchez IM, Nelson MA, Larson JL, Lansing AJ. 1990. Liver tumor induction in
B6C3F1 mice by dichloroacetate and trichloroacetate. Toxicology 63:341-359.
Bull, RJ; Orner, GA; Cheng, RS; et al. (2002) Contribution of dichloroacetate and
trichloroacetate to liver tumor induction in mice by trichloroethylene. Toxicol Appl
Pharmacol 182:55-65.
6/22/2009
1001
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Burg JA, Gist GL. (1997) The potential impact on women from environmental exposures.
Journal of women's health / the official publication of the Society for the Advancement of
Women's Health Research 6(2): 159-161.
Burg JR, Gist GL. (1999) Health effects of environmental contaminant exposure: an intrafile
comparison of the Trichloroethylene Subregistry. Archives of environmental health
54(4):231-241.
Burg JR, Gist GL, Allred SL, Radtke TM, Pallos LL, Cusack CD. (1995) The National Exposure
Registry - Morbidity analyses of noncancer outcomes from the trichloroethylene
subregistry baseline data. Int J Occup Med and Toxicol. 4(2):237-257.
Cai P, Konig R, Boor PJ, Kondraganti S, Kaphalia BS, Khan MF, et al. 2008. Chronic exposure
to trichloroethene causes early onset of SLE-like disease in female MRL +/+ mice.
Toxicol Appl Pharmacol 228:68-75.
Cakmak A, Ekici A, Ekici M, Arsian M, Iteginli A, Kurtipek E, Kara T. 2004. Respiratory
findings in gun factory workers exposed to solvents. Respirat Med 98 :52-56.
Carney, EW; Thorsrud, BA; Dugard, PH; Zablotny, CL. (2006) Developmental toxicity studies
in Crl:Cd (SD) rats following inhalation exposure to trichloroethylene and
perchloroethylene. Birth Defects Research (Part B) 77:405-412.
Carried M, Magosso D, Piccoli P, Zanetti E, Trevisan A, Bartolucci GB. 2007. Acute nonfatal
intoxication with trichloroethylene. Arch Toxicol 81:529-32.
Channel SR, Latendresse JR, Kidney JK, Grabau JH, Lane JW, Steel-Goodwin L, Gothaus MC
(1998). A subchronic exposure to trichloroethylene causes lipid peroxidation and
hepatocellular proliferation in male B6C3F1 mouse liver, Toxicol. Sci. 43: 145-154.
Charbotel B, Fevotte J, Hours M, Martin JL, Bergeret A. 2005. Case-control study on renal cell
cancer and occupational trichloroethylene exposure, in the Arve valley (France). Final
report. Submitted to U.S. EPA by Paul Dugard, Ph.D., Halogenated Solvents Industry
Alliance. #8EHQ-1105-16030.
Charbotel B, Fevotte J, Hours M, Martin JL, Bergeret A. (2006) Case-control study on renal cell
cancer and occupational exposure to trichloroethylene. Part II: Epidemiological aspects.
Ann Occup Hyg 50(8):777-87.
6/22/2009
1002
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Charbotel B, Gad S, Cai'ola D, Beroud C, Fevotte J, Bergeret A, Ferlicot S, Richard S. (2007)
Trichloroethylene exposure and somatic mutations of the VHL gene in patients with
Renal Cell Carcinoma. J Occup Med Toxicol. 2:13.
Charbotel B, Fevotte J, martin JL, Bergeret A. 2009. Cancer du rein et expositions au
trichloroethylene: les valeurs limites d'exposition professionnelle fra9aises en vigueur
sont-elles adaptees. Rev Epidemiol Sante Publique 57:41-47.
Chia SE, Goh VH, Ong CN. (1997) Endocrine profiles of male workers with exposure to
trichloroethylene. Am J Ind Med 32(3):217-222.
Chia SE, Ong CN, Tsakok MF, Ho A. (1996) Semen parameters in workers exposed to
trichloroethylene. Reprod Toxicol 10(4):295-299.
Clewell HJ 3rd, Gentry PR, Covington TR, Gearhart JM. (2000) Development of a
physiologically based pharmacokinetic model of trichloroethylene and its metabolites for
use in risk assessment. Environ Health Perspect 108 Suppl 2:283-305.
Cohn M, Klotz J, Bove F, Berkowitz M, Fagliano J. 1994. Drinking water contamination and
the incidence of leukemia and non-Hodgkin's lymphoma. Environ Health Perspect
102:556-561.
Cooper GS, Stroehla BC. 2003. The epidemiology of autoimmune diseases. Autoimmun Rev
2:119-125.
Cosby NC, Dukelow WR. (1992) Toxicology of maternally ingested trichloroethylene (TCE) on
embryonal and fetal development in mice and of TCE metabolites on in vitro fertilization.
Fundam Appl Toxicol 19:268-274.
Costas K, Knorr RS, Condon SK. (2002) A case-control study of childhood leukemia in Woburn,
Massachusetts: the relationship between leukemia incidence and exposure to public
drinking water. Sci Total Environ 300(l-3):23-35.
Cummings, BS; Lash, LH. (2000) Metabolism and toxicity of trichloroethylene and S (1,2
dichlorovinyl)-L-cysteine in freshly isolated human proximal tubular cells. Toxicol Sci
53(2):458-466.
6/22/2009
1003
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Cummings, BS; Parker, JC; Lash, LH. (2000a) Role of cytochrome P450 and glutathione S
transferase alpha in the metabolism and cytotoxicity of trichloroethylene in rat kidney.
Biochem Pharmacol 59(5):531—543.
Cummings, BS; Zangar, RC; Novak, RF; et al. (2000b) Cytotoxicity of trichloroethylene and S-
(1, 2-dichlorovinyl)-L-cysteine in primary cultures of rat renal proximal tubular and distal
tubular cells. Toxicology 150:83-98.
Cutler JJ, Parker GS, Rosen S, Prenney B, Healey R, Caldwell GG. (1986) Childhood leukemia
in Woburn, Massachusetts. Public Health Rep 101(2):201-205.
Das, RM; Scott, JE. (1994) Trichloroethylene-induced pneumotoxicity in fetal and neonatal
mice. Toxicology Letters 73:227-239.
Davidson IW, Beliles RP. (1991) Consideration of the target organ toxicity of trichloroethylene
in terms of metabolite toxicity and pharmacokinetics. Drug Metab Rev 23(5-6):493-599.
Davis SI, Laszlo Pallos L, Wu JQ, Sapp JH 2nd, Cusack C. (2005) ATSDR's trichloroethylene
subregistry methods and results: 1989-2000. Arch Environ Occup Health 60(3): 130-9.
Dawson BV, Johnson PD, Goldberg SJ, Ulreich JB. (1990) Cardiac teratogenesis of
trichloroethylene and dichloroethylene in a mammalian model. J Am Coll Cardiol
16(5): 1304—1309.
Dawson BV, Johnson PD, Goldberg SJ, Ulreich JB. (1993) Cardiac teratogenesis of halogenated
hydrocarbon-contaminated drinking water. J Am Coll Cardiol 21(6): 1466-1472.
DeAngelo, AB; Daniel, FB; Most, BM; et al. (1996) The carcinogenicity of dichloroacetic acid
in the male Fischer 344 rat. Toxicology 114:207-221.
DeAngelo, AB; Daniel, FB; Most, BM; et al. (1997) Failure of monochloroacetic acid and
trichloroacetic acid administered in the drinking water to produce liver cancer in male
F344/N rats. J Toxicol Environ Health 52:425-445.
DeAngelo AB, George MH, House DE. 1999. Hepatocarcinogenicity in the male B6C3F1
mouse following a lifetime exposure to dichloroacetic acid in the drinking water: dose-
response determination and modes of action. J Toxicol Environ Health 58:485-507.
6/22/2009
1004
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
DeAngelo AB, Daniel FB, Wong DM, George MH. 2008. The induction of hepatocellular
neoplasia by trichloroacetic acid administered in the drinking water of the male B6C3F1
mouse. J Toxicol and Environmental Health Part A 71: 1056-1068.
Dees, C; Travis, C. (1993) The mitogenic potential of trichloroethylene in B6C3F1 mice.
Toxicol Lett 69:129-137.
De Roos AJ, Olshan AF, Teschke K, Poole C, Savitz DA, Blatt J, Bondy ML, Pollock BH.
(2001) Parental occupational exposures to chemicals and incidence of neuroblastoma in
offspring. Am J Epidemiol 154(2): 106-14.
Diot E, Lesire V, Guilmot JL, Metzger MD, Pilore R, Rogier S, et al. 2002. Systemic sclerosis
and occupational risk factors: a case-control study. Occup Environ Med 59:545-549.
Dorfmueller, MA; Henne, SP; York,G; Bornschein, RL; Manson, JM. (1979) Evaluation of
teratogenicity and behavioral toxicity with inhalation exposure of maternal rats to
trichloroethylene. Toxicology 14:153-166.
Dosemeci, M; Cocco, P; Chow, WH. (1999) Gender differences in risk of renal cell carcinoma
and occupational exposures to chlorinated aliphatic hydrocarbons. Am J Ind Med
36(l):54-9.
Drake, VJ; Koprowski, SL; Lough, J; Hu, N; Smith, SM. (2006a) Trichloroethylene exposure
during cardiac valvuloseptal morphogenesis alters cushion formation and cardiac
hemodynamics in the avian embryo. Environ Health Perspect 114(6):842-847.
Drake, VJ; Koprowski, SL; Hu, N; Smith, SM; Lough, J. (2006b) Carcinogenic effects of
trichloroethylene and trichloroacetic acid following exposure during heart specification
of avian development. Toxicol Sci 94(1): 153-162.
Driscoll TR, Hamdan HH, Wang G, Wright PFA, Stacey NH. 1992. Concentrations of
individual serum or plasma bile acids in workers exposed to chlorinated aliphatic
hydrocarbons. Br J Ind Med 49:700-705.
DuTeaux, SB; Berger, T; Hess, RA; Sartin, BL ; Miller, MG. (2004b) Male reproductive toxicity
of trichloroethylene: sperm protein oxidation and decreased fertilizing ability. Biol
Reprod 70:1518-1526.
6/22/2009
1005
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Elcombe, CR; Rose, MS; Pratt, IS. (1985) Biochemical, histological, and ultrastructural changes
in rat and mouse liver following the administration of trichloroethylene: possible
relevance to species differences in hepatocarcinogenicity. Toxicology and Applied
Pharmacology 79:365-376.
El-Ghawabi, SM; Mansoor, MB; El Gamel, MS; El Saharti, AA; El Enany, FF. (1973) Chronic
trichloroethylene exposure. J Egypt Med Assoc 56(11-12):715-724.
El-SeragHB. 2007. Epidemiology of hepatocellular carcinoma in USA. Hepatology Res
37(Suppl 2):S88-S94.
Epstein, DL; Nolen, GA; Randall, JL; Christ, SA; Read, EJ; Stober, JA; Smith, MK. (1992)
Cardiopathic effects of dichloroacetate in the fetal Long-Evans rat. Teratology
46:225-235.
Feldman, RG; Chirico-Post, J; Proctor, SP. (1988) Blink reflex latency after exposure to
trichloroethylene in well water. Arch Environ Health 43(2): 143-148.
Feldman, RG; Niles, C; Proctor, SP; et al. (1992) Blink reflex measurement of effects of
trichloroethylene exposure on the trigeminal nerve. Muscle Nerve 15(4):490-495.
Fisher JW, Whittaker TA, Taylor DH, Clewell HJ 3rd, Andersen ME. (1989) Physiologically
based pharmacokinetic modeling of the pregnant rat: a multiroute exposure model for
trichloroethylene and its metabolite, trichloroacetic acid. Toxicol Appl Pharmacol
99(3):395-414.
Fisher JW, Whittaker TA, Taylor DH, Clewell HJ 3rd, Andersen ME. (1990) Physiologically
based pharmacokinetic modeling of the lactating rat and nursing pup: a multiroute
exposure model for trichloroethylene and its metabolite, trichloroacetic acid. Toxicol
Appl Pharmacol 102(3): 497-513.
Fisher J, Mahle D, Bankston L, Greene R, Gearhart J. (1997) Lactational transfer of volatile
chemicals in breast milk. Am Ind Hyg Assoc J 58(6):425-431.
Fisher JW, Mahle D, Abbas R. (1998) A human physiologically based pharmacokinetic model
for trichloroethylene and its metabolites, trichloroacetic acid and free trichloroethanol.
Toxicol Appl Pharmacol 152(2):339-359.
6/22/2009
1006
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Fisher, JW; Channel, SR; Eggers, JS; Johnson, PD; MacMahon, KL; Goodyear, CD; Sudberry,
GL; Warren, DA; Latendesse, JR; Graeter, LJ. (2001) Trichloroethylene, trichloroacetic
acid, and dichloroacetic acid: do they affect fetal rat heart development? Int J Toxicol
20:257-267.
Forkert PG, Forkert L. 1994. Trichloroethylene induces pulmonary fibrosis in mice. Can J
Physiol Pharmacol. 72: 205-210.
Forkert, P-G; Lash, LH; Nadeau V; Tardif, R; Simmonds, A. (2002) Metabolism and toxicity of
trichloroethylene in epididymis and testis. Toxicol Appl Pharmacol 182:244-254.
Forkert PG, Lash L, Tardif R, Tanphaichitr N, Vandevoort C, Moussa M. (2003) Identification
of trichloroethylene and its metabolites in human seminal fluid of workers exposed to
trichloroethylene. Drug Metab Dispos 31(3):306—311.
Fredriksson, A; Danielsson, BRG; Eriksson, P. (1993) Altered behavior in adult mice orally
exposed to tri- and tetrachloroethylene as neonates. Toxicology Letters 66:13-19.
Fukuda, K; Takemoto, K; Tsuruta, H. (1983) Inhalation carcinogenicity of trichloroethylene in
mice and rats. Ind Health 21:243-254.
Gamberale, F; Annwall, G; Olson, BA. (1976) Exposure to trichloroethylene III. Psychological
functions. Scand J Work Environ Health 2(4):220-224.
Garabrant DH, Held J, Langholz B, Bernstein L. (1988) Mortality of aircraft manufacturing
workers in southern California. Am J Ind Med 13:683-693.
Garabrant DH, Lacey JV Jr, Laing TJ, Gillespie BW, Mayes MD, Cooper BC, Schottenfeld D.
(2003) Scleroderma and solvent exposure among women. Am J Epidemiol
157(6):493-500.
Gash, DM; Rutland, K; Hudson, NL; et al. (2007) Trichloroethylene: Parkinsonism and complex
1 mitochondrial neurotoxicity. Ann Neurol .
George, JD; Reel, JR; Myers, CB; Lawton, AD; Lamb, JC. (1985) Trichloroethylene:
reproduction and fertility assessment in CD-I mice when administered in the feed. RTI
Project No. 310-2344, NTP-86-068. National Institute of Environmental Health Sciences,
National Toxicology Program, RTP, NC.
6/22/2009
1007
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
George, JD; Reel, JR; Myers, CB; Lawton, AD; Lamb, JC. (1986) Trichloroethylene:
reproduction and fertility assessment in F344 rats when administered in the feed. RTI
Project No. 310-2344, NTP-86-085. National Institute of Environmental Health Sciences,
National Toxicology Program, RTP, NC.
George MH, Moore T, Kilburn S, Olson GR, DeAngelo AB. 2000. Carcinogenicity of Chloral
hydrate administered in drinking water to the Male F344/N rat and Male B6C3F1 mouse.
Toxicologic Pathology 28(4): 610-618.
Ghantous, H; Danielsson, BRG; Dencker, L; Gorczak J; Vesterberg, O. (1985) Trichloroacetic
acid accumulates in murine amniotic fluid after tri- and tetrachloroethylene inhalation.
Acta Pharmacol Toxicol (Copenh) 58:105-114.
Gilkeson GS, Keil D, Peden-Adams MM. 2004. Immune effects of trichloroethylene on
autoimmune disease in mice. In: Trichloroethylene - The Scientific Basis of Risk
Assessment (Mohr LC, Hoel DG, Jollow D, eds). Charleston, S.C.: The Medical
University of South Carolina Press, 87-98.
Goel, SK; Rao, GS; Pandya, KP; et al. (1992) Trichloroethylene toxicity in mice: a biochemical,
hematological and pathological assessment. Indian J Exp Biol 30:402-406.
Goh VH, Chia SE, Ong CN. (1998) Effects of chronic exposure to low doses of trichloroethylene
on steroid hormone and insulin levels in normal men. Environ Health Perspect
106(l):41-44.
Goldberg SJ, Lebowitz MD, Graver EJ, Hicks S. (1990) An association of human congenital
cardiac malformations and drinking water contaminants. J Am Coll Cardiol 16(1):
155-164.
Goldsworthy, TL; Popp, JA. (1987) Chlorinated hydrocarbon-induced peroxisomal enzyme
activity in relation to species and organ carcinogenicity. Toxicol Appl Pharmacol
88:225-233.
Grandjean E, Miinchinger R, Turrian V, Haas PA, Knoepfel HK, Rosenmund H. (1955)
Investigations into the effects of exposure to trichloroethylene in mechanical engineering.
Brit J Industr Med 12:131-142.
6/22/2009
1008
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Green T, Dow J, Ong CN, Ng V, Ong HY, Zhuang ZX, Yang XF, Bloemen L. 2004.
Biological monitoring of kidney function among workers occupationally exposed to
trichloroethylene. Occup Environ Med 61:312-317.
Green T. (2000). Pulmonary toxicity and carcinogenicity of trichloroethylene: species
differences and modes of action. Environ Health Perspect. 108 Suppl 2:261-4.
Greenland S, Salvan A, Wegman DH, Hallock MF, Smith TH. (1994) A case-control study of
cancer mortality at the transformer-assembly facility. Int Arch Occup Environ Health
66:49-54.
Griffin JM, Blossom SJ, Jackson SK, Gilbert KM, Pumford NR. 2000a. Trichloroethylene
accelerates an autoimmune response by Thl T cell activation in MRL +/+ mice.
Immunopharmacology 46:123-137.
Griffin JM, Gilbert KM, Lamps LW, Pumford NR. 2000b. CD4(+) T-cell activation and
induction of autoimmune hepatitis following trichloroethylene treatment in MRL+/+
mice. Toxicol Sci 57:345-352.
Gun, RT; Grygorcewicz, C; Nettelbeck, TJ. (1978) Choice reaction time in workers using
trichloroethylene. Med J Aust 1 (10):535—536.
Guyton KZ, Chiu WA, Bateson TF, Jinot J, Scott CS, Brown RC, Caldwell JC. 2009. A
Reexamination of the PPAR-a Activation Mode of Action as a Basis for Assessing
Human Cancer Risks of Environmental Contaminants. Environ Health Perspect.
doi: 10.1289/ehp.0900758 available via http://dx.doi.org/ [Online 15 May 2009]
Hamada T, Tanaka H. (1995) Transfer of methyl chloroform, trichloroethylene and
tetrachloroethylene to milk, tissues and expired air following intraruminal or oral
administration in lactating goats and milk-fed kids. Environ Pollut 87(3):313—8.
Hansen J, Raaschou-Nielsen O, Christensen JM, Johansen I, McLaughlin JK, Lipworth L, Blot
WJ, Olsen JH. (2001) Cancer incidence among Danish workers exposed to
trichloroethylene. J Occup Environ Med 43:133-139.
Hardell L, Bengtsson NO, Johsson U, Eriksson S, Larsson LG. 1984. Aetiological aspects o
primary liver cancer with special regard to alcohol, organic solvents and acute
intermittent porphyria - an epidemiological investigation. Br J Cancer 50:389-297.
6/22/2009
1009
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Hardin, BD; Bond, GP; Sikov, MR; Andrew, FD; Beliles, RP; and Niemeier, RW. (1981)
2 Testing of selected work place chemicals for teratogenic potential. Scand J Work
3 Environ Hlth 7(Suppl 45):66-75.
4 Hardin, BD; Kelman, BJ; Brent, RL. (2005) Trichloroethylene and dichloroethylene: a critical
5 review of teratogenicity. Birth Defects Res A Clin Mol Teratol 73:931-955.
6 Healy, TEJ; Poole, TR; Hopper, A. (1982) Rat fetal development and maternal exposure to
7 trichloroethylene 100 p.p.m. British Journal of Anaesthesiology 54:337-340.
8 Helliwell PJ, Hutton AM. (1950) Trichl or ethylene anaesthesia. Anaesthesia 5(1): 4-13.
9 Henschler D, Romen W, Elsasser HM, Reichaert D, Eder E, Radwan Z. 1980. Carcinogenicity
10 study of trichloroethylene by longterm inhalation in three animal species. Arch Toxicol
11 43:237-248 (1980).
12 Henschler D, Elsasser H, Romen W, Eder E. 1984. Carcinogenicity study of trichloroethylene,
13 with and without epoxide stabilizers, in mice. J Cancer Res Clin Oncol 107:149-156.
14 Henschler D, Vamvakas S, Lammert M, Dekant W, Kraus B, Thomas B, Ulm K. 1995.
15 Increased incidence of renal cell tumors in a cohort of cardboard workers exposed to
16 trichloroethene. Arch Toxicol 69:291-299.
17 Herren-Freund SL, Pereiira MA, Khoury MD, Olson G. 1987. The carcinogenicity of
18 trichloroethylene and its metabolites trichloroacetic acid and dichloroacetic acid, in
19 mouse liver. Toxicol Appl Pharmacol. 90: 183-189.
20 Hill, AB. (1965) The environment and disease: association or causation? Proc R Soc Med
21 58:295-300.
22 Hirsch, AR; Rankin, KM; Panelli, PP. (1996) Trichloroethylene exposure and headache.
23 Headache Quarterly 7(2): 126-13 8.
24 Huang HL, Li LY, Chen B, Huang JX, Huang SR. 2002. New problems caused by occupational
25 trichlorethylene exposure. Int J Immunopath Pharm 15:30-32.
26 Iavicoli I, Marinaccio A, Carelli G. 2005. Effects of occupational trichloroethylene exposure on
27 cytokine levels in workers. J Occup Environ Med 47:453-457.
6/22/2009
1010
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Inoue O, Seiji K, Kawai T, Jin C, Liu YT, Chen Z, Cai SX, Yin SN, Li GL, Nakatsuka H,
Watanabe T, Ikenda M. (1989) Relationship between vapor exposure and urinary
metabolite excretion among workers exposed to trichloroethylene. Am J Ind Med
15(1): 103—10.
Isaacson, LG; Taylor, DH. (1989) Maternal exposure to 1,1,2-trichloroethylene affects myelin in
the hippocampal formation of the developing rat. Brain Res 488(l-2):403-407.
Isaacson, LG; Spohler, SA; Taylor, DH. (1990) Trichloroethylene affects learning and decreases
myelin in the rat hippocampus. Neurotoxicol Teratol 12(4):375—381.
Ito Y, Yamanoshita O, Asaeda N, Tagawa Y, Lee CH, Aoyama T, Ishihara G, Furuhashi K,
Kamijima M, Gonzalez FJ, Nakajima T. 2007. Di(2-ethylhexyl)phthalate induced hepatic
tumorigenesis through a peroxisome proliferator-activated receptor a-independent
pathway. J Occup Health 49:172-182.
Johnson, PD; Dawson, BV; Goldberg, SJ. (1998a) Cardiac teratogenicity of trichloroethylene
metabolites. JACC 72(2):540-545.
Johnson, PD; Dawson, BV; Goldberg, SJ. (1998b) A review: trichloroethylene metabolites:
potential cardiac teratogens. Environmental Health Perspectives 106(suppl. 4):995-999.
Johnson, PD; Goldberg, SJ; Mays, MZ; Dawson, BV. (2003) Threshold of trichloroethylene
contamination in maternal drinking waters affecting fetal heart development in the rat.
Environ Health Perspect 111(3):289-292.
Johnson, PD; Goldberg, SJ; Mays, MZ; Dawson, BV. (2005) Errata. Threshold of
trichloroethylene contamination in maternal drinking waters affecting fetal heart
development in the rat. Environ Health Perspect 113(1): A18.
Kamijima M, Hisanaga N, Wang H, Nakajima T. 2007. Occupational trichloroethylene exposure
as a cause of idiosyncratic generalized skin disorders and accompanying hepatitis similar
to drug hypersensitivities. Int Arch Occup Environ Health 80:357-370.
Kamijima M, Wang H, Huang H, Li L, Shibata E, Lin B, et al. 2008. Trichloroethylene Causes
Generalized Hypersensitivity Skin Disorders Complicated by Hepatitis. J Occup Health.
Kan, FW; Forkert, PG; Wade, MG. (2007) Trichloroethylene exposure elicits damage in
epididymal epithelium and spermatozoa in mice. Histol Histopathol 22(9):977-988.
6/22/2009
1011
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Kaneko T, Wang P-Y, Sato A. 1994. Enzymes induced by ethanol differently affect the
pharmacokinetics of trichloroethylene and 1,1,1-trichloroethane. Occupational and
Environmental Medicine 51:113-119.
Kilburn, KH. (2002a): Is neurotoxicity associated with environmental trichloroethylene (TCE)?
Arch Environ Health 57(2): 113-120.
Kilburn, KH. (2002b): Do duration of exposure, proximity to electronic manufacturing plants,
and involvement in a lawsuit affect chlorinated solvent toxicity? Arch Environ Health
57(2): 121-126.
Kilburn, KH; Thornton, JC. (1996) Prediction equations for simple and visual two-choice
reactions times in environmental neurotoxicology. Arch Environ Health 51(6):439-444.
Kilburn, KH; Warshaw, RH. (1993) Effects on neurobehavioral performance of chronic exposure
to chemically contaminated well water. Toxicol Ind Health 9(3):391-404.
Kilburn, KH. (2002a): Is neurotoxicity associated with environmental trichloroethylene (TCE)?
Arch Environ Health 57(2): 113-120.
Kishi, R; Harabuchi, I; Ikeda, T; et al. (1993) Acute effects of trichloroethylene on blood
concentrations and performance decrements in rats and their relevance to humans. Br J
Ind Med 50(5):470-480.
Kjellstrand P, Kanje M, Mansson L, Bjerkemo M. Mortensen I, Lanke J, Holmquist B. 1981.
Toxicology 21: 105-115.
Kjellstrand P, Holmquist B, Aim P, Kanje M, Romare S, Jonsson I, Mansson L, Bjerkemo M.
(1983a) Trichloroethylene: further studies of the effects on body and organ weights and
plasma butyrylcholinesterase activity in mice. Acta Pharmacol Toxicol (Copenh)
53(5):375-84.
Kjellstrand P, Holmquist B, Mandahl N, Bjerkemo M. (1983b) Effects of continuous
trichloroethylene inhalation on different strains of mice. Acta Pharmacol Toxicol
(Copenh) 53(5):369-74.
Kjellstrand, P; Kanje, M; Bjerkemo, M. (1987) Regeneration of the sciatic nerve in mice and rats
exposed to trichloroethylene. Toxicol Lett 38(1—2): 187—191.
6/22/2009
1012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Klaunig, JE; Babich, MA; Baetcke, KP; et al. (2003) PPARalpha agonist-induced rodent tumors:
modes of action and human relevance. Crit Rev Toxicol 33:655-780.
Krasowski MD, Harrison NL. 2000. The actions of ether, alcohol and alkane general anaesthetics
on GABAA and glycine receptors and the effects of TM2 and TM3 mutations. Br J
Pharmacol. 129(4):731-43.
Kulig, BM. (1987) The effects of chronic trichloroethylene exposure on neurobehavioral
functioning in the rat. Neurotoxicol Teratol 9(2): 171-178.
Kumar, P; Prasad, AK; Saxena, DK; Mani, U; Maji, BK; Dutta, KK. (2000a) Fertility and
general reproduction studies in trichloroethylene exposed rats. Indian J Occup Health
43(3): 117—125.
Kumar, P; Prasad, AK; Dutta, KK. (2000b) Steroidogenic alterations in testes and sera of rats
exposed to trichloroethylene (TCE) by inhalation. Hum Exp Toxicol 19:117-121.
Kumar, P; Prasad, AK; Mani, U; Maji, BK; Dutta, KK. (2001) Trichloroethylene induced
testicular toxicity in rats exposed by inhalation. Hum Exp Toxicol 20:585-589.
Lagakos SW, Wessen BJ, Zelen M. 1986. Analysis of contaminated well water and health effect
in Woburn, Massachusetts. Journal of the American Statistical Association 81:583-596.
Laham, S. (1970) Studies on placental transfer. Trichloroethylene. IMS IndMed Surg. 39(1):
46-49.
Land, PC; Owen, EL; Linde, HW. (1981) Morphologic changes in mouse spermatozoa after
exposure to inhalational anesthetics during early spermatogenesis. Anesthesiology
54:53-56.
Larson JL, Bull RJ. (1989) Effect of ethanol on the metabolism of trichloroethylene. J Toxicol
Environ Health 28(4):395-406.
Lash LH, Fisher JW, Lipscomb JC, Parker JC. (2000) Metabolism of trichloroethylene. Environ
Health Perspect 108 Suppl 2:177-200.
Laslo-Baker D, Barrera M, Knittel-Keren D, Kozer E, Wolpin J, Khattak S, Hackman R, Rovet
J, Koren G. (2004) Child neurodevelopmental outcome and maternal occupational
exposure to solvents. Arch Pediatr Adolesc Med 158(10):956—61.
6/22/2009
1013
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Laughter, AR; Dunn, CS; Swanson, CL; et al. (2004) Role of the peroxisome proliferator-
activated receptor alpha (PPARalpha) in responses to trichloroethylene and metabolites,
trichloroacetate and dichloroacetate in mouse liver. Toxicology 203:83-98.
Leakey JEA, Seng JE, Lantendresse JR, Hussain N, Allen LJ, Allaben WT. 2003a. Dietary
controlled carcinogenicity study of chloral hydrate in male B6C3F1 mice. Toxicology
and Applied Pharmacology 193: 266-280.
Lehmann I, Rehwagen M, Diez U, Seiffart A, Rolle-Kampczyk U, Richter M, et al. (2001)
Enhanced in vivo IgE production and T cell polarization toward the type 2 phenotype in
association with indoor exposure to VOC: results of the LARS study. Int J Hyg Environ
Health 204(4): 211-221.
Lehmann I, Thoelke A, Rehwagen M, Rolle-Kampczyk U, Schlink U, Schulz R, et al. (2002)
The influence of maternal exposure to volatile organic compounds on the cytokine
secretion profile of neonatal T cells. Environ Toxicol 17(3):203-210.
Leigh JP, Jiang W. 1993. liver cirrhosis deaths within occupations and industries in the
California occupational mortality study. Addiction 88:767-779.
Lindbohm ML, Taskinen H, Sallmen M, Hemminki K. (1990) Spontaneous abortions among
women exposed to organic solvents. Am J Ind Med 17(4):449-463.
Lipscomb JC, Garrett CM, Snawder JE. (1997) Cytochrome P450-dependent metabolism of
trichloroethylene: interindividual differences in humans. Toxicol Appl Pharmacol
142(2):311-8.
Liu, YT; Jin, C; Chen, Z; et al. (1988) Increased subjective symptom prevalence among workers
exposed to trichloroethylene at sub-OEL levels. Tohoku J Exp Med 155(2): 183—195.
Loeber, CP; Hendrix, MJC; dePinos, SD; Goldberg, SJ. (1988) Trichloroethylene: a cardiac
teratogen in developing chick embryos. Pediatric Research 24(6):740-744.
Lopreato GF, Phelan R, Borghese CM, Beckstead MJ, Mihic SJ. (2003) Inhaled drugs of abuse
enhance serotonin-3 receptor function. Drug Alcohol Depend. 70(1): 11-5.
Lorente C, Cordier S, Bergeret A, De Walle HE, Goujard J, Ayme S, et al. (2000) Maternal
occupational risk factors for oral clefts. Occupational Exposure and Congenital
Malformation Working Group. Scand J Work Environ Health 26(2): 137-145.
6/22/2009
1014
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Lowengart RA, Peters JM, Cicioni C, Buckley J, Bernstein L, Preston-Martin S, Rappaport E.
(1987) Childhood leukemia and parents' occupational and home exposures. J Natl Cancer
Inst 79(l):39-46.
Lybarger JA, Lichtveld MY, Amler RW. 1999. Biomedical testing of the kidney for persons
exposed to hazardous substances in the environment. Renal Failure 21:263-274.
MADPH (Massachusetts Department of Public Health). (1997) Woburn childhood leukemia
follow-up study. Volume 1: Analyses.
Mahle DA, Gearhart JM, Grigsby CC, Mattie DR, Barton HA, Lipscomb JC, Cook RS. (2007)
Age-dependent partition coefficients for a mixture of volatile organic solvents in
Sprague-Dawley rats and humans. J Toxicol Environ Health A 70(20): 1745-51.
Maltoni, C; Lefemine, G; Cotti, G. (1986) Experimental research on trichloroethylene
carcinogenesis. In: Maltoni, C; Mehlman MA., eds. Vol. 5. Archives of research on
industrial carcinogenesis. Princeton, NJ: Princeton Scientific Publishing;
Manson, JM; Murphy, M; Richdale, N; Smith, MK. (1984) Effects of oral exposure to
trichloroethylene on female reproductive function. Toxicology 32:229-242.
McCarver DG, Byun R, Hines RN, Hichme M, Wegenek W. (1998) A genetic polymorphism in
the regulatory sequences of human CYP2E1: association with increased chlorzoxazone
hydroxylation in the presence of obesity and ethanol intake. Toxicol Appl Pharmacol
152(1):276—81.
McCunney, R.J., Diverse manifestations of trichloroethylene (1988). Br J Ind Med 45:122-126.
McKinney PA, Alexander FE, Cartwright RA, Parker L. (1991) Parental occupations of children
with leukaemia in west Cumbria, north Humberside, and Gateshead. BMJ
302(6778):681-7.
Melnick, RL; Jameson, CW; Goehl, TJ; et al. (1987) Application of microencapsulation for
toxicology studies. II. Toxicity of microencapsulated trichloroethylene in Fischer 344
rats. Fundam Appl Toxicol 8:432-442.
Merrick BA, Robinson M, Condi LW. 1989. Differing hepatotoxicity and lethality after
subacute trichloroethylene exposure in aqueous or corn oil gavage vehicles in B6C3F1
mice. Journal of Applied Toxicology. 9(1): 15—21.
6/22/2009
1015
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Mensing, T; Welge, P; Voss, B; et al. (2002) Renal toxicity after chronic inhalation exposure of
rats to trichloroethylene. Toxicol Lett 128:243-247.
Mhiri, C; Choyakh, F; Ben, HM; et al. (2004) Trigeminal somatosensory evoked potentials in
trichloroethylene-exposed workers. Neurosciences 9(2): 102-107.
Miligi L, Costantini AS, Benvenuti A, Kriebel D, Bolejack V, Tumino R, et al. 2006.
Occupational exposure to solvents and the risk of lymphomas. Epidemiology
17:552-561.
Mirsalis JC, Tyson CK, Steinmetz KL, Loh EK, Hamilton CM, Bakke JP, Spalding JW. 1989.
Measurement of unscheduled DNA synthesis and S-phase synthesis in rodent hepatocytes
following in vivo treatment: testing of 24 compounds. Environmental and Molecular
Mutagenesis 14: 155-164.
Mishima, N; Hoffman, S; Hill, EG; Krug, EL. (2006) Chick embryos exposed to
trichloroethylene in an ex ovo culture model show selective defects in early endocardial
cushion tissue formation. Birth Defects Res A Clin Mol Teratol 76:517-527.
Mitchell, A. B. S., and Parsons-Smith, B. G., (1969) Trichloroethylene Neuropathy. Brit. med. J.
1, 422-423.
Monster AC, Boersma G, Duba WC. (1979) Kinetics of trichloroethylene in repeated exposure
of volunteers. Int Arch Occup Environ Health 42(3-4):283-92.
Moore MM, Harrington-Brock K. (2000) Mutagenicity of trichloroethylene and its metabolites:
implications for the risk assessment of trichloroethylene. Environ Health Perspect. 108
Suppl 2:215-23.
Morgan JW, Cassady RE. (2002) Community cancer assessment in response to long-time
exposure to perchlorate and trichloroethylene in drinking water. J Occup Environ Med.
44(7):616-621.
Morgan RW, Kelsh MA, Zhao K, Heringer S. 1998. Mortality of aerospace workers exposed to
trichloroethylene. Epidemiology 9:424-431.
Moser, Cheek & MacPhail. A multidisciplinary approach to toxicological screening III.
Neurobehavioral toxicity. J Toxicol. Environ. Hlth., 1995, 45, 173-210.
6/22/2009
1016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Moser, VC; MacPhail, RC; Gennings, C. (2003) Neurobehavioral evaluations of mixtures of
trichloroethylene, heptachlor, and di(2-ethylhexyl)phthlate in a full-factorial design.
Toxicology 188(2-3): 125-137.
Miiller G, Spassowski M, Henschler D. (1975) Metabolism of trichloroethylene in man. III.
Interaction of trichloroethylene and ethanol. Arch Toxicol 33(3): 173-89.
Nagaya T, Ishikawa N, Hata H, Otobe T. 1993. Subclinical and reversible hepatic effects of
occupational exposure to trichloroethylene Int Arch Occup Environ Health 64:561-563.
Neghab M, Qu Sm, Bai CL, Caples J, Stacey NH. 1997. Raised concentration of serum bile
acids following occupational exposure to halogenated solvents, 1,1,2-trichloro-1,2,2-
trifluoroethane and trichloroethylene. Int Arch Occup Environ Health 70: 187-194.
Nakajima, T; Okino, T; Okuyama, S; et al. (1988) Ethanol-induced enhancement of
trichloroethylene metabolism and hepatotoxicity: difference from the effect of
phenobarbital. Toxicol Appl Pharmacol 94:227-237.
Nakajima T, Wang RS, Murayama N, Sato A. (1990) Three forms of trichloroethylene-
metabolizing enzymes in rat liver induced by ethanol, phenobarbital, and 3-
methylcholanthrene. Toxicol Appl Pharmacol 102(3):546-52.
Nakajima T, Wang RS, Elovaara E, Park SS, Gelboin HV, Vainio H. (1992b) A comparative
study on the contribution of cytochrome P450 isozymes to metabolism of benzene,
toluene and trichloroethylene in rat liver. Biochem Pharmacol 43(2):251-7.
Nakajima T, Kamijo Y, Usuda N, Liang Y, Fukushima Y, Kametani K, Gonzalez FJ, Aoyama T.
(2000) Sex-dependent regulation of hepatic peroxisome proliferation in mice by
trichloroethylene via peroxisome proliferator-activated receptor alpha (PPARalpha).
Carcinogenesis 21(4):677-82.
Narotsky, MG; Kavlock, RJ. (1995) A multidisciplinary approach to toxicological screening: II.
Developmental toxicity. Journal of Toxicology and Environmental Health 45:145-171.
Narotsky, MG; Weller, EA; Chinchilli, VM; Kavlock, RJ. (1995) Nonadditive developmental
toxicity in mixtures of trichloroethylene, di(2-ethylhexyl) phthalate, and heptachlor in a 5
x 5 x 5 design. Fundam Appl Toxicol 27:203-216.
6/22/2009
1017
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
NCI (National Cancer Institute). (1976) Carcinogenesis bioassay of trichloroethylene. Division
of Cancer Cause and Prevention, National Cancer Institute, U.S. Department of Health,
Education, and Welfare, DHEW Publication No. (NIH) 76-802, Technical Report Series
No. 2, 218 pages; NCI-CG-TR-2; NTIS PB254122.
http://ntp.niehs.nih.gov/ntp/htdocs/LT_rpts/tr002.pdf.
Neghab M, Qu Sm, Bai CL, Caples J, Stacey NH. 1997. Raised concentration of serum bile
acids following occupational exposure to halogenated solvents, 1,1,2-trichloro-1,2,2-
trifluoroethane and trichloroethylene. Int Arch Occup Environ Health 70: 187-194.
Nietert PJ, Sutherland SE, Silver RM, Pandey JP, Knapp RG, Hoel DG, Dosemeci M. (1998) Is
occupational organic solvent exposure a risk factor for scleroderma? Arthritis Rheum
41(6): 111 1—8. Erratum in: Arthritis Rheum 1998 Aug;41(8):1512.
Niklasson, Tham, Larsby, & Eriksson, Effects of toluene, styrene, trichloroethylene and
trichloroethane on the vestibule and opto-oculo motor system in rats, Neurotoxicol.
Teratol., 1993, 15, 327-334.
Noland-Gerbec, EA; Pfohl, RJ; Taylor, DH; Bull RJ. (1986) 2-Deoxyglucose uptake in the
developing rat brain upon pre- and postnatal exposure to trichloroethylene.
Neurotoxicology 7(3): 157-164.
National Research Council (NRC). 2006. Assessing the Human Health Risks of
Trichloroethylene: Key Scientific Issues. Washington DC: National Academy Press.
NTP (National Toxicology Program). (1988) Toxicology and carcinogenesis studies of
trichloroethylene (CAS no. 79-01-6) in four strains of rats (ACI, August, Marshall,
Osborne-Mendel) (gavage studies). Public Health Service, U.S. Department of Health
and Human Services; NTP TR-273; NIH Publication No. 88-2529. Available from the
National Institute of Environmental Health Sciences, Research Triangle Park, NC, and
the National Technical Information Service, Springfield, VA; PB88-218896.
http://ntp.niehs.nih.gov/ntp/htdocs/LT_rpts/tr273.pdf.
NTP (National Toxicology Program). (1990) Carcinogenesis Studies of Trichloroethylene
(Without Epichlorohydrin) (CAS No. 79-01-6) in F344/N Rats and B6C3F1 Mice
(Gavage Study). NTP TR 243. Research Triangle Park, NC: U.S Department of Health
and Human Services.
6/22/2009
1018
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Nunes J, Ehrich M, Robertson J. 2001. Toxicosis associated with dual oral exposure of rats to
lead and trichloroethylene. Toxicologic Pathology. 29(4): 451-457.
Ohta, M; Saito, T; Saito, K; et al. (2001) Effect of trichloroethylene on spatiotemporal pattern of
LTP in mouse hippocampal slices. Int J Neurosci 111(3-4):257-271.
Okino, T; Nakajima, T; Nakano, M. (1991) Morphological and biochemical analyses of
trichloroethylene hepatotoxicity: differences in ethanol- and phenobarbital-pretreated
rats. Toxicol Appl Pharmacol 108:379-389.
Ou, J; Ou, Z; McCarver, G; Hines, RN; Oldham, KT; Ackerman, AW; Pritchard, KA. (2003)
Trichloroethylene decreases heat shock protein 90 interactions with endothelial nitric
oxide synthase: implications for endothelial cell proliferation. Toxicol Sci 73:90-97.
Peden-Adams MM, Eudaly JG, Heesemann LM, Smythe J, Miller J, Gilkeson GS, et al. (2006).
Developmental immunotoxicity of trichloroethylene (TCE): studies in B6C3F1 mice. J
Environ Sci Health A Tox Hazard Subst Environ Eng 41:249-271.
Peden-Adams MM, EuDaly JG, Lee AM, Miller DE, Keil DE, Gilkeson GS. (2008)
Developmental effects of trichloroethylene (TCE) on immunity: studies in MRL (+/+)
mice. J Environ Sci Health A Tox Hazard Subst Environ Eng A43(12): 1402-9.
Pellizzari ED, Hartwell TD, Harris BS 3rd, Waddell RD, Whitaker DA, Erickson MD. (1982)
Purgeable organic compounds in mother's milk. Bull Environ Contam Toxicol
28(3):322-8.
Pesch B, Haerting J, Ranft U, Klimpel A, Oelschlagel B, Schill W. 2000a. Occupational risk
factors for renal cell carcinoma: agent-specific results from a case-control study in
Germany. MURC Study Group. Multicenter urothelial and renal cancer study. Int J
Epidemiol. 29:1014-1024.
Peters FM, Preston-Martin S, Yu MC. (1981) Brain tumors in children and occupational
exposure of parents. Science 213(4504):235-7.
Peters JM, Garabrant DH, Wright WE, Bernstein L, Mack TM. (1985) Uses of a cancer registry
in the assessment of occupational cancer risks. Natl Cancer Inst Monogr 69:157-61.
6/22/2009
1019
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Povey A, Guppy MJ, Wood M, Knight C, Black CM, Silman AJ. (2001) Cytochrome P2
polymorphisms and susceptibility to scleroderma following exposure to organic solvents.
Arthritis Rheum 44(3):662-5.
Price RG, Taylor SA, Chivers I, Arce-Tomas M, Crutcher E, Franchini I, et al., 1996.
Development and validation of new screening tests for nephrotoxic effects. Human
Environ Toxicol 15, Suppl. 1:510-519.
Price RG, Wedeen R, Lichtveld MY, Lybarger JA, Poter GA, Fels L, et al., 1999. Urinary
biomarkers: roles in risk assessment to environmental and occupational nephrotoxins:
monitoring of effects and evaluation of mechanisms of toxicity. Renal Failure
21:xiii-xviii.
Raaschou-Nielsen O, Hansen J, McLaughlin JK, Kolstad H, Christensen JM, Tarone RE, Olsen
JH. (2003) Cancer risk among workers at Danish companies using trichloroethylene: a
cohort study. Am J Epidemiol 158:1182-1192.
Radican L, Wartenberg D, Rhoads, GG, Schneider D, Wedeen R, Stewart P, Blair A. 2006. A
retrospective occupational cohort study of endOstage renal disease in aircraft workers
exposed to trichloroethylene and other hydrocarbons. J Occup Environ Med 48:1-12.
Radican L, Blair A, Stewart P, Wartenberg D. 2008. Mortality follow-up in hydrocarbon-
exposed workers. J Occup Environ Med 50:1306-1319.
Rasmussen, K; Sabroe, S. (1986) Neuropsychological symptoms among metal workers exposed
to halogenated hydrocarbons. Scand J Soc Med 14(3): 161-168.
Rasmussen K, Brogren C-H, Sabroe S. 1993 a. Subclinical affection of liver and kidney function
and solvent exposure. Int Arch Occup Environ Health 64:445-448.
Rasmussen K, Jeppesen HJ, Sabroe S. (1993b) Psychometric tests for assessment of brain
function after solvent exposure. Am J Industr Med 24:553-565.
Rasmussen K, Arlien-Soborg P, Sabroe S. (1993c) Clinical neurological findings among metal
degreasers exposed to chlorinated solvents. Acta Neurol Scand 87(3):200-204.
Rasmussen K, Sabroe S, Wohlert M, Ingerslev HJ, Kappel B, Nielsen J. (1988) A genotoxic
study of metal workers exposed to trichloroethylene. Sperm parameters and chromosome
aberrations in lymphocytes. Int Arch Occup Environ Health 60(6):419-423.
6/22/2009
1020
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Rebert CS, Day VL, Matteucci MJ, Pryor GT (1991). Sensory-evoked potentials in rats
chronically exposed to trichloroethylene: predominant auditory dysfunction.
Neurotoxicol Teratol. 13(1):83—90
Reif JS, Burch JB, Nuckols JR, Metzger L, Ellington D, Anger WK. (2003) Neurobehavioral
effects of exposure to trichloroethylene through a municipal water supply. Environ Res
93(3):248-58.
Richmond, RE; Carter, JH; Carter, HW; et al. (1995) Immunohistochemical analysis of
dichloroacetic acid (DCA)-induced hepatocarcinogenesis in male Fischer (F344) rats.
Cancer Lett 92:67-76.
Ritz B. (1999) Cancer mortality among workers exposed to chemicals during uranium
processing. J Occup Environ Med. 41:556-566.
Rodenbeck SE, Sanderson LM, Rene A. (2000) Maternal exposure to trichloroethylene in
drinking water and birth-weight outcomes. Arch Environ Health 55(3): 188-194.
Rodriguez CE, Mahle DA, Gearhart JM, Mattie DR, Lipscomb JC, Cook RS, Barton HA. (2007)
Predicting age-appropriate pharmacokinetics of six volatile organic compounds in the rat
utilizing physiologically based pharmacokinetic modeling. Toxicol Sci 98(l):43-56.
Rufer, ES; Hacker, T; Lough, J; Smith, SM. (2008) Low-dose trichloroethylene exposure during
valvuloseptal morphogenesis causes ventricular septal defects in hatched chicks.
Toxicologist 102(S—1):316.
Ruijten, M; Verberk, MM; Salle, HJ. (1991) Nerve function in workers with long term exposure
to trichloroethene. Br J Ind Med 48(2):87-92.
Saihan EM, Burton JL, Heaton KW. (1978) A new syndrome with pigmentation, scleroderma,
gynaecomastia, Raynaud's phenomenon and peripheral neuropathy. Br J Dermatol
99(4):437-440.
Sallmen M, Lindbohm ML, Kyyronen P, Nykyri E, Anttila A, Taskinen H, et al. (1995) Reduced
fertility among women exposed to organic solvents. Am J Ind Med 27(5):699-713.
Sallmen M, Lindbohm ML, Anttila A, Kyyronen P, Taskinen H, Nykyri E, Hemminki K. (1998)
Time to pregnancy among the wives of men exposed to organic solvents. Occup Environ
Med 55(l):24-30.
6/22/2009
1021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Sato A, Nakajima T. (1985) Enhanced metabolism of volatile hydrocarbons in rat liver following
food deprivation, restricted carbohydrate intake, and administration of ethanol,
phenobarbital, polychlorinated biphenyl and 3-methylcholanthrene: a comparative study.
Xenobiotica 15(1):67—75.
Sato A, Nakajima T, Koyama Y. (1980) Effects of chronic ethanol consumption on hepatic
metabolism of aromatic and chlorinated hydrocarbons in rats. Br J Ind Med 37(4):382-6.
Sato A, Nakajima T, Koyama Y. (1981) Dose-related effects of a single dose of ethanol on the
metabolism in rat liver of some aromatic and chlorinated hydrocarbons. Toxicol Appl
Pharmacol 60(1):8—15.
Sato A, Nakajima T, Koyama Y. (1983) Interaction between ethanol and carbohydrate on the
metabolism in rat liver of aromatic and chlorinated hydrocarbons. Toxicol Appl
Pharmacol 68(2):242-9.
Sato A, Endoh K, Kaneko T, Johanson G. (1991a) Effects of consumption of ethanol on the
biological monitoring of exposure to organic solvent vapours: a simulation study with
trichloroethylene. Br J Ind Med 48(8):548-556.
Sato A, Endoh K, Kaneko T, Johanson G. (1991b) A simulation study of physiological factors
affecting pharmacokinetic behaviour of organic solvent vapours. Br J Ind Med
48(5):342-7.
Sato, A. (1993) Confounding factors in biological monitoring of exposure to organic solvents. Int
Arch Occup Environ Health 65(1 Suppl):S61-7.
Saygun M, Cakmak A, Ekici A, Pinar T, Bulcun E, Ulu N, Ekici M. 2007. Five annual
observations of respiratory findings in gun factory workers exposed to solvents. J Occup
Environ Med 49:909-912.
Schwetz, BA; Leong, BKJ; Gehring, PJ. (1975) The effect of maternally inhaled
trichloroethylene, perchloroethylene, methyl chloroform, and methylene chloride on
embryonal and fetal development in mice and rats. Toxicol Appl Pharmacol 32:84-96.
Seidler A, Mohner M, Berger J, Mester B, Deeg E, Eisner G, et al. 2007. Solvent exposure and
malignant lymphoma: a population-based case-control study in Germany. J Occup Med
Toxicol 2:2.
6/22/2009
1022
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Shafer, TJ; Bushnell, PJ; Benignus, VA; et al. (2005) Perturbation of voltage-sensitive Ca2+
2 channel function by volatile organic solvents. J Pharmacol Exp Ther 315(3): 1109-1118.
3 Shih, CL, Chen, HH, Chiu, TH. Acute exposure to trichloroethylene differentially alters the
4 susceptibility to chemoconvulsants in mice. Toxicology, 162, 2001, 35-42.
5 Shu XO, Stewart P, Wen WQ, Han D, Potter JD, Buckley JD, Heineman E, Robison LL. (1999)
6 Parental occupational exposure to hydrocarbons and risk of acute lymphocytic leukemia
7 in offspring. Cancer Epidemiol Biomarkers Prev 8(9):783-791.
8 Smith, GF. (1970) The investigation of the mental effects of trichl or ethylene. Ergonomics
9 13(5):580-586.
10 Smith MK; Randall, JL; Read, EJ; Stober, JA. (1989) Teratogenic activity of trichloroacetic acid
11 in the rat. Teratology 40:445-451.
12 Smith MK; Randall, JL; Read, EJ; Stober, JA. (1992) Developmental toxicity of dichloroacetate
13 in the rat. Teratology 46:217-223.
14 Staples, RE. (1974) Detection of visceral alterations in mammalian fetuses. Teratology
15 9:A37-A38.
16 Stewart, R. D., Dodd, H. C., Gay, H. H., and Erley, D. S. (1970). Experimental human exposure
17 to trichloroethylene. Arch Environ Health 20, 64-71.
18 Stewart RD, Hake CL, Peterson JE. (1974) "Degreaser's flush". Arch Environ Health 29(1): 1-5.
19 Stuckhardt, JL; Poppe, SM. (1984) Fresh visceral examination of rat and rabbit fetuses used in
20 teratogenicity testing. Teratog Carcinog Mutagen 4:181-188.
21 Tang XJ, Li LY, Huang JX, Deng YY. 2002. Guinea pig maximization test for trichloroethylene
22 and its metabolites. Biomed Environ Sci 15:113-118.
23 Tang X, Que B, Song X, Li S, Yang X, Wang H, et al. 2008. Characterization of liver injury
24 associated with hypersensitive skin reactions induced by trichloroethylene in the guinea
25 pig maximization test. J Occup Health 50:114-121.
26 Tao L, Yan S, Xie M, Kramer PM, Pereira MA. 2000. Effect of trichloroethylene and its
27 metabolites, dichloroacetic acid and trichloroacetic acid, on the methylation and
6/22/2009
1023
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
expression of c-Jun and c-Myc protooncogenes in mouse liver: prevention by methionine.
Toxicological Sciences 54: 399-407.
Taskinen H, Anttila A, Lindbohm ML, Sallmen M, Hemminki K. (1989) Spontaneous abortions
and congenital malformations among the wives of men occupationally exposed to organic
solvents. Scand J Work Environ Health 15(5): 345-352.
Taskinen H, Kyyronen P, Hemminki K, Hoikkala M, Lajunen K, Lindbohm ML. (1994)
Laboratory work and pregnancy outcome. J Occup Med 36(3): 311-319.
Taylor, DH; Lagory, KE; Zaccaro, DJ; Pfohl, RJ; Laurie, RD. (1985) Effect of trichloroethylene
on the exploratory and locomotor activity of rats exposed during development. Sci Total
Environ 47:415-420.
Tham, R; Larsby, B; Odkvist, LM; et al. (1979) The influence of trichloroethylene and related
drugs on the vestibular system. Acta Pharmacol Toxicol (Copenh) 44(5):336-342.
Tham, R; BUNNFORS, I; Eriksson, B; et al. (1984) VESTIBULO OCULAR DISTURBANCES
IN RATS EXPOSED TO ORGANIC SOLVENTS. Acta Pharmacologica Et
Toxicologica 54(1):58-63.
Thiele DL, Eigenbrodt EH, Ware AJ. 1982. Cirrhosis after reported trichloroethylene and 1,1,
1-trichl or ethane exposure. Gastroenterol 83:926-929.
Till C, Westall CA, Rovet JF, Koren G. (2001) Effects of maternal occupational exposure to
organic solvents on offspring visual functioning: a prospective controlled study.
Teratology 64:134-141.
Tola S, Vilhunen R, Jarvinen E, Korkala ML. (1980) A cohort study on workers exposed to
trichloroethylene. J Occup Med 22(11):737-740.
Triebig, G., Essing, H.-G., Schaller, K.-H., Valentin, H. (1976). Biochemical and psychological
examinations of test subjects exposed to trichloroethylene. Zbl. Bakt. Hyg., I. Abt. Orig.
B 163: 383-416.
Triebig, G., S. Lehrl, W. Kinzel, H. Erzigkeit, J. V. Galster and K. H. Schaller (1977a).
PSYCHO-PATHOMETRIC RESULTS OF FOLLOW-UP STUDIES OF PERSONS
EXPOSED TO TRICHLOROETHYLENE. Zbl. Bakt. Hyg., I. Abt. Orig. B 164,
314-327 (1977). (translated from German)
6/22/2009
1024
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Triebig, G; Schaller, KH; Erzigkeit, H; et al. (1977b). Biochemical investigations and
psychological studies of persons chronically exposed to trichloroethylene with regard to
non-exposure intervals. Int Arch Occup Environ Health 38(3): 149-162. (translated from
German)
Triebig, G; Trautner, P; Weltle, D; Saure, E; Valentin, H. (1982). Investigations on the
Neurotoxicity of processing materials III. Measurement of Motor and Sensory Nerve
Conduction Velocity in Persons with Job-Related Trichloroethylene Exposure. Int Arch
Occup Environ Health (1982) 51: 25-34 (translated from German)
Triebig, G., W. Bestler, P. Baumeister , H. Valentin (1983). Investigations on the neurotoxicity
of chemical substances at the workplace, Part IV: Determination of the motor and sensory
nerve conduction velocity in persons occupationally exposed to a mixture of organic
solvents. Int. Arch. Occup. Environ. Health, 52, 139-150. (translated from German)
Troster, AI; Ruff, RM. (1990) Neuropsychological sequelae of exposure to the chlorinated
hydrocarbon solvents trichloroethylene and trichloroethane. Archives of Clinical
Neuropsychology 5(1):31—47.
Tucker AN, Sanders VM, Barnes DW, Bradshaw TJ, White KL, Sain LE, Borzelleca JF,
Munson AE. 1982. Toxicology of trichloroethylene in the mouse. Toxicology and
Applied Pharmacology 62: 351-357.
U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. Risk Assessment Forum,
Washington, DC; EPA/630/P-03/001B. Available from:
.
Umezu, T; Yonemoto, J; Soma, Y; et al. (1997) Behavioral effects of trichloroethylene and
tetrachloroethylene in mice. Pharmacol Biochem Behav 58(3):665-671.
Vamvakas S, Briining T, Thomasson B, Lammert M, Baumiiller A, Bolt HM et al. 1998. Renal
cell cancer correlated with occupational exposure to trichloroethene. J Cancer Res Clin
Oncol 124:374-382.
Van Duuren BL, Goldschmidt BM, Loewengart G, Smith AC, Meichionne S, Seidman I, Roth
D. 1979. Carcinogenicity of halogenated olefinic and aliphatic hydrocarbons in mice.
JNCI 63(6): 1433-1439.
6/22/2009
1025
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Veeramachaneni, DNR; Palmer, JS; Amann, RP. (2001) Long-term effects on male reproduction
of early exposure to common chemical contaminants in drinking water. Hum Reprod
16(5):979-987.
Vernon RJ, Ferguson RK. (1969) Effects of trichloroethylene on visual-motor performance.
Arch Environ Health 18(6):894-900.
Wang G, Cai P, Ansari GA, Khan MF. 2007b. Oxidative and nitrosative stress in
trichloroethene-mediated autoimmune response. Toxicology 229:186-193.
Wang G, Konig R, Ansari GA, Khan MF. 2008. Lipid peroxidation-derived aldehyde-protein
adducts contribute to trichloroethene-mediated autoimmunity via activation of CD4(+) T
cells. Free Radic Biol Med.
Wang R, Zhang Y, Lan Q, Holdford TR, Leaderer B, Zahm SH, Boyle P, et al. 2009.
Occupational exposure to solvents and risk of non-Hodgkin lymphoma in Connecticut
women. Am J Epidemiol 169:176-185.
Warren, DA; Graeter, LJ, Channel, SR; Eggers, JS; Goodyear, CD; Macmahon, KL; Sudberry,
GL; Latendresse, JR; Fisher, JW; Baker, WH. (2006) Trichloroethylene, trichloroacetic
acid, and dichloroacetic acid: do they affect eye development in the Sprague-Dawley rat?
Int J Toxicol 25(4):279-284.
Waseem, M; Ali, M; Dogra, S; et al. (2001) Toxicity of trichloroethylene following inhalation
and drinking contaminated water. J Appl Toxicol 21(6):441-444.
Watson, RE; Jacobson, CF; Williams, AL; Howard, WB; DeSesso, JM. (2006)
Trichloroethylene-contaminated drinking water and congenital heart defects: a critical
analysis of the literature. Reprod Toxicol 21:117-147.
Westergren, I; Kjellstrand, P; Linder, LE; Johansson, BB. (1984) Reduction of brain specific
gravity in mice prenatally exposed to trichloroethylene. Toxicology Letters 23:223-226.
White JF, Carlson GP. (1981) Epinephrine-induced cardiac arrhythmias in rabbits exposed to
trichloroethylene: potentiation by ethanol. Toxicol Appl Pharmacol 60(3):466-71.
White RF, Feldman RG, Eviator, II, Jabre JF, Niles CA. (1997) Hazardous waste and
neurobehavioral effects: a developmental perspective. Environ Res 73(1-2): 113-124.
6/22/2009
1026
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Wiesenhiitter B, Selinski S, Golka K, Briining T, Bolt HM. (2007) Re-assessment of the
influence of polymorphisms of phase-II metabolic enzymes on renal cell cancer risk of
trichloroethylene-exposed workers. Int Arch Occup Environ Health 81:247-251.
Windham GC, Shusterman D, Swan SH, Fenster L, Eskenazi B. (1991) Exposure to organic
solvents and adverse pregnancy outcome. Am J Ind Med 20(2):241-259.
Windham GC, Zhang L, Gunier R, Croen LA, Grether JK. (2006) Autism spectrum disorders in
relation to distribution of hazardous air pollutants in the San Francisco Bay area. Environ
Health Perspect 114(9): 1438-1444.
Withey JR, Karpinski K. (1985) The fetal distribution of some aliphatic chlorinated
hydrocarbons in the rat after vapor phase exposure. Biol Res Pregnancy Perinatol. 6(2):
79-88.
Wolff & Siegmund, the effect of trichloroethylene on the spontaneous locomotor activity of mice
and of tetrachloroethane on the mortality of mice as a function of the time of day. Biol.
Zbl, 97, 345-351, 1978.
Woolhiser, MR; Krieger, SM; Thomas, J; Hotchkiss, JA. (2006) Trichloroethylene (TCE):
Immunotoxicity potential in CD rats following a 4-week vapor inhalation exposure. Dow
Chemical Company, Toxicology & Environmental Research and Consulting, Midland,
MI, Study ID 031020, July 5, 2006, unpublished.
Wu, KL; Berger, T. (2007) Trichloroethylene metabolism in the rat ovary reduces oocyte
fertilizibility. Chemico-Biological Interactions 170:20-30.
Xu, H; Tanphaichitr, N; Forkert, P-G; Anupriwan, A; Weerachatyanukul, W; Vincent, R;
Leader, A; Wade, MG. (2004) Exposure to trichloroethylene and its metabolites causes
impairment of sperm fertilizing ability in mice. Toxicol Sci 82:590-597.
Yang Q, Ito S, Gonzalez FJ. 2007. Hepatocyte-restricted constitutive activation of PPAR alpha
induces hepatoproliferation but not hepatocarcinogenesis. Carcinogenesis. 28(6): 1171 -
1177.
Yauck JS, Malloy ME, Blair K, Simpson PM, McCarver DG. (2004) Proximity of residence to
trichloroethylene-emitting sites and increased risk of offspring congenital heart defects
among older women. Birth Defects Res A Clin Mol Teratol 70(10) 808-814.
6/22/2009
1027
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 Yoon M, Madden MC, Barton HA. (2007) Extrahepatic metabolism by CYP2E1 in PBPK
2 modeling of lipophilic volatile organic chemicals: impacts on metabolic parameter
3 estimation and prediction of dose metrics. J Toxicol Environ Health A 70(18):1527-41.
4 Zenick H, Blackburn K, Hope E, Richdale N, Smith MK. (1984) Effects of trichloroethylene
5 exposure on male reproductive function in rats. Toxicology 31(3-4):237-50.
6 Zhao Y, Krishnadasan A, Kennedy N, Morgenstern H, Ritz B. 2005. Estimated effects of
7 solvents and mineral oils on cancer incidence and Mortality in a cohort of aerospace
8 workers. Am J Ind Med 48:249-258.
9 Zielinski, A. (1973) General health state of women professionally exposed to trichloroethylene
10 vapours. Med Pr 24:263-271.
11
6/22/2009
1028
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5 Dose-Response Assessment
5.1 Dose-Response Analyses for Non-Cancer Endpoints
Because of the large number of non-cancer health effects associated with TCE exposure
and the large number of studies reporting on these effects, a screening process, described below,
was used to reduce the number of endpoints and studies to those that would best inform the
selection of the critical effects for the inhalation reference concentration (RfC) and oral reference
dose (RfD).25 The screening process helped identify the more sensitive endpoints for different
types of effects within each health effect domain (e.g., different target systems) and provided
information on the exposure levels that could contribute to the most sensitive effects, used for the
RfC and RfD, as well as to additional non-cancer effects as exposure increases. These more
sensitive endpoints were also used to investigate the impacts of pharmacokinetic uncertainty and
variability.
The general process used to derive the RfD and RfC was as follows (see Figure 5.1.1):
(1) Consider all studies described in Chapter 4 which report adverse non-cancer health
effects and provide quantitative dose-response data.
(2) Consider for each study/endpoint possible points of departure (PODs) on the basis of
applied dose, with the order of preference being first a benchmark dose (BMD)26 derived
from empirical modeling of the dose-response data, then a no observed adverse effect
level (NOAEL), and lastly a lowest observed adverse effect level (LOAEL).
(3) Adjust each POD by endpoint/study-specific "uncertainty factors" (UFs), accounting for
uncertainties and adjustments in the extrapolation from the study conditions to conditions
of human exposure, to derive candidate RfCs (cRfCs) or RfDs (cRfDs) intended to be
protective for each endpoint (individually) on the basis of applied dose.
(4) Array the cRfCs and cRfDs across the following health effect domains: (i) neurotoxic
effects; (ii) systemic (body weight) and organ toxicity (kidney, liver) effects; (iii)
immunotoxic effects; (iv) reproductive effects; and (v) developmental effects.
(5) Select as candidate critical effects those endpoints with the lowest cRfCs or cRfDs,
within each of these effect domains, taking into account the confidence in each estimate.
25 In EPA non-cancer health assessments, the RfC [RfD] is an estimate (with uncertainty spanning perhaps an order
of magnitude) of a continuous inhalation [daily oral] exposure to the human population (including sensitive
subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime. It can be derived
from a NOAEL, LOAEL, or benchmark concentration [dose], with uncertainty factors generally applied to reflect
limitations of the data used.
26 more precisely, it is the BMDL, i.e., the (one-sided) 95% lower confidence bound on the dose corresponding to
the benchmark response for the effect, that is used as the POD
6/22/2009
1029
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
When there are alternative estimates available for a particular endpoint, preference is
given to studies whose design characteristics (e.g., species, statistical power, exposure
level(s) and duration, endpoint measures) are better suited for determining the most
sensitive human health effects of chronic TCE exposure.
(6) For each candidate critical effect selected in step (5), use, to the extent possible, the
PBPK model developed in Section 3.5 to calculate an internal dose POD (iPOD) for
plausible internal dose metrics that were selected on the basis of what is understood about
the role of different TCE metabolites in toxicity and the MOA for toxicity.
(7) For each iPOD for each candidate critical effect, use the PBPK model to estimate inter-
species and within-human pharmacokinetic variability (or just within-human variability
for human-based PODs). The results of this calculation are 99th percentile estimates of
the human equivalent concentration and human equivalent dose (HEC99 and HED99) for
each candidate critical effect.
(8) Adjust each HEC99 or HED99 by endpoint/study-specific UFs [which, due to the use of
the PBPK model, may differ from the UFs used in step (3)] to derive a PBPK model-
based candidate RfCs (p-cRfC) and RfD (p-cRfD) for each candidate critical effect.
(9) Characterize the uncertainties in the cRfCs, cRfDs, p-cRfCs, and p-cRfDs, with the
inclusion of quantitative uncertainty analyses of pharmacokinetic uncertainty and
variability as derived from the Bayesian population analysis using the PBPK model; and
(10) Evaluate the most sensitive cRfCs, p-cRfCs, cRfDs, and p-cRfDs, taking into account the
confidence in the estimates, to arrive at a RfC and RfD for TCE.
In contrast to the approach used in most assessments, in which the RfC and RfD are each based
on a single critical effect, the final RfC and RfD for TCE were based on multiple critical effects
that resulted in very similar candidate RfC and RfD values at the low end of the full range of
values. This approach was taken here because it provides robust estimates of the RfC and RfD
and because it highlights the multiple effects that are all yielding very similar candidate values.
The results of this process are summarized in the sections below, with technical details presented
in Appendix F.
6/22/2009
1030
-------
1
2
3
4
5
6
7
8
9
10
11
12
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
studies
(2^ Points of
Departure
^(PODs)
Wl~
Apply
'Uncertainty
Factors
(UFs)
Candidate RfC
(cRfCs) &
candidate RfDs
(cRfDs)
[applied dose]
Xt)
Lowest
values within
each domain
©
Candidate
critical effects/
studies,
cRfCs & cRfDs
Apply PBPK
model;
Update UFs
PBPK-based
candidate RfCs
(p-cRfDs) &
candidate RfDs,
(p-cRfDs)
(g\Consider and evaluate most sensitive
^ estimates across domains and their
uncertainties
RfC and RfD for non-
cancer effects
Figure 5.1.1
Flow-chart of the process used to derive the RfD and RfC for non-cancer effects.
5.1.1 Modeling approaches and uncertainty factors for developing candidate reference
values based on applied dose
This section summarizes the general methodology used with all the TCE studies and
endpoints for developing cRfCs and cRfDs on the basis of applied dose. A detailed discussion of
the application of these approaches to the studies and endpoints for each health effect domain
follows in the next section (5.1.2).
Standard adjustments27 were made to the applied doses to obtain continuous inhalation
exposures and daily average oral doses over the study exposure period (see Appendix F for
27 Discontinuous exposures (e.g., gavage exposures once a day, 5 days/week, or inhalation exposures for 5
days/week, 6 hrs/day) were adjusted to the continuous exposure yielding the same cumulative exposure. For
inhalation studies, these adjustments are equivalent to those recommended by U.S. EPA (1994) for deriving a human
equivalent concentration for a Category 3 gas for which the blood:air partition coefficient in laboratory animals is
greater than that in humans (see Section 3.1 for discussion of the TCE blood:air partition coefficient).
6/22/2009
1031
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
details), except for effects for which there was sufficient evidence that the effect was more
closely associated with administered exposure level (e.g., changes in visual function). The PODs
based on applied dose in the following sections and in Appendix F are presented in terms of the
adjusted doses (except where noted).
As described above, wherever possible28, benchmark dose modeling was conducted to
obtain BMDLs to serve as PODs for the cRfCs and cRfDs. Note that not all quantitative dose-
response data are amenable to benchmark dose modeling. We did not consider, for example,
non-numerical data (e.g., data presented in line or bar graphs rather than in tabular form). In
addition, sometimes the available models used do not provide an adequate fit to the data. For the
benchmark dose modeling for this assessment, we used U.S. EPA's BenchMark Dose Software
(BMDS), which is freely available at www.epa.gov/ncea/bmds. For dichotomous responses, we
fitted the loglogistic, multistage, and Weibull models. This subset of BMDS dichotomous
models was used to reduce modeling demands, and these particular models were selected
because, as a group, they have been found to be capable of describing the great majority of dose-
response datasets, and specifically for some TCE datasets (Filipsson and Victorin, 2003). For
continuous responses, we fitted the distinct models available in BMDS - the power, polynomial,
and Hill models. For some reproductive and developmental datasets, we also fitted two nested
models (the nested logistic and the Rai & Van Ryzin models in BMDS29) to examine and
account for potential intra-litter correlations. Models with unconstrained power parameters <1
were considered when the dose-response relationship appeared supralinear, but these models
often yield very low BMDL estimates and there was no situation in which an unconstrained
model with a power parameter <1 was selected for the datasets modeled here. In most cases, a
constrained model or the Hill model provided an adequate fit to such a dose-response
relationship. In a few cases, the highest dose group was dropped to obtain an improved fit to the
lower dose groups. See Appendix F for further details on model fitting and parameter
constraints.
After fitting these models to the datasets, we applied the recommendations for model
selection set out in EPA's Benchmark Dose Technical Guidance Document (Inter-Agency
Review Draft, US EPA, 2008b). First, models were rejected if the p-value for goodness of fit
was < 0.10.30 Second, models were rejected if they did not appear to adequately fit the low-dose
region of the dose-response relationship, based on an examination of graphical displays of the
28 An exception was for the systemic effect of decreased body weight, which was observed in multiple chronic
studies. Dose-response data were available, but the resources were not invested into modeling these data because
the endpoint appeared a priori to be less sensitive than others and was not expected to be a critical effect.
29 the NCTR model failed with the TCE datasets.
30 in a few cases in which none of the models fit the data with p>0.10, linear models were selected on the basis of an
adequate visual fit overall.
6/22/2009
1032
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
data and scaled residuals. If the BMDL estimates from the remaining models were "sufficiently
close" (we used a criterion of within 2-fold for "sufficiently close"), then the model with the
lowest AIC was selected.31 If the BMDL estimates from the remaining models are not
sufficiently close, some model dependence is assumed. With no clear biological or statistical
basis to choose among them, the lowest BMDL was chosen as a reasonable conservative
estimate, as suggested in the Benchmark Dose Technical Guidance Document, unless the lowest
BMDL appeared to be an outlier, in which case further judgments were made. Additionally, for
continuous models, constant variance models were used for model parsimony unless the p-value
for the test of homogenous variance was <0.10, in which case the modeled variance models were
considered.
For benchmark response (BMR) selection, we took statistical and biological
considerations into account, in accordance with the Benchmark Dose Technical Guidance
Document (Inter-Agency Review Draft, US EPA, 2008b). For dichotomous responses, our
general approach was to use 10% extra risk as the BMR for borderline or minimally adverse
effects and either 5% or 1% extra risk for adverse effects, with 1% reserved for the most severe
effects. For continuous responses, the preferred approach for defining the BMR is to use a pre-
established cut-point for the minimal level of change in the endpoint at which the effect is
generally considered to become biologically significant (e.g., there is substantial precedence for
using a 10% change in weight for organ and body weights and a 5% change in weight for fetal
weight). In the absence of a well-established cut-point, we generally selected a BMR of 1
(control) standard deviation (SD) change from the control mean, or 0.5 SD for effects considered
to be more serious. For one neurological effect (traverse time), a doubling (i.e., 2-fold change)
was selected because the control SD appeared unusually small.
After the PODs were determined for each study/endpoint, UFs were applied to obtain the
cRfCs and cRfDs. Uncertainty factors are used to address differences between study conditions
and conditions of human environmental exposure (U.S. EPA, 2002). These include:
(a) Extrapolating from laboratory animals to humans: If a POD is derived from experimental
animal data, it is divided by an UF to reflect pharmacokinetic and pharmacodynamic
differences that may make humans more sensitive than laboratory animals. For oral
exposures, the standard value for the interspecies UF is 10, which breaks down
(approximately) to a factor of 3 for pharmacokinetic differences and a factor of 3 for
31 Akaike Information Criteria - a measure of information loss from a dose-response model that can be used to
compare a set of models. Among a specified set of models, the model with the lowest AIC is considered the "best".
If 2 or more models share the lowest AIC, the BMD Technical Guidance Document (US EPA, 2008b) suggests that
an average of the BMDLs could be used, but averaging was not used in this assessment (for the one occasion in
which models shared the lowest AIC, a selection was made based on visual fit).
6/22/2009
1033
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
pharmacodynamic differences. For inhalation exposures, ppm equivalence across species is
generally assumed, in which case pharmacokinetic differences are considered to be
negligible, and the standard value used for the interspecies UF is 3, which is ascribed to
pharmacodynamic differences32. These standard values were used for all the cRfCs and
cRfDs based on laboratory animal data in this assessment.
(b) Human (intraspecies) variability: RfCs and RfDs apply to the human population, including
sensitive subgroups, but studies rarely examine sensitive humans. Sensitive humans could be
adversely affected at lower exposures than a general study population; consequently, PODs
from general-population studies are divided by an UF to address sensitive humans.
Similarly, the animals used in most laboratory animal studies are considered to be "typical"
or "average" responders, and the human (intraspecies) variability UF is also applied to PODs
from such studies to address sensitive subgroups. The standard value for the human
variability UF is 10, which breaks down (approximately) to a factor of 3 for pharmacokinetic
variability and a factor of 3 for pharmacodynamic variability. This standard value was used
for all the PODs in this assessment with the exception of the PODs for a few immunological
effects that were based on data from a sensitive (autoimmune-prone) mouse strain; for those
PODs, an UF of 3 was used for human variability.
(c) Uncertainty in extrapolating from subchronic to chronic exposures: RfCs and RfDs apply to
lifetime exposure, but sometimes the best (or only) available data come from less-than-
lifetime studies. Lifetime exposure can induce effects that may not be apparent or as large in
magnitude in a shorter study; consequently, a dose that elicits a specific level of response
from a lifetime exposure may be less than the dose eliciting the same level of response from
a shorter exposure period. Thus, PODs based on subchronic exposure data are generally
divided by a sub chronic-to-chronic UF, which has a standard value of 10. If there is
evidence suggesting that exposure for longer time periods does not increase the magnitude of
an effect, a lower value of 3 or 1 might be used. For some reproductive and developmental
effects, chronic exposure is that which covers a specific window of exposure that is relevant
for eliciting the effect, and subchronic exposure would correspond to an exposure that is
notably less than the full window of exposure.
(d) Uncertainty in extrapolating from LOAELs to NOAELs: PODs are intended to be estimates
of exposure levels without appreciable risk under the study conditions so that, after the
application of appropriate UFs for interspecies extrapolation, human variability, and/or
32 Note that the full attribution of the scaling effect, under the assumption that response scales across species in
accordance with ppm equivalence, to pharmacokinetics is an oversimplification and is only one way to think about
how to interpret cross-species scaling. See Section 5.1.3.1 for further discussion of scaling issues.
6/22/2009
1034
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
duration extrapolation, the absence of appreciable risk is conveyed to the RfC or RfD
exposure level to address sensitive humans with lifetime exposure. Under the
NOAEL/LOAEL approach to determining a POD, however, adverse effects are sometimes
observed at all study doses. If the POD is a LOAEL, it is divided by an UF to better estimate
a NOAEL. The standard value for the LOAEL-to-NOAEL UF is 10, although sometimes a
value of 3 is used if the effect is considered minimally adverse at the response level observed
at the LOAEL or even 1 if the effect is an early marker for an adverse effect. For one POD
in this assessment, a value of 30 was used for the LOAEL-to-NOAEL UF because the
incidence rate for the adverse effect was > 90% at the LOAEL.
(e) Additional database uncertainties: Sometimes a database UF of 3 or 10 is used to reflect
other factors contributing uncertainties that are not explicitly treated by the UFs described
above. Such factors include lack of completeness of the overall database, minimal sample
size, or poor exposure characterization. No database UF was used in this assessment. See
Section 5.1.4.1 for additional discussion of the uncertainties associated with the overall
database for TCE.
5.1.2 Candidate Critical Effects by Effect Domain
A large number of endpoints and studies were considered within each of the 5 health
effect domains. A comprehensive list of all endpoints/studies which were considered for
developing cRfCs and cRfDs is shown in Tables 5.1.1-5.1.5. These Tables also summarize the
PODs for the various study endpoints, the UFs applied, and the resulting cRfCs or cRfDs.
Inhalation and oral studies are presented together so that the extent of the available data, as well
as concordance or lack thereof in the responses across routes of exposure, is evident. In addition,
the PBPK model developed in Section 3.5 will be applied to each candidate critical effect to
develop a POD based on internal dose (iPOD); and subsequent extrapolation of the iPOD to
pharmacokinetically sensitive humans is performed for both inhalation and oral human
exposures, regardless of the route of exposure in the original study.
The sections below discuss the cRfCs and cRfDs developed from the effects and studies
identified in the hazard characterization (Chapter 4) that were suitable for the derivation of
reference values (i.e., that provided quantitative dose-response data). Because the general
approach for applying UFs was discussed above, the sections below only discuss the selection of
particular UFs when there are study characteristics that require additional judgment as to the
appropriate UF values and possible deviations from the standard values usually assigned.
6/22/2009
1035
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.2.1 Candidate critical neurological effects on the basis of applied dose
As summarized in Section 4.10.1.1, both human and experimental animal studies have
associated TCE exposure with effects on several neurological domains. The strongest
neurological evidence of hazard is for changes in trigeminal nerve function or morphology and
impairment of vestibular function. There is also evidence for effects on motor function, changes
in auditory, visual, and cognitive function or performance, structural or functional changes in the
brain, and neurochemical and molecular changes. Studies with numerical dose-response
information, with their corresponding cRfCs or cRfDs, are shown in Table 5.1.1. Because
impairment of vestibular function occurs at higher exposures, such changes were not considered
candidate critical effects; but the other neurological effect domains are represented.
For trigeminal nerve effects, cRfC estimates based on two human studies are in a similar
range of 0.4-0.5 ppm (Mhiri et al. 2004; Ruitjen et al. 1991). There remains some uncertainty as
to the exposure characterization, as shown by the use of an alternative POD for Mhiri et al.
(2004) based on urinary TCA resulting in a 5-fold smaller cRfC. However, the overall
confidence in these estimates is relatively high because they are based on humans exposed under
chronic or nearly chronic conditions. Other human studies (e.g., Barret et al. 1984), while
indicative of hazard, did not have adequate exposure information for quantitative estimates of an
inhalation POD. A cRfD of 0.2 mg/kg/d was developed from the only oral study demonstrating
trigeminal nerve changes, an acute study in rats (Barret et al. 1992). This estimate required
multiple extrapolations with a composite uncertainty factor of 10,00033.
For auditory effects, a high confidence cRfC of about 0.7 ppm was developed based on
BMD modeling of data from Crofton and Zhao (1997); and cRfCs developed from two other
auditory studies (Albee et al., 2006; Rebert et al., 1991) were within about 4-fold. No oral data
were available for auditory effects. For psychomotor effects, the available human studies (e.g.,
Rasmussen et al. 1983) did not have adequate exposure information for quantitative estimates of
an inhalation POD. However, a relatively high confidence cRfC of 0.5 ppm was developed from
a study in rats (Waseem et al. 2001). Two cRfDs within a narrow range of 0.7-1.7 mg/kg/d were
developed based on two oral studies reporting psychomotor effects (Nunes et al. 2001; Moser et
al. 1995), although varying in degree of confidence.
For the other neurological effects, the estimated cRfCs and cRfDs were more uncertain,
as there were fewer studies available for any particular endpoint, and the PODs from several
33 U.S. EPA's report on the RfC and RfD processes (U.S. EPA, 2002) recommends not deriving reference values
with a composite UF of greater than 3000; however, composite UFs exceeding 3000 are considered here because the
derivation of the cRfCs and cRfDs is part of a screening process and the subsequent application of the PBPK model
for candidate critical effects will reduce the values of some of the individual UFs
6/22/2009
1036
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
studies required more adjustment to arrive at a cRfC or cRfD. However, the endpoints in these
studies also tended to be indicative of more sensitive effects and, therefore, they need to be
considered. The lower cRfCs fall in the range 0.01-0.1 ppm and were based on effects on visual
function in rabbits (Blain et al. 1994), wakefulness in rats (Arito et al. 1994), and regeneration of
the sciatic nerve in mice and rats (Kjellstrand et al. 1987). Of these, altered wakefulness (Arito
et al. 1994) has both the lowest POD and the lowest cRfC. There is relatively high confidence in
this study, as it shows a clear dose-response trend, with effects persisting post-exposure. For the
subchronic-to-chronic UF, a value of 3 was used because, even though it was just a 6-week
study, there was no evidence of a greater impact on wakefulness following 6 weeks of exposure
than there was following 2 weeks of exposure at the LOAEL, although there was an effect of
repeated exposure on the post-exposure period impacts of higher exposure levels. The cRfDs, in
the range 0.005-0.07, were based on demyelination in the hippocampus (Isaacson et al. 1990)
and degeneration of dopaminergic neurons (Gash et al. 2007), both in rats. In both these cases,
adjusting for study design characteristics led to a composite uncertainty factor of 10,00034, so the
confidence in these cRfDs is lower. However, no other studies of these effects are available.
In summary, although there is high confidence both in the hazard and the cRfCs and
cRfDs for trigeminal nerve, auditory, or psychomotor effects, the available data suggest that the
more sensitive indicators of TCE neurotoxicity are changes in wakefulness, regeneration of the
sciatic nerve, demyelination in the hippocampus and degeneration of dopaminergic neurons.
Therefore, these more sensitive effects are considered the candidate critical effects for
neurotoxicity, albeit with more uncertainty in the corresponding cRfCs and cRfDs. Of these
more sensitive effects, for the reasons discussed above, there is greater confidence in the changes
in wakefulness reported by Arito et al. (1994). In addition, trigeminal nerve effects are
considered a candidate critical effect because this is the only type of neurological effect for
which human data are available, and the POD for this effect is similar to that from the most
sensitive rodent study (Arito et al., 1994, for changes in wakefulness). Between the two human
studies of trigeminal nerve effects, Ruitjen et al. (1991) is preferred for deriving non-cancer
reference values because its exposure characterization is considered more reliable.
34 U.S. EPA's report on the RfC and RfD processes (U.S. EPA, 2002) recommends not deriving reference values
with a composite UF of greater than 3000; however, composite UFs exceeding 3000 are considered here because the
derivation of the cRfCs and cRfDs is part of a screening process and the subsequent application of the PBPK model
for candidate critical effects will reduce the values of some of the individual UFs
6/22/2009
1037
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.1. Neurological effects in studies suitable for dose-response, and corresponding cRfCs and cRfDs.
Effect type Species POD PODa UFSC UFis UFh UF|0aei UFdb UFcompb cRfC cRfD Effect; comments
Supporting studies type (ppm) (mg/kg/d)
Triaeminal Nerve Effects
Mhiri et al. 2004
human
LOAEL
40
1
1
10
10
1
100
0.40
Abnormal trigeminal somatosensory evoked potentials;
preferred POD based on middle of reported range of 50-150
human
LOAEL
6
1
1
10
10
1
100
0.06
ppm.
Alternate POD based on U-TCA and Ikeda et al. (1972).
Ruitjen et al. 1991
human
LOAEL
14
1
1
10
3
1
30
0.47
Trigeminal nerve effects; POD based on mean cumulative
exposure and mean duration, UFIoael = 3 due to early marker
effect and minimal degree of change
Barret et al. 1992
rat
LOAEL
1800
10
10
10
10
1
10000c
0.18
Morphological changes; uncertain adversity; some effects
consistent with demyelination
Auditorv Effects
Rebert et al. 1991
rat
NOAEL
800
10
3
10
1
1
300
2.7
Albee et al. 2006
rat
NOAEL
140
10
3
10
1
1
300
0.47
Crofton & Zhao 1997
rat
BMDL
274
10
3
10
1
1
300
0.91
Preferred, due to better dose-response data, amenable to
BMD modeling. BMR=10dB absolute change.
Psychomotor Effects
Waseem et al. 2001
rat
LOAEL
45
1
3
10
3
1
100
0.45
Changes in locomotor activity; transient, minimal degree of
Nunes et al. 2001
rat
LOAEL
2000
10
10
10
3
1
3000
0.67
adversity; no effect reported in same study for oral exposures
(210 mg/kg/d).
t foot splaying; minimal adversity
Moser et al. 1995
rat
BMDL
248
3
10
10
1
1
300
0.83
t # rears (standing on hindlimbs); BMR=1sd change
rat
NOAEL
500
3
10
10
1
1
300
1.7
t severity score for neuromuscular changes
Visual Function Effects
Blain et al. 1994
rabbit
LOAEL
350
10
3
10
10
1
3000
0.12
POD not adjusted to continuous exposure because visual
effects more closely associated with administered exposure
Coanitive Effects
Kulig et al. 1987
rat
NOAEL
500
1
3
10
1
1
30
17
t time in 2-choice visual discrim test; test involves multiple
systems but largely visual so not adjusted to continuous
exposure
Isaacson et al. 1990
rat
LOAEL
47
10
10
10
10
1
10000c
0.0047
demyelination in hippocampus
Mood and Sleep
Disorders
Albee et al. 2006
rat
NOAEL
140
10
3
10
1
1
300
0.47
hyperactivity
Arito et al. 1994
rat
LOAEL
12
3
3
10
10
1
1000
0.012
Changes in wakefulness
Other neuroloaical
effects
Kjellstrand et al. 1987
rat
LOAEL
300
10
3
10
10
1
3000
0.10
I regeneration of sciatic nerve
mouse
LOAEL
150
10
3
10
10
1
3000
0.050
I regeneration of sciatic nerve
Gash et al. 2007
rat
LOAEL
710
10
10
10
10
1
10000c
0.071
degeneration of dopaminergic neurons
a Adjusted to continuous exposure unless otherwise noted. For inhalation studies, adjustments yield a POD that is a human equivalent concentration as recommended for a Category 3 gas in U.S. EPA
(1994) in the absence of PBPK modeling. Same units as cRfC (ppm) or cRfD (mg/kg/d).
b Product of individual uncertainty factors
6/22/2009
1038
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
0 U.S. EPA's report on the RfC and RfD processes (U.S. EPA, 2002) recommends not deriving reference values with a composite UF of greater than 3000; however, composite UFs exceeding 3000 are
considered here because the derivation of the cRfCs and cRfDs is part of a screening process and the subsequent application of the PBPK model for candidate critical effects will reduce the values of
some of the individual UFs
UFsc = subchronic-to-chronic UF; UFis = interspecies UF; UFh = human variability UF; UF|0aei = LOAEL-to-NOAEL UF; UFdb = database UF
Shaded studies/endpoints were selected as candidate critical effects/studies.
6/22/2009
1039
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.2.2 Candidate critical kidney effects on the basis of applied dose
As summarized in Sections 4.10.1.2, multiple lines of evidence support TCE
nephrotoxicity in the form of tubular toxicity, mediated predominantly through the GSH
conjugation product DCVC. Available human studies, while providing evidence of hazard, did
not have adequate exposure information for quantitative estimates of PODs. Several studies in
rodents, some of chronic duration, have shown histological changes, nephropathy, or increased
kidney/body weight ratios, and were suitable for deriving cRfCs and cRfDs, shown in Table
5.1.2.
The cRfCs developed from three suitable inhalation studies, one reporting
meganucleocytosis in rats (Maltoni et al. 1986) and two others reporting increased kidney
weights in mice (Kjellstand et al. 1983b) and rats (Woolhiser et al. 2006)35, are in a narrow range
of 0.5-1.3 ppm. All three utilized BMD modeling and, thus, take into account statistical
limitations of the Woolhiser et al. (2006) and Kjellstrand et al. (1983b) studies, such as
variability in responses or the use of low numbers of animals in the experiment. The response
used for kidney weight increases was the organ weight as a percentage of body weight, to
account for any commensurate decreases in body weight, although the results did not generally
differ much when absolute weights were used instead. Although the two studies reporting
kidney weight changes were subchronic, longer-term experiments by Kjellstrand et al. (1983b)
did not report increased severity, so no subchronic-to-chronic uncertainty factor was used in the
derivation of the cRfC. The high response level of 73% at the lowest dose for
meganucleocytosis in the chronic study of Maltoni et al. (1986) implies more uncertainty in the
low-dose extrapolation. However, strengths of this study include the presence of
histopathological analysis and relatively high numbers of animals per dose group.
The suitable oral studies give cRfDs within a narrow range of 0.09-0.4 mg/kg/d, as
shown in Table 5.1.2, although the degree of confidence in the cRfDs varies considerably. For
cRfDs based on NTP (1990) and NCI (1976) chronic studies in rodents, extremely high response
rates of >90% precluded BMD modeling. An UF of 10 was applied for extrapolation from a
LOAEL to a NOAEL in the NTP (1990) study because the effect (cytomegaly and
karyomegaly), although minimally adverse, was observed at such a high incidence. An UF of 30
was applied for extrapolation from a LOAEL to a NOAEL in the NCI (1976) study because of
the high incidence of a clearly adverse effect (toxic nephrosis). There is more confidence in the
cRfDs based on meganucleocytosis reported in Maltoni et al. (1986) and toxic nephropathy NTP
35 Woolhiser et al. (2006) is an OECD guideline immunotoxicity study performed by the Dow Chemical Company,
certified by Dow as conforming to Good Laboratory Practices as published by the U.S. EPA for the Toxic
Substances Control Act.
6/22/2009
1040
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
(1988), as BMD modeling was used to estimate BMDLs. Because these two oral studies
measured somewhat different endpoints, but both were sensitive markers of nephrotoxic
responses, they were considered to have similarly strong weight. For meganucleocytosis, a BMR
of 10% extra risk was selected because the effect was considered to be minimally adverse. For
toxic nephropathy, a BMR of 5% extra risk was used because toxic nephropathy is a severe toxic
effect. This BMR required substantial extrapolation below the observed responses (about 60%);
however, the response level seemed warranted for this type of effect and the ratio of the BMD to
the BMDL was not large (1.56).
In summary, there is high confidence both in the hazard and the cRfCs and cRfDs for
histopathological and weight changes in the kidney, and these are considered to be candidate
critical effects for several reasons. First, they appear to be the most sensitive indicators of
toxicity that are available for the kidney. In addition, as discussed in Section 3.5,
pharmacokinetic data indicate substantially more production of GSH-conjugates thought to
mediate TCE kidney effects in humans relative to rats and mice. As discussed above, several
studies are considered reliable for developing cRfCs and cRfDs for these endpoints. For
histopathological changes, the most sensitive were selected as candidate critical studies. These
were the only available inhalation study (Maltoni et al. 1986), the NTP (1988) study in rats, and
the NCI (1976) study in mice. While the NCI (1976) study has greater uncertainty, as discussed
above, with a high response incidence at the POD that necessitates greater low-dose
extrapolation, it is included to add a second species to the set of candidate critical effects. For
kidney weight changes, both available studies were chosen as candidate critical studies.
5.1.2.3 Candidate critical liver effects on the basis of applied dose
As summarized in Sections 4.10.1.3, while there is only limited epidemiologic evidence
of TCE hepatotoxicity, TCE clearly leads to liver toxicity in laboratory animals, likely through
its oxidative metabolites. Available human studies contribute to the overall weight of evidence
of hazard, but did not have adequate exposure information for quantitative estimates of PODs.
In rodent studies, TCE causes a wide array of hepatotoxic endpoints, including increased liver
weight, small transient increases in DNA synthesis, changes in ploidy, cytomegaly, increased
nuclear size, and proliferation of peroxisomes. Increased liver weight (hepatomegaly, or
specifically increased liver/body weight ratio) has been the most studied endpoint across a range
of studies in both sexes of rats and mice, with a variety of exposure routes and durations.
Hepatomegaly was selected as the critical liver effect for multiple reasons. First, it has been
consistently reported in multiple studies in rats and mice following both inhalation and oral
routes of exposure. In addition, it appears to accompany the other hepatic effects at the doses
6/22/2009
1041
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
tested, and hence constitutes a hepatotoxicity marker of similar sensitivity to the other effects.
Finally, in several studies, there are good dose-response data for BMD modeling.
As shown in Table 5.1.2, cRfCs for hepatomegaly developed from the two most suitable
subchronic inhalation studies (Woolhiser et al. 2006; Kjellstrand et al. 1983b), while in different
species (rats and mice, respectively), are both based on similar PODs derived from BMD
modeling, have the same composite uncertainty factor of 30, and result in similar cRfC estimates
of about 0.8 ppm. The cRfD for hepatomegaly developed from the oral study of Buben and
O'Flaherty (1985) in mice also was based on a POD derived from BMD modeling and resulted
in a cRfD estimate of 0.8 mg/kg/d. Among the studies reporting liver weight changes (reviewed
in Section 4.4 and Appendix E), this study had by far the most extensive dose-response data.
The response used in each case was the liver weight as a percentage of body weight, to account
for any commensurate decreases in body weight, although the results did not generally differ
much when absolute weights were used instead.
There is high confidence in all these candidate reference values. BMD modeling takes
into account statistical limitations such as variability in response or low numbers of animals and
standardizes the response rate at the POD. Although the studies were subchronic, hepatomegaly
occurs rapidly with TCE exposure, and the degree of hepatomegaly does not increase with
chronic exposure (Kjellstrand et al. 1983b), so no subchronic-to-chronic uncertainty factor was
used.
In summary, there is high confidence both in the hazard and the cRfCs and cRfDs for
hepatomegaly. Hepatomegaly also appears to be the most sensitive indicator of toxicity that is
available for the liver and is therefore considered a candidate critical effect. As discussed above,
several studies are considered reliable for developing cRfCs and cRfDs for this endpoint, and,
since they all indicated similar sensitivity but represented different species and/or routes of
exposure, were all considered candidate critical studies.
5.1.2.4 Candidate critical body weight effects on the basis of applied dose
The chronic oral bioassays NCI (1976) and NTP (1990) reported decreased body weight
with TCE exposure, as shown in Table 5.1.2. However, the lowest doses in these studies were
quite high, even on an adjusted basis (see PODs in Table 5.1.2). These were not considered
critical effects because they are not likely to be the most sensitive non-cancer endpoints, and
were not considered candidate critical effects.
6/22/2009
1042
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.2. Kidney, liver, and body weight effects in studies suitable for dose-response, and corresponding cRfCs and cRfDs.
Effect type Species POD PODa UFSC UFis UFh UF|0aei UFdb UFcompb cRfC cRfD Effect; comments
Supporting studies type (ppm) (mg/kg/d)
Histological changes in
kidney
Maltoni 1986
rat
BMDL
40.2
1
3
10
1
1
30
1.3
meganucleocytosis; BMR=10% extra risk
rat
BMDL
34
1
10
10
1
1
100
0.34
meganucleocytosis; BMR=10% extra risk
NTP 1990
rat
LOAEL
360
1
10
10
10
1
1000
0.36
cytomegaly & karyomegaly; considered minimally adverse, but
UF|0aei = 10 due to high response rate (> 98%) at LOAEL; also
in mice, but use NCI 1976 for that species
NCI 1976
mouse
LOAEL
620
1
10
10
30
1
3000
0.21
toxic nephrosis; UFIoael = 30 due to >90% response at
LOAELfor severe effect
NTP 1988
rat
BMDL
9.45
1
10
10
1
1
100
0.0945
toxic nephropathy; female Marshall (most sensitive sex/strain);
BMR = 5% extra risk
t kidnev/bodv weight
ratio
Kjellstrand et al. 1983b
mouse
BMDL
34.7
1
3
10
1
1
30
1.2
BMR=10% increase; 30 d, but 120 d @ 120 ppm not more
severe so UFsc = 1; results are for males, which were slightly
more sensitive, and yielded better fit to variance model
Woolhiser et al. 2006
rat
BMDL
15.7
1
3
10
1
1
30
0.52
BMR=10% increase; UFsc = 1 based on Kjellstrand et al.
(1983b) result
t liver/body weight
ratio
Kjellstrand et al. 1983b
mouse
BMDL
21.6
1
3
10
1
1
30
0.72
BMR=10% increase; UFsc = 1
based on not more severe at 4
months
Woolhiser et al. 2006
rat
BMDL
25.2
1
3
10
1
1
30
0.84
BMR=10% increase; UFsc = 1
based on Kjellstrand et al.
(1983b) result
Buben & O'Flaherty 1985
mouse
BMDL
81.5
1
10
10
1
1
100
0.82 BMR=10% increase; UFsc = 1
based on Kjellstrand et al.
(1983b) result
Decreased body weight
NTP 1990 mouse LOAEL 710 1 10 10 10 1 1000 0.71
NCI 1976 rat LOAEL 360 1 10 10 10 1 1000 0.36 Reflects several, but not all, strains/sexes.
a Adjusted to continuous exposure unless otherwise noted. For inhalation studies, adjustments yield a POD that is a human equivalent concentration as recommended for a Category 3 gas in U.S. EPA
(1994) in the absence of PBPK modeling. Same units as cRfC (ppm) or cRfD (mg/kg/d).
b Product of individual uncertainty factors.
UFsc = subchronic-to-chronic UF; UFis = interspecies UF; UFh = human variability UF; UF|0aei = LOAEL-to-NOAEL UF; UFdb = database UF
Shaded studies/endpoints were selected as candidate critical effects/studies.
6/22/2009
1043
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.2.5 Candidate critical immunological effects on the basis of applied dose
As summarized in 4.10.1.4, the human and experimental animal studies of TCE and
immune-related effects provide strong evidence for a role of TCE in autoimmune disease and in
a specific type of generalized hypersensitivity syndrome, while there are fewer data pertaining to
immunosuppressive effects. Available human studies, while providing evidence of hazard, did
not have adequate exposure information for quantitative estimates of PODs. Several studies in
rodents were available on autoimmune and immunosuppressive effects that were adequate for
deriving cRfCs and cRfDs, which are summarized in Table 5.1.3.
For decreased thymus weights, a cRfD from the only suitable study (Keil et al. 2009) is
0.00035 mg/kg/d based on results from non-autoimmune-prone B6C3F1 mice, with a composite
uncertainty factor of 1000 for a POD that is a LOAEL (the dose-response relationship is
sufficiently supralinear that attempts at BMD modeling did not result in adequate fits to these
data). Thymus weights were not significantly affected in autoimmune prone mice in the same
study, consistent with the results reported by Kaneko et al. (2000) in autoimmune-prone mice. In
addition, Keil et al. (2009) and Peden-Adams et al. (2008) reported that for several
immunotoxicity endpoints associated with TCE, the autoimmune-prone strain appeared to be less
sensitive than the non-autoimmune prone B6C3F1 strain. In rats, Woolhiser et al. (2006)
reported no significant change in thymus weights in the CD strain. These data are consistent
with normal mice being sensitive to this effect as compared to autoimmune-prone mice or CD
rats, so the results of Keil et al. (2009) are not necessarily discordant with the other studies.
For autoimmune effects, the cRfC from the only suitable inhalation study (Kaneko et al.
2000) is 0.07 ppm. This study reported changes in immunoreactive organs (i.e., liver and spleen)
in autoimmune-prone mice. BMD modeling was not feasible, so a LOAEL was used as the
POD. The standard value of 10 was used for the LOAEL-to-NOAEL UF because the
inflammation was reported to include sporadic necrosis in the hepatic lobules at the LOAEL, so
this was considered an adverse effect. A value of 3 was used for the human (intraspecies)
variability UF because the effect was induced in autoimmune-prone mice, a sensitive mouse
strain for such an effect. The cRfDs from the oral studies (Keil et al. 2009; Griffin et al. 2000;
Cai et al. 2008) spanned about a 100-fold range from 0.004-0.5 mg/kg/d. Each of the studies
used different markers for autoimmune effects, which may explain the over 100-fold range of
PODs (0.4-60 mg/kg/d). The most sensitive endpoint, reported by Keil et al. (2009), was
increases in anti-dsDNA and anti-ssDNA antibodies, early markers for systemic lupus
erythematosus (SLE), in B6C3Flmice exposed to the lowest tested dose of 0.35 mg/kg/d,
yielding a cRfD of 0.004 mg/kg/d. Therefore, the results of Keil et al. (2009) are not discordant
6/22/2009
1044
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
with the higher PODs and cRfDs derived from the other oral studies that examined more frank
autoimmune effects.
For immunosuppressive effects, the only suitable inhalation study (Woolhiser et al. 2006)
gave a cRfC of 0.08 ppm. The cRfDs from the only suitable oral study (Sanders et al. 1982)
ranged from 0.06 mg/kg/d to 2 mg/kg/d, based on different markers for immunosuppression.
Woolhiser et al. (2006) reported decreased PFC response in rats. Data from Woolhiser et al.
(2006) were amenable to BMD modeling, but there is notable uncertainty in the modeling. First,
it is unclear what should constitute the cut-point for characterizing the change as minimally
biologically significant, so a BMR of 1 control SD change was used. In addition, the dose-
response relationship is supralinear, and the highest exposure group was dropped to improve the
fit to the low-dose data points. Nonetheless, the uncertainty in the BMD modeling is no greater
than the uncertainty inherent in the use of a LOAEL or NOAEL. The more sensitive endpoints
reported by Sanders et al. (1982), both of which were in female mice exposed to a LOAEL of 18
mg/kg/day TCE in drinking water for 4 months, were decreased cell-mediated response to sheep
red blood cells (sRBC) and decreased stem cell bone recolonization, a sign of impaired bone
marrow function. The cRfD based on these endpoints is 0.06 mg/kg/d, with a LOAEL-to-
NOAEL UF of 3 because, although the immunosuppressive effects may not be adverse in and of
themselves, multiple effects were observed suggesting potentially less resilience to an insult
requiring an immunological response.
In summary, there is high qualitative confidence for TCE immunotoxicity and moderate
confidence in the cRfCs and cRfDs that can be derived from the available studies. Decreased
thymus weight reported at relatively low exposures in non-autoimmune-prone mice is a clear
indicator of immunotoxity (Keil et al. 2009), and is therefore considered a candidate critical
effect. A number of studies have also reported changes in markers of immunotoxicity at
relatively low exposures. Therefore, among markers for autoimmune effects, the more sensitive
measures of autoimmune changes in liver and spleen (Kaneko et al. 2000) and increased anti-
dsDNA and anti-ssDNA antibodies (Keil et al. 2009) are considered the candidate critical
effects. Similarly, for markers of immunosuppression, the more sensitive measures of decreased
PFC response (Woolhiser et al. 2006), decreased stem cell bone marrow recolonization, and
decreased cell-mediated response to sRBC (both from Sanders et al. 1982 ) are considered the
candidate critical effects.
6/22/2009
1045
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.3. Immunological effects in studies suitable for dose-response, and corresponding cRfCs and cRfDs.
Effect type Species POD PODa UFSC UFis UFh UF|0aei UFdb UFcompb cRfC cRfD Effect; comments
Supporting studies type (ppm) (mg/kg/d)
I thymus weight
Keil et al. (2009)
Mouse
LOAEL
0.35
1
10
10
10
1
1000
0.00035
i thymus weight; corresponding decrease in total thymic
cellularity reported at 10x higher dose.
Autoimmunity
Kaneko et al. 2000
mouse
(MRL-
Ipr/lpr)
LOAEL
70
10
3
3
10
1
1000
0.070
Changes in immunoreactive organs - liver (incl. sporadic
necrosis in hepatic lobules), spleen; UFh=3 due to
autoimmune-prone mouse
Keil et al. (2009)
mouse
LOAEL
0.35
1
10
10
1
1
100
0.0035
t anti-dsDNA and anti-ssDNA Abs (early markers for SLE)
(B6C3F1 mouse); UFIoael=1 due to early marker
Griffin et al. 2000
mouse
(MRL+/+)
BMDL
13.4
1
10
3
1
1
30
0.45
various signs of autoimmune hepatitis; BMR=10% extra risk
for > minimal effects
Cai et al. 2008
mouse
(MRL+/+)
LOAEL
60
1
10
3
10
1
300
0.20
Inflammation in liver, kidney, lungs, and pancreas, which may
lead to SLE-like disease; UFh=3 due to autoimmune-prone
mouse; UFIoael = 10 since some hepatic necrosis
immunosuppression
Woolhiser et al. 2006
rat
BMDL
31.2
10
3
10
1
1
300
0.10
[ PFC response; BMR=1SD change; highest dose dropped
Sanders et al. 1982
mouse
NOAEL
190
1
10
10
1
1
100
1.9
[ humoral response to sRBC; largely transient during
exposure
mouse
LOAEL
18
1
10
10
3
1
300
0.060
[ stem cell bone marrow recolonization (sustained); females
more sensitive
mouse
LOAEL
18
1
10
10
3
1
300
0.060
[ cell-mediated response to sRBC (largely transient during
exposure); females more sensitive
a Adjusted to continuous exposure unless otherwise noted. For inhalation studies, adjustments yield a POD that is a human equivalent concentration as recommended for a Category 3 gas in U.S. EPA
(1994) in the absence of PBPK modeling. Same units as cRfC (ppm) or cRfD (mg/kg/d).
b Product of individual uncertainty factors.
UFsc = subchronic-to-chronic UF; UFis = interspecies UF; UFh = human variability UF; UF|0aei = LOAEL-to-NOAEL UF; UFdb = database UF
Shaded studies/endpoints were selected as candidate critical effects/studies.
6/22/2009
1046
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.2.6 Candidate critical respiratory tract effects on the basis of applied dose
As summarized in Section 4.10.1.5, available data are suggestive of TCE causing
respiratory tract toxicity, based primarily on short-term studies in mice and rats. However, these
studies are generally at high inhalation exposures and over durations of less than 2 weeks. Thus,
these were not considered critical effects because such data are not necessarily indicators of
longer-term effects at lower exposure and are not likely to be the most sensitive non-cancer
endpoints for chronic exposures. Therefore, cRfCs and cRfDs were not developed for them.
5.1.2.7 Candidate critical reproductive effects on the basis of applied dose
As summarized in Section 4.10.1.6, both human and experimental animal studies have
associated TCE exposure with adverse reproductive effects. The strongest evidence of hazard is
for effects on sperm and male reproductive outcomes, with evidence from multiple human
studies and several experimental animal studies. There is also substantial evidence for effects on
the male reproductive tract and male serum hormone levels, as well as evidence for effects on
male reproductive behavior. There are fewer data and more limited support for effects on female
reproduction. The PODs, UFs, and resulting cRfDs and cRfCs for the effects from the suitable
reproductive studies are summarized in Table 5.1.4.
5.1.2.7.1 Male reproductive effects (effects on sperm and reproductive tract)
A number of available studies have reported functional and structural changes in sperm
and male reproductive organs and effects on male reproductive outcomes following TCE
exposure (Table 5.1.4). A cRfC of 0.014 ppm was derived based on hyperzoospermia reported
in the available human study (Chia et al. 1996), but there is substantial uncertainty in this
estimate due to multiple issues.36 Among the rodent inhalation studies, the cRfC of 0.2 ppm
based on increased abnormal sperm in the mouse reported by Land et al. (1981) is considered
relatively reliable because it is based on BMD modeling rather than a LOAEL or NOAEL.
However, increased sperm abnormalities do not appear to be the most sensitive effect, as Kumar
et al. (2000a, 2000b, 2001) reported a similar POD to be a LOAEL for reported multiple effects
on sperm and testes, as well as altered testicular enzyme markers in the rat. Although there are
greater uncertainties associated with the cRfC of 0.02 ppm for this effect and a composite UF of
36 Mean exposure estimates for the exposure groups were limited because they were defined in terms of ranges and
because they were based on mean urinary TCA (mg/g creatinine). There is substantial uncertainty in the conversion
of urinary TCA to TCE exposure level (see discussion of Mhiri et al. 2004, for neurotoxicity, above). In addition,
there was uncertainty about the adversity of the effect being measured. While rodent evidence supports effects of
TCE on sperm, and hyperzoospermia has reportedly been associated with infertility, the adversity of the
hyperzoospermia (i.e., high sperm density) outcome measured in the Chia et al. (1996) study is unclear.
Furthermore, the cut-point used to define hyperzoospermia in this study (i.e., > 120 million sperm per mL ejaculate)
is lower than some other reported cut-points, such as 200 and 250 million/mL. A BMR of 10% extra risk was used
on the assumption that this is a minimally adverse effect, but biological significance of this effect level is unclear.
6/22/2009
1047
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
3000 was applied to the POD, the uncertainties are generally typical of those encountered in RfC
derivations. Standard values of 3, 10, and 10 were used for the interspecies UF, the human
variability UF, and the LOAEL-to-NOAEL UF, respectively. In addition, although the study
would have qualified as a chronic exposure study based on its duration of 24 weeks (i.e., > 10%
of lifetime), statistically significant decreases in testicular weight and in sperm count and
motility were already observed from subchronic exposure (12 weeks) to the same TCE exposure
concentration and these effects became more severe after 24 weeks of exposure. Moreover,
several testicular enzyme markers associated with spermatogenesis and germ cell maturation had
significantly altered activities after 12 weeks of exposure, with more severe alterations at 24
weeks, and histological changes were also observed in the testes at 12 weeks, with the testes
being severely deteriorated by 24 weeks. Thus, since the single exposure level used was already
a LOAEL from subchronic exposure, and the testes were even more seriously affected by longer
exposures, a subchronic-to-chronic UF of 10 was applied.37 Note that for the cRfC derived for
pre-and post-implantation losses reported by Kumar et al. (2000a), the subchronic-to-chronic UF
was not applied because the exposure covered the time period for sperm development. This
cRfC was 0.2 ppm, similar to that derived from Land et al. (1981) based on BMD modeling of
increases in abnormal sperm.
At a higher inhalation POD, Xu et al. (2004) reported decreased fertilization following
exposure in male mice, and Forkert et al. (2002) and Kan et al. (2007) reported effects on the
epididymal epithelium in male mice. Kan et al. (2007) reported degenerative effects on the
epididymis as early as 1 week into exposure that became more severe at 4 weeks of exposure
when the study ended; increases in abnormal sperm were also observed. As with the cRfC
developed from the Kumar et al. (2000a, 2000b, 2001), a composite UF of 3000 was applied to
these data, but the uncertainties are again typical of those encountered in RfC derivations.
Standard values of 3 for the interspecies UF, 10 for the human variability UF, 10 for the
LOAEL-to-NOAEL UF, and 10 for the subchronic-to-chronic UF were applied to each of the
study PODs.
Among the oral studies, cRfDs derived for decreased sperm motility and changes in
reproductive organ weights in rodents reported by George et al. (1985, 1986) were relatively
high (2-4 mg/kg/d), and these effects were not considered candidate critical effects. The
remaining available oral study of male reproductive effects is DuTeaux et al. (2004b), which
reported decreased ability of sperm from TCE-exposed rats to fertilize eggs in vitro. This effect
37 Alternatively, the value of the LOAEL-to-NOAEL UF could have been increased above 10 to reflect the extreme
severity of the effects at the LOAEL after 24 weeks; however, the comparison of the 12-week and 24-week results
gives such a clear depiction of the progression of the effects, it was more compelling to frame the issue as a
subchronic-to-chronic extrapolation issue.
6/22/2009
1048
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
occurred in the absence of changes in combined testes/epididymes weight, sperm concentration
or motility, or histological changes in the testes or epididymes. DuTeaux et al. (2004b)
hypothesize that the effect is due to oxidative damage to the sperm. A LOAEL was used as the
POD, and the standard UF values of 10 were used for each of the UFs, i.e., the subchronic-to-
chronic UF (14-day study; substantially less than the 70-day time period for sperm
development), the interspecies UF for oral exposures, the human variability UF, and the
LOAEL-to-NOAEL UF. The resulting composite UF was 10,00 038, and this yielded a cRfD of
0.01 mg/kg/day. The excessive magnitude of the composite UF, however, highlights the
uncertainty in this estimate.
In summary, there is high qualitative confidence for TCE male reproductive tract toxicity
and lower confidence in the cRfCs and cRfDs that can be derived from the available studies.
Relatively high PODs are derived from several studies reporting less sensitive endpoints (George
et al. 1985, 1986; Land et al. 1981), and correspondingly higher cRfCs and cRfDs suggest that
they are not likely to be critical effects. The studies reporting more sensitive endpoints also tend
to have greater uncertainty. For the human study by Chia et al. (1996), as discussed above, there
are uncertainties in the characterization of exposure and the adversity of the effect measured in
the study. For the Kumar et al. (2000a, 2000b, 2001), Forkert et al. (2002) and Kan et al. (2007)
studies, the severity of the sperm and testes effects appears to be continuing to increase with
duration even at the end of the study, so it is plausible that a lower exposure for a longer duration
may elicit similar effects. For the DuTeaux et al. (2004b) study, there is also duration- and low-
dose extrapolation uncertainty due to the short duration of the study in comparison to the time
period for sperm development as well as the lack of a NOAEL at the tested doses. Overall, even
though there are limitations in the quantitative assessment, there remains sufficient evidence to
consider these to be candidate critical effects.
5.1.2.7.2 Other reproductive effects
With respect to female reproductive effects, several studies reporting decreased maternal
weight gain were suitable for deriving candidate reference values (Table 5.1.4). The cRfCs from
the two inhalation studies (Carney et al. 2006; Schwetz et al. 1975) yielded virtually the same
estimate 0.3-0.4 ppm, although the Carney et al. (2006) result is preferred due to the use of
BMD modeling, which obviates the need for the 10-fold LOAEL-to-NOAEL UF used for
Schwetz et al. (1975) (the other UFs, with a product of 30, were the same) The cRfDs for this
endpoint from the three oral studies were within 3-fold of each other (1-3 mg/kg/d), with the
38 U.S. EPA's report on the RfC and RfD processes (U.S. EPA, 2002) recommends not deriving reference values
with a composite UF of greater than 3000; however, composite UFs exceeding 3000 are considered here because the
derivation of the cRfCs and cRfDs is part of a screening process and the subsequent application of the PBPK model
for candidate critical effects will reduce the values of some of the individual UFs
6/22/2009
1049
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
same composite UFs of 100. The most sensitive estimate of Narotsky et al. (1995) is preferred
due to the use of BMD modeling and the apparent greater sensitivity of the rat strain used.
With respect to other reproductive effects, the most reliable cRfD estimates of about 2
mg/kg/d, derived from BMD modeling with composite UFs of 100, are based on decreased
litters/pair and decreased live pups/litter in rats reported in the continuous breeding study of
George et al. (1986). Both of these effects were considered severe adverse effects, so a BMR of
a 0.5 control SD shift from the control mean was used. Somewhat lower cRfDs of 0.4-1
mg/kg/d were derived based on delayed parturition in females (Narotsky et al. 1995), decreased
copulatory performance in males (Zenick et al. 1984), and decreased mating for both exposed
males and females in cross-over mating trials (George et al. 1986), all with composite UFs of
100 or 1000 depending on whether a LOAEL or NOAEL was used.
In summary, there is moderate confidence both in the hazard and the cRfCs and cRfDs
for reproductive effects other than the male reproductive effects discussed previously. While
there are multiple studies suggesting decreased maternal body weight with TCE exposure, this
systemic change may not be indicative of more sensitive reproductive effects. None of the
estimates developed from other reproductive effects is particularly uncertain or unreliable.
Therefore, delayed parturition (Narotsky et al. 1995) and decreased mating (George et al. 1986),
which yielded the lowest cRfDs, were considered candidate critical effects. These effects were
also included so that candidate critical reproductive effects from oral studies would not include
only that reported by DuTeaux et al. (2004b), from which deriving the cRfD entailed a higher
degree of uncertainty.
6/22/2009
1050
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 Table 5.1.4. Reproductive effects in studies suitable for dose-response, and corresponding cRfCs and cRfDs.
Effect type Species POD PODa UFSC UFis UFh UF|0aei UFdb UFcompb cRfC cRfD Effect; comments
Supporting studies type (ppm) (mg/kg/d)
Effects on sperm, male
reproductive outcomes
Chia et al. 1996
human
BMDL
1.43
10
1
10
1
1 100
0.014
hyperzoospermia; exposure estimates based on U-TCA from
Ikeda et al. (1972); BMR=10% extra risk
Land et al. 1981
mouse
BMDL
46.9
10
3
10
1
1 300
0.16
t abnormal sperm; BMR=0.5SD
Kan et al. 2007
mouse
LOAEL
180
10
3
10
10
1 3000
0.060
t abnormal sperm; Land et al. (1981) cRfC preferred due to
BMD modeling
Xu et al. 2004
mouse
LOAEL
180
10
3
10
10
1 3000
0.060
I fertilization
Kumar et al. 2000a 2001 b
rat
LOAEL
45
10
3
10
10
1 3000
0.015
multiple sperm effects, increasing severity from 12 to 24 weeks
rat
LOAEL
45
1
3
10
10
1 300
0.15
pre- and post-implantation losses; UFsc = 1 due to exposure
covered time period for sperm development; higher response
for pre-implantation losses
George et al. 1985
mouse
NOAEL
362
1
10
10
1
1 100
3.6
I sperm motility
DuTeaux et al. 2004
rat
LOAEL
141
10
10
10
10
1 10000c
0.014
I ability of sperm to fertilize in vitro
Male reproductive tract
effects
Forkert et al. 2002, Kan et
mouse
LOAEL
180
10
3
10
10
1 3000
0.060
effects on epididymis epithelium
al. 2007
Kumar et al. 2000a 2001 b
rat
LOAEL
45
10
3
10
10
1 3000
0.015
testes effects, altered testicular enzyme markers, increasing
severity from 12 to 24 weeks
George et al. 1985
mouse
NOAEL
362
1
10
10
1
1 100
3.6
I testis/seminal vesicle weights
George et al. 1986
rat
NOAEL
186
1
10
10
1
1 100
1.9
t testis/epididymis weights
Female maternal weiaht
gain
Carney et al. 2006
rat
BMDL
10.5
1
3
10
1
1 30
0.35
I BWgain; BMR=10% decrease
Schwetz et al. 1975
rat
LOAEL
88
1
3
10
10
1 300
0.29
I mat BW; Carney et al. 2006 cRfC preferred due to BMD
modeling
Narotsky et al. 1995
rat
BMDL
108
1
10
10
1
1 100
1.1
I BWgain; BMR=10% decrease
Manson et al. 1984
rat
NOAEL
100
1
10
10
1
1 100
1.0
I BW gain; Narotsky et al. 1995 preferred due to BMD
modeling (different strain)
George et al. 1986
rat
NOAEL
186
1
10
10
1
1 100
1.9
I postpartum BW; Narotsky et al. 1995 cRfD preferred due to
BMD modeling
Female reproductive
outcomes
Narotsky et al. 1995
rat
LOAEL
475
1
10
10
10
1 1000
0.48
delayed parturition
Reproductive behavior
Zenick et al. 1984
rat
NOAEL
100
1
10
10
1
1 100
1.0
I copulatory performance in males
George et al. 1986
rat
LOAEL
389
1
10
10
10
1 1000
0.39
I mating (both sexes exposed)
6/22/2009
1051
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Reproductive effects
from exposure to both
sexes
George etal. 1986 rat BMDL 179 1 10 10 1 1 100 1.8 | # litters/pair; BMR=0.5SD
rat BMDL 152 1 10 10 1 1 100 1.5 | live pups/litter; BMR=0.5SD
a Adjusted to continuous exposure unless otherwise noted. For inhalation studies, adjustments yield a POD that is a human equivalent concentration as recommended for a Category 3 gas in U.S. EPA
(1994) in the absence of PBPK modeling. Same units as cRfC (ppm) or cRfD (mg/kg/d).
b Product of individual uncertainty factors.
0 U.S. EPA's report on the RfC and RfD processes (U.S. EPA, 2002) recommends not deriving reference values with a composite UF of greater than 3000; however, composite UFs exceeding 3000 are
considered here because the derivation of the cRfCs and cRfDs is part of a screening process and the subsequent application of the PBPK model for candidate critical effects will reduce the values of
some of the individual UFs.
UFsc = subchronic-to-chronic UF; UFis = interspecies UF; UFh = human variability UF; UF|0aei = LOAEL-to-NOAEL UF; UFdb = database UF
Shaded studies/endpoints were selected as candidate critical effects/studies.
6/22/2009
1052
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.2.8 Candidate critical developmental effects on the basis of applied dose
As summarized in Section 4.10.1.7, both human and experimental animal studies have
associated TCE exposure with adverse developmental effects. Weakly suggestive epidemiologic
data and fairly consistent experimental animal data support TCE exposure posing a hazard for
increased pre-natal or post-natal mortality and decreased pre- or post-natal growth. In addition,
congenital malformations following maternal TCE exposure have been reported in a number of
epidemiologic and experimental animal studies. There is also some support for TCE effects on
neurological and immunological development. Available human studies, while indicative of
hazard, did not have adequate exposure information for quantitative estimates of PODs, so only
experimental animal studies are considered here. The PODs, UFs, and resulting cRfDs and
cRfCs for the effects from the suitable developmental studies are summarized in Table 5.1.5.
For pre- and post-natal mortality and growth, a cRfC of 0.06 ppm for resorptions,
decreased fetal weight, and variations in skeletal development indicative of delays in ossification
was developed based on the single available (rat) inhalation study considered (Healy et al. 1982)
and utilizing the composite UF of 300 for an inhalation POD that is a LOAEL. The cRfDs for
pre- and post-natal mortality derived from oral studies were within about a 10-fold range of 0.4-
5 mg/kg/d, depending on the study and specific endpoint assessed. Of these, the estimate based
on Narotsky et al. (1995) rat data was both the most sensitive and most reliable cRfD. The dose
response for increased full-litter resorptions from this study is based on BMD modeling.
Because of the severe nature of this effect, a BMR of 1% extra risk was used. The ratio of the
resulting BMD to the BMDL was 5.7, which is on the high side, but given the severity of the
effect and the low background response, a judgment was made to use 1% extra risk.
Alternatively, a 10% extra risk could have been used, in which case the POD would have been
considered more analogous to a LOAEL than a NOAEL, and a LOAEL-to-NOAEL UF of 10
would have been applied, ultimately resulting in the same cRfD estimate. The cRfDs for altered
pre- and post-natal growth developed from the oral studies ranged about 10-fold from 0.8-8
mg/kg/d, all utilizing the composite UFs for the corresponding type of POD. The cRfDs for
decreased fetal weight, both of which were based on NOAELs, were consistent, being about 2-
fold apart (Narotsky et al. 1995; George et al. 1985). The cRfD based on post-natal growth at 21
days, reported in George et al. (1986), was lower and is preferred because it was based on BMD
modeling. A BMR of 5% decrease in weight was used for post-natal growth at 21 days because
decreases in weight gain so early in life were considered similar to effects on fetal weight.
For congenital defects, there is relatively high confidence in the cRfD for eye defects in
rats reported in Narotsky et al. (1995), derived using a composite UF of 100 for BMD modeling
in a study of duration that encompasses the full window of eye development. However, the most
6/22/2009
1053
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
sensitive developmental effect by far was heart malformations in the rat reported by Johnson et
al. (2003), yielding a cRfD estimate of 0.0002 mg/kg/d, also with a composite UF of 100. As
discussed in detail in Section 4.7 and summarized in Section 4.10.1.7, although this study has
important limitations, the overall weight of evidence supports an effect of TCE on cardiac
development, and this is the only study of heart malformations available for conducting dose-
response analysis. Individual data were kindly provided by Dr. Johnson (personal
communication from Paula Johnson, University of Arizona, to Susan Makris, U.S. EPA, 25
August 2008), and, for analyses for which the pup was the unit of measure, BMD modeling was
done using nested models because accounting for the intralitter correlation improved model fit.
For these latter analyses, a 1% extra risk of a pup having a heart malformation was used as the
BMR because of the severity of the effect, since some of the types of malformations observed
could have been fatal. The ratio of the resulting BMD to the BMDL was about 3.
For developmental neurotoxicity, the cRfD estimates based on the 4 oral studies span a
wide range from 0.02 to 0.8 mg/kg/d. The most reliable estimate, with a composite UF of 100, is
based on BMD modeling of decreased locomotor activity in rats reported in George et al. (1986),
although a non-standard BMR of a 2-fold change was selected because the control SD appeared
unusually small. The cRfDs developed for decreased rearing post-exposure in mice
(Fredricksson et al. 1993), increased exploration post-exposure in rats (Taylor et al. 1985) and
decreased myelination in the hippocampus of rats (Isaacson and Taylor 1989), while being more
than 10-fold lower, are all within a 3-fold range of 0.02-0.05 mg/kg/d. Importantly, there is
some evidence from adult neurotoxicity studies of TCE causing demyelination, so there is
additional biological support for the latter effect. There is greater uncertainty in the Fredricksson
et al. (1993), the cRfD for which utilized a subchronic-to-chronic UF of 3 rather than 1, because
exposure during PND 10-16 does not cover the full developmental window (Rice and Barone
2000). The cRfDs derived from Taylor et al. (1985) and (Isaacson and Taylor 1989) used the
composite UF of 1000 for a POD that is a LOAEL. While there is greater uncertainty in these
endpoints, none of the uncertainties is particularly high, and they also appear to be more
sensitive indicators of developmental neurotoxicity than that from George et al. (1986).
A cRfD of 0.0004 mg/kg/d was developed from the study (Peden-Adams et al. 2006) that
reported developmental immunotoxicity. The main effects observed were significantly
decreased PFC response and increased delayed-type hypersensitivity. The data on these effects
were kindly provided by Dr. Peden-Adams (personal communication from Margie Peden-
Adams, Medical University of South Carolina, to Jennifer Jinot, U.S. EPA, 26 August 2008);
however, the dose-response relationships were sufficiently supralinear that attempts at BMD
modeling did not result in adequate fits to these data. Thus, the LOAEL was used as the POD.
Although decreased PFC response may not be considered adverse in and of itself, a LOAEL-to-
6/22/2009
1054
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
NOAEL UF of 10 was used because of the increased delayed-type hypersensitivity at the same
dose. While there is uncertainty in this estimate, it is notable that decreased PFC response was
also observed in an immunotoxicity study in adult animals (Woolhiser et al., 2006), lending
biological plausibility to the effect.
In summary, there is moderate-to-high confidence both in the hazard and the cRfCs and
cRfDs for developmental effects of TCE. It is also noteworthy that the PODs for the more
sensitive developmental effects were similar to or, in most cases, lower than the PODs for the
more sensitive reproductive effects, suggesting that developmental effects are not a result of
paternal or maternal toxicity. Among inhalation studies, cRfCs were only developed for effects
in rats reported in Healy et al. (1982), so the effects of resorptions, decreased fetal weight, and
delayed skeletal ossification were considered candidate critical developmental effects. Because
resorptions were also reported in oral studies, the most sensitive (rat) oral study (and most
reliable for dose-response analysis) of Narotsky et al. (1995) was also selected as a candidate
critical study for this effect. The confidence in the oral studies and candidate reference values
developed for more sensitive endpoints is more moderate, but still sufficient for consideration as
candidate critical effects. The most sensitive endpoints by far are the increased fetal heart
malformations in rats reported by Johnson et al. (2003) and the developmental immunotoxicity in
mice reported by Peden-Adams et al. (2006), and these are both considered candidate critical
effects. Neurodevelopmental effects are a distinct type among developmental effects. Thus, the
next most sensitive endpoints of decreased rearing post-exposure in mice (Fredricksson et al.
1993), increased exploration post-exposure in rats (Taylor et al. 1985) and decreased myelination
in the hippocampus of rats (Isaacson and Taylor 1989) are also considered candidate critical
effects.
6/22/2009
1055
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 Table 5.1.5. Developmental effects in studies suitable for dose-response, and corresponding cRfCs and cRfDs.
Effect type Species POD PODa UFSC UFis UFh UF|0aei UFdb UFcompb cRfC cRfD Effect; comments
Supporting studies type (ppm) (mg/kg/d)
Pre- and post-natal
mortality
George et al. 1985
mouse
NOAEL
362
1
10
10
1
100
3.6
t perinatal mortality
Narotsky et al. 1995
rat
LOAEL
475
1
10
10
10
1000
0.48
post-natal mortality; Manson et al. 1984 cRfD preferred for
same endpoint due to NOAEL vs. LOAEL
Manson et al. 1984
rat
NOAEL
100
1
10
10
1
100
1.0
t neonatal death
Healey etal. 1982
rat
LOAEL
17
1
3
10
10
300
0.057
resorptions
Narotsky et al. 1995
rat
BMDL
469
1
10
10
1
100
4.7
pre-natal loss; BMR=1% extra risk
rat
BMDL
32.2
1
10
10
1
100
0.32
resorptions; BMR=1% extra risk
Pre- and post-natal
arowth
Healey etal. 1982
rat
LOAEL
17
1
3
10
10
300
0.057
[ fetal weight; skeletal effects
Narotsky et al. 1995
rat
NOAEL
844
1
10
10
1
100
8.4
[ fetal weight
George et al. 1985
mouse
NOAEL
362
1
10
10
1
100
3.6
I fetal weight
George et al. 1986
rat
BMDL
79.7
1
10
10
1
100
0.80
I BW at d21; BMR=5% decrease
Conaenital defects
Narotsky et al. 1995
rat
BMDL
60.1
1
10
10
1
100
0.60
eye defects; low BMR (1 %), but severe effect and low bkgd
rate (<1%)
Johnson et al. 2003
rat
BMDL
0.0146
1
10
10
1
100
0.00015
heart malformations (litters); BMR=10% extra risk (only ~1/10
from each litter affected); highest dose group (1000-fold
higher than next highest) dropped to improve model fit.
rat
BMDL
0.0207
1
10
10
1
100
0.00021
heart malformations (pups); BMR=1% extra risk; preferred
due to accounting for intra-litter effects via nested model and
pups being the unit of measure; highest dose group (1000-
fold higher than next highest) dropped to improve model fit
Developmental
neurotoxicity
George et al. 1986
rat
BMDL
72.6
1
10
10
1
100
0.73
I locomotor activity; BMR = doubling of traverse time; results
from females (males similar with BMDL=92)
Fredricksson et al. 1993
mouse
LOAEL
50
3
10
10
10
3000
0.017
I rearing post-exp; pup gavage dose; No effect at tested
doses on locomotion behavior; UFsc=3 because exposure
only during PND10-16
Taylor et al. 1985
rat
LOAEL
45
1
10
10
10
1000
0.045
t exploration post-exp; estimated dam dose; Less sensitive
than Isaacson&Taylor (1989), but included because exposure
is pre-weaning, so can utilize PBPK model.
Isaacson&Taylor 1989
rat
LOAEL
16
1
10
10
10
1000
0.016
1 myelination in hippocampus; estimated dam dose
Developmental
immunotoxicitv
Peden-Adams et al. 2006
mouse
LOAEL
0.37
1
10
10
10
1000
0.00037
I PFC, fDTH; POD is estimated dam dose (exp thruout gest
and lactation + to 3 or 8 wks of age); UF loael = 10 since f
DTH and also multiple immuno effects
6/22/2009
1056
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
a Adjusted to continuous exposure unless otherwise noted. For inhalation studies, adjustments yield a POD that is a human equivalent concentration as recommended for a Category 3 gas in U.S. EPA
(1994) in the absence of PBPK modeling. Same units as cRfC (ppm) or cRfD (mg/kg/d).
b Product of individual uncertainty factors.
UFsc = subchronic-to-chronic UF; UFis = interspecies UF; UFh = human variability UF; UF|0aei = LOAEL-to-NOAEL UF; UFdb = database UF
Shaded studies/endpoints were selected as candidate critical effects/studies.
6/22/2009
1057
-------
1
2
3
4
5
6
7
8
9
10
11
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.2.9 Summary of cRfCs, cRfDs, and Candidate Critical Effects
An overall summary of the cRfCs, cRfDs, and candidate critical effects across the health
effect domains is shown in Table 5.1.6-5.1.7. These tables present, for each type of non-cancer
effect, the relative ranges of the cRfC and cRfD developed for the different endpoints. The
candidate critical effects selected above for each effect domain are shown in bold. As discussed
above, these effects were generally selected to represent the most sensitive endpoints, across
species where possible. From these candidate critical effects, candidate reference values based
on internal dose metrics from the PBPK model (p-cRfCs and p-cRfDs) were developed where
possible. Application of the PBPK model is discussed in the next section.
6/22/2009
1058
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.6. Ranges of cRfCs based on applied dose for various noncancer effects associated with inhalation TCE exposure.
cRfC range
(ppm)
Neurological
Systemic/organ-
specific
Immunological
Reproductive
Developmental
10-100
impaired visual
discrimination (rat)
1-10
kidney
meganucleocytosis
(rat)
| kidney weight
(mouse)
0.1-1
ototoxicity (rat)
hyperactivity (rat)
changes in locomotor
activity (rat)
trigeminal nerve
effects (human)
impaired visual function
(rabbit)
I regeneration of
sciatic nerve (rat)
| liver weight (rat)
| liver weight (mouse)
| kidney weight (rat)
I PFC response (rat)
i maternal body weight
gain (rat)
t abnormal sperm
(mouse)
pre/post-implantation
losses (male rat exp)
0.01-0.1
I regeneration of
sciatic nerve
(mouse)
disturbed wakefulness
(rat)
autoimmune changes
(MRL -Ipr/lpr
mouse)
effects on epididymis
epithelium (mouse)
I fertilization (male
mouse exp)
testes & sperm effects
(rat)
hyperzoospermia
(human)
resorptions (female rat)
I fetal weight (rat)
skeletal effects (rat)
Endpoints in bold were selected as candidate critical effects (see sections 5.1.2.1-5.1.2.8).
6/22/2009
1059
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.7. Ranges of cRfDs based on applied dose for various noncancer effects associated with oral TCE exposure.
cRfD range
(mg/kg/day)
Neurological
Systemic/organ-
specific
Immunological
Reproductive
Developmental
1-10
t neuromuscular changes
(rat)
I BW (mouse)
i humoral response to
sRBC (mouse)
i testis/seminal vesicle
weight (mouse)
i sperm motility (mouse)
t testis/epididymis
weight (rat)
i litters/pair (rat)
i live pups/litter (rat)
i BW gain (rat)
i copulatory
performance (rat)
i fetal weight (rat)
pre-natal loss (rat)
i fetal weight (mouse)
t neonatal mortality
(mouse, rat)
0.1 - 1
t # rears (rat)
t foot splaying (rat)
trigeminal nerve effect
(rat)
| liver weight (mouse)
I BW (mouse)
| BW (rat)
toxic nephropathy &
meganucleocytosis
(other rat strains/sexes
& mouse)
signs of autoimmune
hepatitis (MRL +/+
mouse)
inflamm in various
tissues (MRL +/+
mouse)
delayed parturition
(rat)
I mating (rat)
| BW atPND21 (rat)
I locomotor activity (rat)
eye defects (rat)
resorptions (rat)
0.01-0.1
degeneration of
dopaminergic
neurons (rat)
toxic nephropathy
(female Marshall rat)
I cell-mediated
response to sRBC
(mouse)
I stem cell bone
marrow
recolonization
(mouse)
I ability of sperm to
fertilize (rat)
| exploration (post exp)
(rat)
I rearing (post exp)
(mouse)
I myelination in
hippocampus (rat)
0.001-0.01
demyelination in
hippocampus (rat)
| anti-dsDNA & anti-
ssDNA Abs (early
marker for SLE)
(mouse)
10"4- 0.001
I thymus weight
(mouse)
immunotox (J, PFC, |
DTH) (B6C3F1
mouse)
heart malformations (rat)
Endpoints in bold were selected as candidate critical effects (see sections 5.1.2.1-5.1.2.8).
6/22/2009
1060
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.3 Application of PBPK model to inter- and intra-species extrapolation for candidate
critical effects
For the candidate critical effects, the use of PBPK modeling of internal doses could
justify, where appropriate, replacement of the uncertainty factors for pharmacokinetic inter- and
intra-species extrapolation. For more details on PBPK modeling used to estimate levels of dose
metrics corresponding to different exposure scenarios in rodents and humans, see Section 3.5.
Quantitative analyses of the uncertainties, from a Bayesian analysis of the PBPK model, are
discussed separately in Section 5.1.4.
5.1.3.1 Selection of dose metrics for different endpoints
One area of scientific uncertainty in non-cancer dose-response assessment is the
appropriate scaling between rodent and human doses for equivalent responses. Another way one
could regard the UF for inter-species extrapolation discussed above for applied dose is that it
reflects the combination of an adjustment factor due to the expected scaling of toxicologically-
equivalent doses across species (commonly attributed to pharmacokinetics) and a factor
accounting for uncertainty in the appropriate inter-species extrapolation for specific noncancer
effects from a specific chemical exposure (commonly attributed to pharmacodynamics). For
considering how to scale internal doses predicted by a PBPK model across species, it is useful to
consider two possible interpretations of the "adjustment" component (UFis_adj), and their
consequent implications for the remaining "uncertainty" component (UFis_unc) of the interspecies
UF.
The first (denoted "empirical dosimetry") interpretation is that the "adjustment" is based
on the empirical finding that scaling the delivered dose rate by body weight to the 3/4 power
results in equivalent toxicity (e.g., Travis and White, 1988; USEPA, 1992), since the 3-fold
factor comprising this UFis.adj component is similar to what would result from body weight -3/4
power-scaling from rats to humans (an adjustment of mg/kg/d dose by (70/0.4)'/4 = 3.6). The
scaling of dose by body weight to the 3/4 power is supported biologically by data showing that the
rates of both kinetic and dynamic physiologic processes are generally consistent with 3/4 power of
body weight scaling across species (USEPA 1992). Note also that this applies to inhalation
exposure because the delivered dose rate in that case is the air concentration multiplied by the
ventilation rate, which scales by body weight to the 3/4 power. Applying this interpretation to
internal doses would imply that the dose rate of the active moiety delivered to the target tissue,
scaled by body weight to the 3/4 power, would be assumed to result in equivalent responses.
Under this interpretation, the "uncertainty" component, UFis_unc, of the interspecies UF (which is
6/22/2009
1061
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
still retained for reference values using PBPK modeling) reflects the possible deviations from the
empirically-based "adjustment" due to the kinetics or dynamics for a particular non-cancer effect
for a particular chemical in the particular species from which human risk is being extrapolated.
The second (denoted "concentration equivalence dosimetry") interpretation is consistent
with the further hypothesis that the empirical finding (and hence the "adjustment" component of
the inter-species UF) is largely pharmacokinetically-driven, so UFis_adj = UFis-pk (e.g., IPCS,
2005). Under this interpretation, it is hypothesized that, due to the body weight to the 3/4 scaling
of physiologic flows (cardiac output, ventilation rate, glomerular filtration, etc.) and metabolic
rates (enzyme-mediated biotransformation), the "adjustment" component is intended to result in
average internal concentrations of the active moiety at the target tissue, which in turn results in
equivalent toxicity (NRC, 1986; NRC, 1987). Applying this interpretation to internal doses
would imply that equal (average) concentrations of the active moiety or moieties at the target
tissue would result in equivalent responses. Under this interpretation, the "uncertainty"
component of the interspecies UF (which is still retained for reference values using PBPK
modeling) reflects the possible deviations from the empirically-based "adjustment" due to the
pharmacodynamics (and not pharmacokinetics) for a particular non-cancer effect for a particular
chemical in the particular species from which human risk is being extrapolated, so UFis_unc =
UFis-pd-
To the extent that production and clearance of the active moiety or moieties all scale by
body weight to the 3/4 power, these two dosimetry interpretations both lead to the same dose
metrics and quantitative results. However, these interpretations may lead to different
quantitative results when there are deviations of the underlying physiologic or metabolic
processes from body weight to the 3/4 power scaling. For instance, as discussed in Section 3.5,
the PBPK model predictions for AUC of TCE in blood deviate from the body weight to the 3/4
scaling (the scaling is closer to mg/kg/d than mg/kgy4/d), so use of this dose metric implicitly
assumes the "concentration equivalence dosimetry." In addition, as discussed below, in most
cases involving TCE metabolites, only the rate of production of the active moiety(ies) or the rate
of transformation through a particular metabolic pathway can be estimated using the PBPK
model, and the actual concentration of the active moiety(ies) cannot be estimated due to data
limitations. Under "empirical dosimetry," these metabolism rates, which are estimates of the
systemic or tissue-specific delivery of the active moiety(ies), would be scaled by body weight to
the 3/4 power to yield equivalent toxicological response. Under "concentration equivalence
dosimetry," additional assumptions about the rate of clearance are necessary to specify the
scaling that would yield concentration equivalence. In the absence of data, active metabolites are
assumed to be sufficiently stable so that clearance is via enzyme-catalyzed transformation or
systemic excretion (e.g., blood flow, glomerular filtration), which scale approximately by body
6/22/2009
1062
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
weight to the 3/4 power. Therefore, under "concentration equivalence dosimetry," the metabolism
rates would also be scaled by body weight to the 3/4 power in the absence of additional data.
For toxicity that is associated with local (in situ) production of "reactive" metabolites
whose concentrations cannot be directly measured in the target tissue, an alternative approach,
under "concentration equivalence dosimetry," of scaling by unit tissue mass has been proposed
(e.g., Andersen et al. 1987). As discussed by Travis (1990), scaling the rate of local metabolism
across species and individuals by tissue mass is appropriate if the metabolites are sufficiently
reactive and are cleared by "spontaneous" deactivation (i.e., changes in chemical structure
without the need of biological influences). Thus, use of this alternative scaling approach requires
that (i) the active moiety or moieties do not leave the target tissue in appreciable quantities (i.e.,
are cleared primarily by in situ transformation to other chemical species and/or binding
to/reactions with cellular components); and (ii) the clearance of the active moieties from the
target tissue is governed by biochemical reactions whose rates are independent of body weight
(e.g., purely chemical reactions). If these conditions are met, then under the "concentration
equivalence dosimetry," the relevant metabolism rates estimated by the PBPK model would be
scaled by tissue mass, rather than by body weight to the 3/4 power.
To summarize, the internal dose metric for equivalent toxicological responses across
species can be specified by invoking one of two alternative interpretations of the "adjustment"
component of the inter-species UF: "empirical dosimetry" based on the rate at which the active
moiety(ies) is(are) delivered to the target tissue scaled by body weight to the 3/4 power or
"concentration equivalence dosimetry" based on matching internal concentrations of the active
moiety(ies) in the target tissue. If the active moiety(ies) is TCE itself or a putatively reactive
metabolite, the choice of interpretation will affect the choice of internal dose metric. In the
discussions of dose metric selections for the individual endpoints below, the implications of both
"empirical dosimetry" and "concentration equivalence dosimetry" are discussed.
The use of these dose metrics was then also deemed to obviate the need for the
pharmacokinetic component, UFh-pk, of the UF for human (intraspecies) variability. Because all
the dose metrics used for TCE are for adults, and the dose metrics are not very sensitive to the
plausible range of adult body weight, for convenience the body weight 3/4 scaling used for inter-
species extrapolation was retained for characterization of human variability. However, it should
be emphasized that this intra-species characterization is of pharmacokinetics only, and not
pharmacody nami cs.
In general, an attempt was made to use tissue-specific dose metrics representing
particular pathways or metabolites identified from available data on the role of metabolism in
toxicity for each endpoint (discussed in more detail below). The selection was limited to dose
metrics that could be adequately estimated by the PBPK model (see Section 3.5). For most
6/22/2009
1063
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
endpoints, sufficient information on the role of metabolites or MOA was not available to identify
likely relevant dose metrics, and more "upstream" metrics representing either parent compound
or total metabolism had to be used. The "primary" or "preferred" dose metric referred to in
subsequent Tables has the greater biological support for its involvement in toxicity, whereas
"alternative" dose metrics are those which may also be plausibly involved (discussed further
below). A discussion of the dose metrics selected for particular non-cancer endpoints follows.
5.1.3.1.1 Kidney toxicity (meganucleuocytosis, increased kidney weight, toxic nephropathy)
As discussed in Sections 4.3.6-4.3.7, there is sufficient evidence to conclude that TCE-
induced kidney toxicity is caused predominantly by GSH conjugation metabolites either
produced in situ in or delivered systemically to the kidney. As discussed in Section 3.3.3.2,
bioactivation of DCVG, DCVC, and NAcDCVC within the kidney, either by beta-lyase, FMO,
or P450s, produces reactive species, any or all of which may cause nephrotoxicity. Therefore,
multiple lines of evidence support the conclusion that renal bioactivation of DCVC is the
preferred basis for internal dose extrapolations for TCE-induced kidney toxicity. However,
uncertainties remain as to the relative contribution from each bioactivation pathway; and
quantitative clearance data necessary to calculate the concentration of each species are lacking.
Under "empirical dosimetry," the rate of renal bioactivation of DCVC would be scaled by
body weight to the 3/4 power. As discussed above, under "concentration equivalence dosimetry,"
when the concentration of the active moiety cannot be estimated, qualitative data on the nature of
clearance of the active moiety or moieties can be used to inform whether to scale the rate of
metabolism by body weight to the 3/4 power or by the target tissue weight. For the beta-lyase
pathway, Dekant et al. (1988) reported in trapping experiments that the postulated reactive
metabolites decompose to stable (unreactive) metabolites in the presence of water. Moreover,
the necessity of a chemical trapping mechanism to detect the reactive metabolites suggests a very
rapid reaction such that it is unlikely that the reactive metabolites leave the site of production.
Therefore, these data support the conclusion that, for this bioactivation pathway, clearance is
chemical in nature and hence species-independent. If this were the only bioactivation pathway,
then scaling by kidney weight would be supported. With respect to the FMO bioactivation
pathway, Sausen and Elfarra (1991) reported that after direct dosing of the postulated reactive
sulfoxide (DCVC sulfoxide), the sulfoxide was detected as an excretion product in bile. These
data suggest that reactivity in the tissue to which the sulfoxide was delivered (the liver, in this
case) is insufficient to rule out a significant role for enzymatic or systemic clearance. Therefore,
according to the criteria outlined above, for this bioactivation pathway, the data support scaling
the rate of metabolism by body weight to the % power. For P450-mediated bioactivation
producing N-acetyl DCVC (mercapturic acid) sulfoxide, the only relevant data on clearance are
from a study of the structural analogue to DCVC, FDVE (Sheffels et al. 2004), which reported
6/22/2009
1064
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
that the postulated reactive sulfoxide was detected in urine. This suggests that the sulfoxide is
sufficiently stable to be excreted by the kidney and supports the scaling of the rate of metabolism
by body weight to the 3/4 power.
Therefore, because the contributions to TCE-induced nephrotoxicity from each possible
bioactivation pathway are not clear, and, even under "concentration equivalence dosimetry," the
scaling by body weight to the 3/4 power is supported for two of the three bioactivation pathways,
it is decided here to scale the DCVC bioactivation rate by body weight to the 3/4 power. The
primary internal dose metric for TCE-induced kidney tumors is thus the weekly rate of DCVC
bioactivation per unit body weight to the 3/4 power (ABioactDCVCBW34 [mg/kgyYwk]).
However, it should be noted that due to the larger relative kidney weight in rats as compared to
humans, scaling by kidney weight instead of body weight to the 3/4 power would only change the
quantitative inter-species extrapolation by about 2-fold,39 so the sensitivity of the results to the
scaling choice is relatively small. In addition, quantitative estimates for this dose metric are only
available in rats and humans, and not in mice. Accordingly, this metric was only used for
extrapolating results from rat toxicity studies.
To summarize, under the "empirical dosimetry" approach, the underlying assumption for
the ABioactDCVCBW34 dose metric is that equalizing the rate of renal bioactivation of DCVC
(i.e., local production of active moiety(ies) in the target tissue), scaled by the 3/4 power of body
weight, accounts for the "adjustment" component of the interspecies UF and the
"pharmacokinetic" component of the intraspecies UF. Under "concentration equivalence
dosimetry," the underlying assumptions for the ABioactDCVCBW34 dose metric are that (i)
matching the average concentration of reactive species in the kidney accounts for the
"adjustment" component of the interspecies UF and the "pharmacokinetic" component of the
intraspecies UF ; and (ii) the rates of clearance of these reactive species scale by the 3/4 power of
body weight (e.g., assumed for enzyme-activity or blood-flow).
39 The range of the difference is 2.1-2.4-fold using the posterior medians for the relative kidney weight in rats and
humans from the PBPK model described in Section 3.5 (Table 3.5.7), and body weights of 0.3-0.4 kg for rats and
60-70 kg for humans.
6/22/2009
1065
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
An alternative dose metric that also involves the GSH conjugation pathway is the amount
of GSH conjugation scaled by the 3/4 power of body weight (AMetGSHBW34 [mg/kgyYwk]).
This dose metric uses the total flux of GSH conjugation as the toxicologically-relevant dose, and
thus incorporates any direct contributions from DCVG and DCVC, which are not addressed in
the DCVC bioactivation metric. Under the "empirical dosimetry" approach, the underlying
assumption for the AMetGSHBW34 dose metric is that equalizing the (whole body) rate of
production of GSH conjugation metabolites (i.e., systemic production of active moiety(ies)),
scaled by the 3/4 power of body weight, accounts for the "adjustment" component of the
interspecies UF and the "pharmacokinetic" component of the intraspecies UF. Under
"concentration equivalence dosimetry," the AMetGSHBW34 dose metric is consistent with the
assumptions that (i) matching the same average concentration of the (relatively) stable upstream
metabolites DCVG or DCVC in the kidney (the PBPK model assumes all DCVG and DCVC
produced translocates to the kidney) accounts for the "adjustment" component of the interspecies
UF and the "pharmacokinetic" component of the intraspecies UF; and (ii) the rate of clearance of
DCVG or DCVC scales by the 3/4 power of body weight (as is assumed for enzyme activity or
blood flow). Because of the lack of availability of the DCVC bioactivation dose metric in mice,
the GSH conjugation metric is used as the primary dose metric for the nephrotoxicity endpoint in
studies of mice.
Another alternative dose metric is the total amount of TCE metabolism (oxidation and
GSH conjugation together) scaled by the 3/4 power of body weight (TotMetabBW34
[mg/kgyYwk]). This dose metric uses the total flux of TCE metabolism as the toxicologically-
relevant dose, and thus incorporates the possible involvement of oxidative metabolites, acting
either additively or interactively, in addition to GSH conjugation metabolites in nephrotoxicity
(see Section 4.3.6). However, this dose metric is given less weight than those involving GSH
conjugation because, as discussed in Sections 4.3.6, the weight of evidence supports the
conclusion that GSH conjugation metabolites play a predominant role in nephrotoxicity. Under
the "empirical dosimetry" approach, the underlying assumption for the TotMetabBW34 dose
metric is that equalizing the (whole body) rate of production of all metabolites (i.e., systemic
production (and distribution) of active moiety(ies)), scaled by the 3/4 power of body weight,
accounts for the "adjustment" component of the interspecies UF and the "pharmacokinetic"
component of the intraspecies UF. Under "concentration equivalence dosimetry," the
TotMetabBW34 dose metric is consistent with the assumptions that (i) the relative proportions
and blood:tissue partitioning of the active metabolites is similar across species; (ii) matching the
average concentration of one or more metabolites in the kidney accounts for the "adjustment"
component of the interspecies UF and the "pharmacokinetic" component of the intraspecies UF;
6/22/2009
1066
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
and (iii) the rate of clearance of active metabolites scales by the 3/4 power of body weight (e.g.,
assumed for enzyme-activity or blood-flow).
5.1.3.1.2 Liver weight increases (hepatomegaly)
As discussed in Section 4.4.6, there is substantial evidence that oxidative metabolism is
involved in TCE hepatotoxicity, based primarily on similarities in non-cancer effects with a
number of oxidative metabolites of TCE (e.g., CH, TCA, and DC A). While TCA is a stable,
circulating metabolite, CH and DCA are relatively short-lived, although enzymatically cleared
(see Section 3.3.3.1). As discussed in section 4.4.6.2.1, there is substantial evidence that TCA
alone does not adequately account for the hepatomegaly induced by TCE; therefore, unlike in
previous dose-response analyses (Barton and Clewell 2000, Clewell and Andersen 2004), the
AUC of TCA in plasma or in liver were not considered as dose metrics. However, there are
inadequate data across species to quantify the dosimetry of CH and DCA, and other
intermediates of oxidative metabolism (such as TCE-oxide or dichloroacetylchloride) may be
involved in hepatomegaly. Thus, due to uncertainties as to the active moiety(ies), but given the
strong evidence associating TCE liver effects with oxidative metabolism in the liver, hepatic
oxidative metabolism is the preferred basis for internal dose extrapolations of TCE-induced liver
weight increases. Under "empirical dosimetry," the rate of hepatic oxidative metabolism would
be scaled by body weight to the 3/4 power. As discussed above, under "concentration equivalence
dosimetry," when the concentration of the active moiety cannot be estimated, qualitative data on
the nature of clearance of the active moiety or moieties can be used to inform whether to scale
the rate of metabolism by body weight to the 3/4 power or by the target tissue weight. However,
several of the oxidative metabolites are stable and systemically available, and several of those
that are cleared rapidly are metabolized enzymatically, so, according to the criteria discussed
above, there are insufficient data to support the conclusions that the active moiety or moieties do
not leave the target tissue in appreciable quantities and are cleared by mechanisms whose rates
are independent of body weight. Thus, scaling the rate of oxidative metabolism by body weight
to the 3/4 power would also be supported under "concentration equivalence dosimetry."
Therefore, the primary internal dose metric for TCE-induced liver weight changes is selected to
be the weekly rate of hepatic oxidation per unit body weight to the 3/4 power (AMetLivlBW34
[mg/kgyYwk]). The use of this dose metric is also supported by the analysis in Section 4.4.6.2.1
showing much more consistency in the dose-response relationships for TCE-induced
hepatomegaly across studies and routes of exposure using this metric and the total oxidative
metabolism dose metric (discussed below) as compared to the AUC of TCE in blood. It should
be noted that due to the larger relative liver weight in mice as compared to humans, scaling by
liver weight instead of body weight to the 3/4 power would only change the quantitative inter-
6/22/2009
1067
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
species extrapolation by about 4-fold,40 so the sensitivity of the results to the scaling choice is
relatively modest.
To summarize, under the "empirical dosimetry" approach, the underlying assumption for
the AMetLivlBW34 dose metric is that equalizing the rate of hepatic oxidation of TCE (i.e.,
local production of active moiety(ies) in the target tissue), scaled by the 3/4 power of body weight,
accounts for the "adjustment" component of the interspecies UF and the "pharmacokinetic"
component of the intraspecies UF. Under "concentration equivalence dosimetry," the
AMetLivlBW34 dose metric is consistent with the assumptions that (i) oxidative metabolites are
primarily generated in situ in the liver; (ii) the relative proportions and blood:tissue partitioning
of the active oxidative metabolites are similar across species; (iii) matching the average
concentration of the active oxidative metabolites in the liver accounts for the "adjustment"
component of the interspecies UF and the "pharmacokinetic" component of the intraspecies UF;
and (iv) the rates of clearance of the active oxidative metabolites scale by the 3/4 power of body
weight (e.g., assumed for enzyme-activity or blood-flow).
It is also known that the lung has substantial capacity for oxidative metabolism, with
some proportion of the oxidative metabolites produced there entering systemic circulation. Thus,
it is possible that extra-hepatic oxidative metabolism can contribute to TCE-induced
hepatomegaly. Therefore, the total amount of oxidative metabolism of TCE scaled by the 3/4
power of body weight (TotOxMetabBW34 [mg/kgy7wk]) was selected as an alternative dose
metric (the justification for the body weight to the 3/4 power scaling is analogous to that for
hepatic oxidative metabolism, above). Under the "empirical dosimetry" approach, the
underlying assumption for the TotOxMetabBW34 dose metric is that equalizing the rate of total
oxidation of TCE (i.e., systemic production of active moiety(ies)), scaled by the 3/4 power of
body weight, accounts for the "adjustment" component of the interspecies UF and the
"pharmacokinetic" component of the intraspecies UF. Under "concentration equivalence
dosimetry," this dose metric is consistent with the assumptions that (i) oxidative metabolites may
be generated in situ in the liver or delivered to the liver via systemic circulation; (ii) the relative
proportions and blood:tissue partitioning of the active oxidative metabolites is similar across
species; (iii) matching the average concentration of the active oxidative metabolites in the liver
accounts for the "adjustment" component of the interspecies UF and the "pharmacokinetic"
component of the intraspecies UF; and (iv) the rates of clearance of the active oxidative
metabolites scale by the 3/4 power of body weight (e.g., enzyme-activity or blood-flow).
40 The range of the difference is 3.5-3.9-fold using the posterior medians for the relative liver weight in mice and
humans from the PBPK model described in Section 3.5 (Table 3.5.7), and body weights of 0.03-0.04 kg for mice
and 60-70 kg for humans.
6/22/2009
1068
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.3.1.3 Developmental toxicity - heart malformations
As discussed in Section 4.7.3.2.1, several studies have reported that the prenatal exposure
to TCE oxidative metabolites TCA or DCA also induces heart malformations, suggesting that
oxidative metabolism is involved in TCE-induced heart malformations. However, there are
inadequate data across species to quantify the dosimetry of DCA, and it is unclear if other
products of TCE oxidative metabolism are involved. Therefore, the total amount of oxidative
metabolism of TCE scaled by the 3/4 power of body weight (TotOxMetabBW34 [mg/kgy4/wk])
was selected as the primary dose metric. Under the "empirical dosimetry" approach, the
underlying assumption for the TotOxMetabBW34 dose metric is that equalizing the rate of total
oxidation of TCE (i.e., systemic production of active moiety(ies), the same proportion of which
is assumed to be delivered to the fetus across species/individuals), scaled by the 3/4 power of body
weight, accounts for the "adjustment" component of the interspecies UF and the
"pharmacokinetic" component of the intraspecies UF. Under "concentration equivalence
dosimetry," this dose metric is consistent with the assumptions that (i) oxidative metabolites are
delivered to the fetus via systemic circulation; (ii) the relative proportions and blood:tissue
partitioning of the active oxidative metabolites is similar across species; (iii) matching the
average concentration of the active oxidative metabolites in the fetus accounts for the
"adjustment" component of the interspecies UF and the "pharmacokinetic" component of the
intraspecies UF; and (iv) the rates of clearance of the active oxidative metabolites scale by the 3/4
power of body weight (e.g., enzyme-activity or blood-flow).
An alternative dose metric that is considered here is the AUC of TCE in (maternal) blood
(AUCCBld [mg-hr/l/d]). Under either "empirical dosimetry" or "concentration equivalence
dosimetry," this dose metric would account for the possible role of local metabolism, which is
determined by TCE delivered in blood via systemic circulation to the target tissue (the flow rate
of which scales as body weight to the 3/4 power). Moreover, the placenta is a highly perfused
tissue, and TCE is known to cross the placenta to the fetus, with rats showing similar (within 2-
fold) maternal and fetal blood TCE concentrations (see Section 3.2). Under the "concentration
equivalence dosimetry," this dose metric also accounts for the possible role of TCE itself. This
dose metric of AUC of TCE in blood is therefore consistent with the assumptions that (i)
maternal blood:fetal partitioning of TCE is similar across species, so that similar blood
concentrations imply similar fetal concentrations; (ii) to the extent that local metabolism in the
placenta or fetus is involved, both in situ metabolism of TCE and clearance of active oxidative
metabolites scale by the 3/4 power of (adult) body weight (e.g., enzyme-activity or blood-flow);
and therefore, (iii) matching the average concentrations of TCE in blood accounts for the
"adjustment" component of the interspecies UF and the "pharmacokinetic" component of the
intraspecies UF.
6/22/2009
1069
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.3.1.4 Reproductive toxicity - decreased ability of sperm to fertilize oocytes
The decreased ability of sperm to fertilize oocytes observed by DuTeaux et al. (2004)
occurred in the absence of changes in combined testes/epididymes weight, sperm concentration
or motility, or histological changes in the testes or epididymes. However, there was evidence of
oxidative damage to the sperm, and DuTeaux et al. (2003) previously reported the ability of the
rat epididymis and efferent ducts to metabolize TCE oxidatively. Based on this evidence,
DuTeaux et al. (2004) hypothesize that the decreased ability to fertilize is due to oxidative
damage to the sperm from local metabolism. Thus, the primary dose metric for this endpoint is
selected to be the AUC of TCE in blood (AUCCBld [mg-hr/l/d]), based on the assumption that in
situ oxidation of systemically-delivered TCE (the flow rate of which scales as body weight to the
3/4 power) is the determinant of toxicity. Under either "empirical dosimetry" or "concentration
equivalence dosimetry," this dose metric is therefore consistent with the assumptions that (i)
blood:tissue partitioning of TCE is similar across species, so that similar blood concentrations
imply similar tissue concentrations; (ii) in situ oxidation of TCE and clearance of active
oxidative metabolites scale by the % power of body weight (e.g., enzyme-activity or blood-flow);
and, therefore, (iii) matching the average concentrations of TCE in blood accounts for the
"adjustment" component of the interspecies UF and the "pharmacokinetic" component of the
intraspecies UF.
Because metabolites causing oxidative damage may be delivered systemically to the
target tissue, an alternative dose metric that is considered here is total oxidative metabolism of
TCE scaled by the % power of body weight (TotOxMetabBW34 [mg/kgy4/d]). Under the
"empirical dosimetry" approach, the underlying assumption for the TotOxMetabBW34 dose
metric is that equalizing the rate of total oxidation of TCE (i.e., systemic production of active
moiety(ies), the same proportion of which is assumed to be delivered to the target tissue across
species/individuals), scaled by the % power of body weight, accounts for the "adjustment"
component of the interspecies UF and the "pharmacokinetic" component of the intraspecies UF.
Under "concentration equivalence dosimetry," this dose metric is consistent with the
assumptions that (i) oxidative metabolites are delivered to the target tissue via systemic
circulation; (ii) the relative proportions and blood:tissue partitioning of the active oxidative
metabolites is similar across species; (iii) matching the average concentrations of the active
oxidative metabolites in the target tissue accounts for the "adjustment" component of the
interspecies UF and the "pharmacokinetic" component of the intraspecies UF; and (iv) the rates
of clearance of the active oxidative metabolites scale by the % power of body weight (e.g.,
enzyme-activity or blood-flow). Because oxidative metabolites make up the majority of TCE
metabolism, total metabolism gives very similar results (within 1.2-fold) to total oxidative
metabolism and is therefore not included as a dose metric.
6/22/2009
1070
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.3.1.5 Other reproductive and developmental effects and neurological effects and
immunologic effects
For all other candidate critical endpoints listed in Tables 5.1.6-5.1.7, including
developmental effects other than heart malformations and reproductive effects other than
decreased ability of sperm to fertilize, there is insufficient information for site-specific
determinations of an appropriate dose metric. While TCE metabolites and/or metabolizing
enzymes have been reported in some of these tissues (e.g., male reproductive tract), their general
roles in toxicity in the respective tissues have not been established. The choice of total
metabolism as the primary dose metric is based on the observation that, in general, TCE toxicity
is associated with metabolism rather than the parent compound. It is acknowledged that there is
no compelling evidence that definitively establishes one metric as more plausible than the other
in any particular case. Nonetheless, as a general inference in the absence of specific data, total
metabolism is viewed as more likely to be involved in toxicity than the concentration of TCE
itself.
Therefore, given that the majority of the toxic and carcinogenic responses in many tissues
to TCE appears to be associated with metabolism, the primary dose metric is selected to be total
metabolism of TCE scaled by the 3/4 power of body weight (TotMetabBW34 [mg/kgy4/d]). Under
the "empirical dosimetry" approach, the underlying assumption for the TotOxMetabBW34 dose
metric is that equalizing the rate of total oxidation of TCE (i.e., systemic production of active
moiety(ies), the same proportion of which is assumed to be delivered to the target tissue across
species/individuals), scaled by the 3/4 power of body weight, accounts for the "adjustment"
component of the interspecies UF and the "pharmacokinetic" component of the intraspecies UF.
Under "concentration equivalence dosimetry," this dose metric is consistent with the
assumptions that (i) metabolites are delivered to the target tissue via systemic circulation; (ii) the
relative proportions and blood:tissue partitioning of the active metabolites is similar across
species; (iii) matching the average concentrations of the active metabolites in the target tissue
accounts for the "adjustment" component of the interspecies UF and the "pharmacokinetic"
component of the intraspecies UF; and (iv) the rates of clearance of the active metabolites scale
by the 3/4 power of body weight (e.g., enzyme-activity or blood-flow). Because oxidative
metabolites make up the majority of TCE metabolism, total oxidative metabolism gives very
similar results (within 1.2-fold) to total metabolism and is therefore not included as a dose
metric.
An alternative dose metric that is considered here is the AUC of TCE in blood
(AUCCBld [mg-hr/l/d]). Under either "empirical dosimetry" or "concentration equivalence
dosimetry," this dose metric would account for the possible role of local metabolism, which is
determined by TCE delivered in blood via systemic circulation to the target tissue (the flow rate
6/22/2009
1071
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
of which scales as body weight to the 3/4 power). Under the "concentration equivalence
dosimetry," this dose metric also accounts for the possible role of TCE itself. This dose metric is
consistent with the assumption that matching the average concentrations of TCE in blood
accounts for the "adjustment" component of the interspecies UF and the "pharmacokinetic"
component of the intraspecies UF. This dose metric would also be most applicable to tissues
which have similar tissue:blood partition coefficients across and within species.
Because the PBPK model described in Section 3.5 did not include a fetal compartment,
the maternal internal dose metric is taken as a surrogate for developmental effects in which
exposure was before or during pregnancy (Taylor et al., 1985; Fredricksson et al., 1993;
Narotsky et al., 1995; Johnson et al, 2003). This was considered reasonable because TCE and
the major circulating metabolites (TCA, TCOH) appear to cross the placenta (see Sections 3.2,
3.3, and 4.9 [Ghantous et al. 1986, Fisher et al. 1989]), and maternal metabolizing capacity is
generally greater than that of the fetus (see Section 4.9). In the cases where exposure continues
after birth (Issacson and Taylor, 1989; Peden-Adams et al, 2006), no PBPK model-based internal
dose was used. Because of the complicated fetus/neonate dosing that includes transplacental,
lactational, and direct (if dosing continues post-weaning) exposure, the maternal internal dose is
no more accurate a surrogate than applied dose in this case.
5.1.3.2 Methods for inter- and intra-species extrapolation using internal doses
As shown in Figures 5.1.2 and 5.1.3, the general approach taken to use the internal dose
metrics in deriving human equivalent concentrations (HECs) and human equivalent doses
(HEDs) was to first apply the rodent PBPK model to get rodent values for the dose metrics
corresponding to the applied doses in a study reporting non-cancer effects. The internal dose
POD (iPOD) is then obtained either directly from the internal dose corresponding to the applied
dose LOAEL or NOAEL, or by dose-response modeling of responses with respect to the internal
doses to derive a BMDL in terms of internal dose. Separately, the human PBPK model is run for
a range of continuous exposures from 10"1 to 2xl03 ppm or mg/kg/d to obtain the relationship
between human exposure and internal dose for the same dose metric used for the rodent. The
human equivalent exposure (HEC or HED) corresponding to the iPOD is derived by
interpolation. It should be noted that median values of dose metrics were used for rodents,
whereas both median and 99th percentile values were used for humans. As discussed in Section
3.5, the rodent population model characterizes study-to-study variation, while, within a study,
animals with the same sex/species/strain combination were assumed to be identical
pharmacokinetically and represented by the group average (typically the only data reported).
Therefore, use of median dose metric values can be interpreted as assuming that the animals in
the non-cancer toxicity study were all "typical" animals and the iPOD is for a rodent that is
6/22/2009
1072
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
pharmacokinetically "typical." In practice, the use of median or mean internal doses for rodents
did not make much difference except when the uncertainty in the rodent dose metric was high.
The impact of the uncertainty in the rodent PBPK dose metrics is analyzed quantitatively in
Section 5.1.4.2.
The human population model characterizes individual-to-individual variation, in addition
to its uncertainty. The "median" value for the HEC or HED was calculated as a point of
comparison but was not actually used for derivation of candidate reference values. Because the
RfC and RfD are intended to characterize the dose below which a sensitive individual would
likely not experience adverse effects, the overall 99th percentile of the combined uncertainty and
variability distribution was used for deriving the HEC and HED (denoted HEC99 and HED99)
from each iPOD.41 As shown in Figures 5.1.2 and 5.1.3., the HEC99 or HED99 replaces the
quantity POD/(UFis_adj x UFh-pk) in the calculation of the RfC or RfD, i.e., the pharmacokinetic
components of the UFs representing inter-species extrapolation and human inter-individual
variability. As calculated, the extrapolated HEC99 and HED99 can be interpreted as being the
dose or exposure for which there is 99% likelihood that a randomly selected individual will have
an internal dose less than or equal to the iPOD derived from the rodent study. The separate
contributions of uncertainty and variability in the human PBPK model are analyzed
quantitatively, along with the uncertainty in the rodent PBPK dose metrics as mentioned above,
in Section 5.1.4.2.
Because they are derived from rodent internal dose estimates, the HEC and HED are
derived in the same manner independent of the route of administration of the original rodent
study. Therefore, a route-to-route extrapolation from an oral (inhalation) study in rodents to a
HEC (HED) in humans is straight-forward. As shown in Tables 5.1.8-5.1.13, route-to-route
extrapolation was performed for a number of endpoints with low cRfCs and cRfDs to derive p-
cRfDs and p-cRfCs.
For the candidate critical studies using human data (Chia et al., 1996), the PBPK model
was used only for intra-species extrapolation and route-to-route extrapolation. The internal dose
POD was defined as the internal dose of the median individual exposed at the applied dose POD
41 There is no explicit guidance on the selection of the percentile for human toxicokinetic variability. Ideally, all
sources of uncertainty and variability would be included, and a lower percentile would be selected so as to be more
in line with the levels of risk at which cancer dose-response is typically characaterized (e.g., 106 to 104). However,
only toxicokinetic variability is assessed quantitatively. In addition, percentiles greater than the 99th are likely to be
progressively more uncertain due to the unknown shape of the tail of the input distributions for the PBPK model
parameters (which were largely assumed to be normal or lognormal), and the fact that only 42 individuals were
incorporated in the PBPK model for characterization of inter-individual variability (see Section 3.5). This concern
is somewhat ameliorated because the candidate reference values also incorporate use of UFs to account for inter-
and intra-species toxicodynamic sensitivity.
6/22/2009
1073
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 (LOAEL or BMDL). Then, as with the rodent studies, the HEC99 or HED99 is the lower 99th
2 percentile applied dose corresponding to same internal dose.
6/22/2009
1074
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
distribution (combined
ncertainty and variability)
Rodent
non-cancer
Rodent
model
parameters
study
experimental
paradigm
[distribution
PBPK
model
Rodent
non-cancer
study
responses
Rodent
internal
dose
edian
Dose-Response Model
or
OAEUNOAE
iPOD (internal
dose unit)
BMDLor
LOAEL or
NOAEL
0.1-2000 ppm
in air or
0.1-2000
mg/kg-d
continuous
exposure
istribution
Human
model
parameters
PBPK
model
Human
internal
dose as
function of
applied do
distribution (separate
luncertainty and variability)
invert functions of dose
or concentration
Overall
median ^
"Typical"
human internal
dose as
function
of applied
dose
Overall
'^99%-ile
"Sensitive"
human internal
dose as
function
of applied
dose
"Typical"
human
equivalent
dose or
concentration
A.
"Sensitive"
human
equivalent
dose or
concentration
2
3
4
5
6
7
HEC50 or
HED50
(replaces
POD/UF,
is-adj
HEC99 or
HED99
[replaces
P°D/(UFis.adj*UFh.pk)]
Figure 5.1.2
Flow-chart for dose-response analyses of rodent non-cancer effects using PBPK model-based
dose metrics. Square nodes indicate point values, circle nodes indicate distributions, and the
inverted triangle indicates a (deterministic) functional relationship.
6/22/2009
1075
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Uncertainty &
variability
distribution
Human internal
dose
Human inhalation
exposure (ppm)
Rodent internal
dose
55\ed\an
Uncertain*
variabil1 y
distribut.on
Lower 99th
percentile
iPOD
Uncertainty &
variability
distribution
Human internal
dose
Study dose groups
LOAEL/
NOAEL
Human oral exposure
(mg/kg/d)
=HED
Lower 99 th
percentile
Figure 5.1.3
Schematic of combined inter-species, intra-species, and route-to-route extrapolation from a
rodent study LOAEL or NOAEL. In the case where BMD modeling is performed, the applied
dose values are replaced by the corresponding median internal dose estimate, and the iPOD is the
modeled BMDL in internal dose units.
5.1.3.3 Results and discussion of p-RfCs andp-RfDs for candidate critical effects
Tables 5.1.8-5.1.13 present the p-cRfCs and p-cRfDs developed using the PBPK internal
dose metrics, along with the cRfCs and cRfDs based on applied dose for comparison, for each
health effect domain.
The greatest impact of using the PBPK model was, as expected, for kidney effects, since
as discussed in Sections 3.3 and 3.5, toxicokinetic data indicate substantially more GSH
conjugation of TCE and subsequent bioactivation of GSH-conjugatates in humans relative to rats
or mice. In addition, as discussed in Sections 3.3 and 3.5, the available in vivo data indicate high
inter-individual variability in the amount of TCE conjugated with GSH. The overall impact is
that the p-cRfCs and p-cRfDs based on the preferred dose metric of bioactivated DCVC are 300-
to 400-fold lower than the corresponding cRfCs and cRfDs based on applied dose. As shown in
Figure 3.5.6 in Section 3.5, for this dose metric there is about a 30- to 100-fold difference
(depending on exposure route and level) between rats and humans in the "central estimates" of
interspecies differences for the fraction of TCE that is bioactivated as DCVC. The uncertainty in
the human central estimate is only on the order of 2-fold (in either direction), while that in the rat
6/22/2009
1076
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
central estimate is substantially greater, about 10-fold (in either direction). In addition, the inter-
individual variability about the human median estimate is on the order of 10-fold (in either
direction). Because of the high confidence in the PBPK model, as well as the high confidence in
GSH conjugation and subsequent bioactivation being the appropriate dose metric for TCE kidney
effects, there is also high confidence in the p-cRfCs and p-RfDs for these effects.
In addition, in two cases in which BMD modeling was employed, using internal dose
metrics led to a sufficiently different dose-response shape so as to change the resulting references
value by greater than 5-fold. For the Woolhiser et al. (2006) decreased PFC response, this
occurred with the AUC of TCE in blood dose metric, leading to a p-cRfC 17-fold higher than the
cRfC based on applied dose. However, the model fit for this effect using this metric was
substantially worse than the fit using the preferred metric of Total oxidative metabolism.
Moreover, whereas an adequate fit was obtained with applied dose only with the highest dose
group dropped, all the dose groups were included when the total oxidative metabolism dose
metric was used while still resulting in a good model fit. Therefore, it appears that using this
metric resolves some of the low-dose supralinearity in the dose-response curve. Nonetheless, the
overall impact of the preferred metric was minimal, as the p-cRfC based on the Total oxidative
metabolism metric was less than 1.4-fold larger than the cRfC based on applied dose. The
second case in which BMD modeling based on internal doses changed the candidate reference
value by more than 5-fold was for resorptions reported by Narotsky et al. (1995). Here, the p-
cRfDs were 7- to 8-fold larger than the corresponding cRfD based on applied dose. However,
for applied dose there is substantial uncertainty in the low-dose curvature of the dose-response
curve. This uncertainty persisted with the use of internal dose metrics, so the BMD remains
somewhat uncertain (see Figures in Appendix F).
In the remaining cases, which generally involved the "generic" dose metrics of total
metabolism and AUC of TCE in blood, the p-cRfCs and p-cRfDs were within 5-fold of the
corresponding cRfC or cRfD based on applied dose, with the vast majority within 3-fold. This
suggests that the standard UFs for inter- and intra-species pharmacokinetic variability are fairly
accurate in capturing these differences for these TCE studies.
6/22/2009
1077
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.8. cRfCs and cRfDs (based on applied dose) and p-cRfCs and p-cRfDs (based on PBPK modeled internal dose metrics) for
candidate critical neurological effects.
Effect type Species POD POD, UFSC UFis UFh UF|0aei UFdb UFcompb cRfC or cRfD or Candidate critical effect; comments [dose metric]
Candidate critical type HECgg, p-cRfC p-cRfD
studies or (ppm) (mg/kg/d)
HED99a
Trigeminal Nerve Effects
Ruitjen et al. 1991 human
Cognitive Effects
Isaacson et al. 1990 rat
Mood and Sleep
Disorders
Arito et al. 1994 rat
Other neurological
effects
Kjellstrand et al. 1987 rat
Gash et al. 2007
rat
LOAEL
14
1
1
10
3
1 30
0.47
Trigeminal nerve effects
hec99
5.3
1
1
3
3
1 10
0.53
[TotMetabBW34]
hec99
8.3
1
1
3
3
1 10
0.83
[AUCCBId]
hed99
7.3
1
1
3
3
1 10
0.73
[TotMetabBW34] (route-to-route)
hed99
14
1
1
3
3
1 10
1.4
[AUCCBId] (route-to-route)
LOAEL
47
10
10
10
10
1 10000c
0.0047
demyelination in hippocampus
hed99
9.2
10
3
3
10
1 1000
0.0092
[TotMetabBW34]
hed99
4.3
10
3
3
10
1 1000
0.0043
[AUCCBId]
hec99
7.1
10
3
3
10
1 1000
0.0071
[TotMetabBW34] (route-to-route)
hec99
2.3
10
3
3
10
1 1000
0.0023
[AUCCBId] (route-to-route)
LOAEL
12
3
3
10
10
1 1000
0.012
Changes in wakefulness
hec99
4.8
3
3
3
10
1 300
0.016
[TotMetabBW34]
hec99
9.0
3
3
3
10
1 300
0.030
[AUCCBId]
hed99
6.5
3
3
3
10
1 300
0.022
[TotMetabBW34] (route-to-route)
hed99
15
3
3
3
10
1 300
0.051
[AUCCBId] (route-to-route)
LOAEL
300
10
3
10
10
1 3000
0.10
I regeneration of sciatic nerve
hec99
93
10
3
3
10
1 1000
0.093
[TotMetabBW34]
hec99
257
10
3
3
10
1 1000
0.26
[AUCCBId]
hed99
97
10
3
3
10
1 1000
0.097
[TotMetabBW34] (route-to-route)
hed99
142
10
3
3
10
1 1000
0.14
[AUCCBId] (route-to-route)
LOAEL
150
10
3
10
10
1 3000
0.050
I regeneration of sciatic nerve
hec99
120
10
3
3
10
1 1000
0.12
[TotMetabBW34]
hec99
108
10
3
3
10
1 1000
0.11
[AUCCBId]
hed99
120
10
3
3
10
1 1000
0.12
[TotMetabBW34] (route-to-route)
hed99
76
10
3
3
10
1 1000
0.076
[AUCCBId] (route-to-route)
LOAEL
710
10
10
10
10
1 10000c
0.071
degeneration of dopaminergic neurons
hed99
53
10
3
3
10
1 1000
0.053
[TotMetabBW34]
hed99
192
10
3
3
10
1 1000
0.19
[AUCCBId]
hec99
47
10
3
3
10
1 1000
0.047
[TotMetabBW34] (route-to-route)
hec99
363
10
3
3
10
1 1000
0.36
[AUCCBId] (route-to-route)
6/22/2009
1078
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
a Applied dose POD adjusted to continuous exposure unless otherwise noted. POD, HEC99, and HED99 have same units as cRfC (ppm) or cRfD (mg/kg/d).
b Product of individual uncertainty factors, rounded to 3, 10, 30, 100, 300, 1000, 3000, or 10,000 [see fotenote (c) below],
0 U.S. EPA's report on the RfC and RfD processes (U.S. EPA, 2002) recommends not deriving reference values with a composite UF of greater than 3000; however, composite UFs exceeding 3000
considered here because the derivation of the cRfCs and cRfDs is part of a screening process and the application of the PBPK model for candidate critical effects reduces the values of some of the
individual UFs for the p-cRfCs and p-cRfDs.
UFSC = subchronic-to-chronic UF; UFis = interspecies UF; UFh = human variability UF; UF|0aei = LOAEL-to-NOAEL UF; UFdb = database UF
Shaded rows represent the p-cRfC or p-cRfD using the preferred PBPK model dose metric.
6/22/2009
1079
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.9. cRfCs and cRfDs (based on applied dose) and p-cRfCs and p-cRfDs (based on PBPK modeled internal dose metrics) for
candidate critical kidney effects.
Effect type Species POD POD, UFSC UFis UFh UF|0aei UFdb UFcompb cRfC or cRfD or Candidate critical effect; comments [dose metric]
Candidate critical type HECgg, p-cRfC p-cRfD
studies or (ppm) (mg/kg/d)
HED99a
Histological changes in
kidney
Maltoni 1986 rat
NCI 1976
NTP 1988
rat
t kidnev/bodv weight
ratio
Kjellstrand et al. 1983b
Woolhiser et al. 2006
rat
BMDL
40.2 1
3
10
1
30
1.3
meganucleocytosis; BMR=10%
HEC99
0.038
3
3
1
10
0.0038
[ABioactDCVCBW34]
hec99
0.058 1
3
3
1
10
0.0058
[AMetGSHBW34]
hec99
15.3
3
3
1
10
1.5
[TotMetabBW34]
hed99
0.023
3
3
1
10
0.0023
[ABioactDCVCBW34] (route-to-route)
hed99
0.036 1
3
3
1
10
0.0036
[AMetGSHBW34] (route-to-route)
hed99
19
3
3
1
10
1.9
[TotMetabBW34] (route-to-route)
LOAEL
620
10
10
30
3000
0.21
toxic nephrosis
hed99
0.30
3
3
30
300
0.00101
[AMetGSHBW34]
hed99
48
3
3
30
300
0.160
[TotMetabBW34]
hec99
0.50
3
3
30
300
0.00165
[AMetGSHBW34] (route-to-route)
hec99
42
3
3
30
300
0.140
[TotMetabBW34] (route-to-route)
BMDL
9.45
10
10
1
100
0.0945
toxic nephropathy; BMR = 5%; female Marshall (most sensitive
sex/strain)
hed99
0.0034
3
3
1
10
0.00034
[ABioactDCVCBW34]
hed99
0.0053
3
3
1
10
0.00053
[AMetGSHBW34]
hed99
0.74
3
3
1
10
0.074
[TotMetabBW34]
hec99
0.0056
3
3
1
10
0.00056
[ABioactDCVCBW34] (route-to-route)
hec99
0.0087 1
3
3
1
10
0.00087
[AMetGSHBW34] (route-to-route)
hec99
0.51
3
3
1
10
0.051
[TotMetabBW34] (route-to-route)
BMDL
34.7 1
3
10
1
30
1.2
BMR=10%
HEC99
0.12 1
3
3
1
10
0.012
[AMetGSHBW34]
hec99
21
3
3
1
10
2.1
[TotMetabBW34]
hed99
0.070
3
3
1
10
0.0070
[AMetGSHBW34] (route-to-route)
hed99
25 1
3
3
1
10
2.5
[TotMetabBW34] (route-to-route)
BMDL
15.7 1
3
10
1
30
0.52
BMR=10%
HEC99
0.013
3
3
1
10
0.0013
[ABioactDCVCBW34]
hec99
0.022 1
3
3
1
10
0.0022
[AMetGSHBW34]
hec99
11
3
3
1
10
1.1
[TotMetabBW34]
hed99
0.0079 1
3
3
1
10
0.00079
[ABioactDCVCBW34] (route-to-route)
hed99
0.013 1
3
3
1
10
0.0013
[AMetGSHBW34] (route-to-route)
hed99
14 1
3
3
1
10
1.4
[TotMetabBW34] (route-to-route)
6/22/2009
1080
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
a Applied dose POD adjusted to continuous exposure unless otherwise noted. POD, HEC99, and HED99 have same units as cRfC or cRfD.
b Product of individual uncertainty factors, rounded to 3, 10, 30, 100, 300, 1000, or 3000.
UFSC = subchronic-to-chronic UF; UFis = interspecies UF; UFh = human variability UF; UF|0aei = LOAEL-to-NOAEL UF; UFdb = database UF
Shaded rows represent the p-cRfC or p-cRfD using the preferred PBPK model dose metric
6/22/2009
1081
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.10. cRfCs and cRfDs (based on applied dose) and p-cRfCs and p-cRfDs (based on PBPK modeled internal dose metrics) for
candidate critical liver effects.
Effect type
Candidate critical
studies
Species
POD
type
POD,
HEC99,
or
HED99a
UFS(
UFis UFh
UF|C
UFd
UFC(
cRfC or
p-cRfC
(ppm)
cRfD or
p-cRfD
(mg/kg/d)
Candidate critical effect; comments [dose metric]
t liver/body weight ratio
Kjellstrand et al. 1983b
Woolhiser et al. 2006
rat
Buben & O'Flaherty 1985 mouse
BMDL
21.6
3
10
1 1
30
0.72
BMR=10% increase
hec99
9.1
3
3
1 1
10
0.91
[AMetLivl BW34]
hec99
24.9
3
3
1 1
10
2.5
[TotOxMetabBW34]
hed99
7.9
3
3
1 1
10
0.79
[AMetLivl BW34] (route-to-route)
hed99
25.7
3
3
13 1
10
2.6
[TotOxMetabBW34] (route-to-route)
BMDL
25
3
10
1 1
30
0.83
BMR=10% increase
hec99
19
3
3
1 1
10
1.9
[AMetLivl BW34]
hec99
16
3
3
1 1
10
1.6
[TotOxMetabBW34]
hed99
16
3
3
1 1
10
1.6
[AMetLivl BW34] (route-to-route)
hed99
17
3
3
1 1
10
1.7
[TotOxMetabBW34] (route-to-route)
BMDL
82
10
10
1 1
100
0.82
BMR=10% increase
hed99
10
3
3
1 1
10
1.0
[AMetLivl BW34]
hed99
13
3
3
1 1
10
1.3
[TotOxMetabBW34]
hec99
11
3
3
1 1
10
1.1
[AMetLivl BW34] (route-to-route)
hec99
11
3
3
1 1
10
1.1
[TotOxMetabBW34] (route-to-route)
a Applied dose POD adjusted to continuous exposure unless otherwise noted. POD, HEC99, and HED99 have same units as cRfC (ppm) or cRfD (mg/kg/d).
b Product of individual uncertainty factors, rounded to 3, 10, 30, 100, 300, 1000, or 3000.
UFsc = subchronic-to-chronic UF; UFis = interspecies UF; UFh = human variability UF; UF|0aei = LOAEL-to-NOAEL UF; UFdb = database UF
Shaded rows represent the p-cRfC or p-cRfD using the preferred PBPK model dose metric
6/22/2009
1082
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.11. cRfCs and cRfDs (based on applied dose) and p-cRfCs and p-cRfDs (based on PBPK modeled internal dose metrics) for
candidate critical immunological effects.
Effect tvoe
Candidate critical
studies
Species
POD
type
POD,
hec99,
or
HED99a
UFSC
UFis
UFh
UF|oael
UFdb UFcompb
cRfC or
p-cRfC
(ppm)
cRfD or
p-cRfD
(mg/kg/d)
Candidate critical effect; comments [dose metric]
1 thvmus weiaht
Keil et al. 2009
mouse
LOAEL
0.35
1
10
10
10
1 1000
0.00035
I thymus weight
hed99
0.048
1
3
3
10
1 100
0.00048
[TotMetabBW34]
hed99
0.016
1
3
3
10
1 100
0.00016
[AUCCBId]
hec99
0.033
1
3
3
10
1 100
0.00033
[TotMetabBW34] (route-to-route)
Autoimmunity
Kaneko et al. 2000
mouse
hec99
LOAEL
0.0082
70
1
10
3
3
3
3
10
10
1 100
1 1000
0.000082
0.070
[AUCCBId] (route-to-route)
Changes in immunoreactive organs - liver (including sporadic
necrosis in hepatic lobules), spleen; UFh=3 due to
autoimmune-prone mouse
hec99
37
10
3
1
10
1 300
0.12
[TotMetabBW34]
hec99
69
10
3
1
10
1 300
0.23
[AUCCBId]
hed99
42
10
3
1
10
1 300
0.14
[TotMetabBW34] (route-to-route)
Keil et al. 2009
mouse
hed99
LOAEL
57
0.35
10
1
3
10
1
10
10
1
1 300
1 100
0.19
0.0035
[AUCCBId] (route-to-route)
t anti-dsDNA & anti-ssDNA Abs (early markers for SLE)
hed99
0.048
1
3
3
1
1 10
0.0048
[TotMetabBW34]
hed99
0.016
1
3
3
1
1 10
0.0016
[AUCCBId]
hec99
0.033
1
3
3
1
1 10
0.0033
[TotMetabBW34] (route-to-route)
Immunosuppression
Woolhiser et al. 2006
rat
hec99
BMDL
0.0082
24.9
1
10
3
3
3
10
1
1
1 10
1 300
0.00082
0.083
[AUCCBId] (route-to-route)
I PFC response; BMR=1 SD change; dropped highest dose
hec99
11
10
3
3
1
1 100
0.11
[TotMetabBW34]; all dose groups
hec99
140
10
3
3
1
1 100
1.4
[AUCCBId]; all dose groups
hed99
14
10
3
3
1
1 100
0.14
[TotMetabBW34] (route-to-route); all dose groups
Sanders et al. 1982
mouse
hed99
LOAEL
91
18
10
1
3
10
3
10
1
3
1 100
1 300
0.91
0.060
[AUCCBId] (route-to-route); all dose groups
I stem cell bone marrow recolonization (sustained); J, cell-
mediated response to sRBC (largely transient during
exposure); females more sensitive
hed99
2.5
1
3
3
3
1 30
0.083
[TotMetabBW34]
hed99
0.84
1
3
3
3
1 30
0.028
[AUCCBId]
hec99
1.7
1
3
3
3
1 30
0.057
[TotMetabBW34] (route-to-route)
hec99
0.43
1
3
3
3
1 30
0.014
[AUCCBId] (route-to-route)
a Applied dose POD adjusted to continuous exposure unless otherwise noted. POD, HEC99, and HED99 have same units as cRfC (ppm) or cRfD (mg/kg/d).
b Product of individual uncertainty factors, rounded to 3, 10, 30, 100, 300, 1000, or 3000.
UFsc = subchronic-to-chronic UF; UFis = interspecies UF; UFh = human variability UF; UF|0aei = LOAEL-to-NOAEL UF; UFdb = database UF
Shaded rows represent the p-cRfC or p-cRfD using the preferred PBPK model dose metric
6/22/2009 1083
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.12. cRfCs and cRfDs (based on applied dose) and p-cRfCs and p-cRfDs (based on PBPK modeled internal dose metrics) for
candidate critical reproductive effects.
Effect type Species POD POD, UFSC UFis UFh UF|0aei UFdb UFcompb cRfC or cRfD or Candidate critical effect; comments [dose metric]
Candidate critical type HECgg, p-cRfC p-cRfD
studies or (ppm) (mg/kg/d)
HED99a
Effects on sperm, male
reproductive outcomes
Chia et al. 1996
human
BMDL
1.43
10
1
10
1
1
100
0.014
hyperzoospermia; BMR=10% extra risk
hec99
0.50
10
1
3
1
1
30
0.0017
[TotMetabBW34]
hec99
0.83
10
1
3
1
1
30
0.0028
[AUCCBId]
hed99
0.73
10
1
3
1
1
30
0.024
[TotMetabBW34] (route-to-route)
hed99
1.6
10
1
3
1
1
30
0.053
[AUCCBId] (route-to-route)
Xu et al. 2004
mouse
LOAEL
180
10
3
10
10
1
3000
0.060
I fertilization
hec99
67
10
3
3
10
1
1000
0.067
[TotMetabBW34]
hec99
170
10
3
3
10
1
1000
0.17
[AUCCBId]
hed99
73
10
3
3
10
1
1000
0.073
[TotMetabBW34] (route-to-route)
hed99
104
10
3
3
10
1
1000
0.10
[AUCCBId] (route-to-route)
Kumar et al. 2000a 2001 b
rat
LOAEL
45
10
3
10
10
1
3000
0.015
multiple sperm effects, increasing severity from 12 to 24 weeks
hec99
13
10
3
3
10
1
1000
0.013
[TotMetabBW34]
hec99
53
10
3
3
10
1
1000
0.053
[AUCCBId]
hed99
16
10
3
3
10
1
1000
0.016
[TotMetabBW34] (route-to-route)
hed99
49
10
3
3
10
1
1000
0.049
[AUCCBId] (route-to-route)
DuTeaux et al. 2004
rat
LOAEL
141
10
10
10
10
1
10000c
0.014
I ability of sperm to fertilize in vitro
hed99
16
10
3
3
10
1
1000
0.016
[AUCCBId]
hed99
42
10
3
3
10
1
1000
0.042
[TotOxMetabBW34]
hec99
9.3
10
3
3
10
1
1000
0.0093
[AUCCBId] (route-to-route)
hec99
43
10
3
3
10
1
1000
0.043
[TotOxMetabBW34] (route-to-route)
Male reproductive tract
effects
Forkert et al. 2002, Kan et
mouse
LOAEL
180
10
3
10
10
1
3000
0.060
effects on epididymis epithelium
al. 2007
hec99
67
10
3
3
10
1
1000
0.067
[TotMetabBW34]
hec99
170
10
3
3
10
1
1000
0.17
[AUCCBId]
hed99
73
10
3
3
10
1
1000
0.073
[TotMetabBW34] (route-to-route)
hed99
104
10
3
3
10
1
1000
0.10
[AUCCBId] (route-to-route)
Kumar et al. 2000a 2001 b
rat
LOAEL
45
10
3
10
10
1
3000
0.015
testes effects, testicular enzyme markers, increasing severity
from 12 to 24 weeks
hec99
13
10
3
3
10
1
1000
0.013
[TotMetabBW34]
hec99
53
10
3
3
10
1
1000
0.053
[AUCCBId]
6/22/2009
1084
-------
INTER-AGENCY REVIEW DRAFT-
DO NOT QUOTE OR CITE
hed99
16
10
3
3
10 1
1000
0.016
[TotMetabBW34] (route-to-route)
hed99
49
10
3
3
10 1
1000
0.049
[AUCCBId] (route-to-route)
Female reproductive
outcomes
Narotsky et al. 1995
rat
LOAEL
475
1
10
10
10 1
1000
0.48
delayed parturition
hed99
44
1
3
3
10 1
100
0.44
[TotMetabBW34]
hed99
114
1
3
3
10 1
100
1.1
[AUCCBId]
hec99
37
1
3
3
10 1
100
0.37
[TotMetabBW34] (route-to-route)
hec99
190
1
3
3
10 1
100
1.9
[AUCCBId] (route-to-route)
Reproductive behavior
George et al. 1986
rat
LOAEL
389
1
10
10
10 1
1000
0.39
I mating (both sexes exposed)
hed99
77
1
3
3
10 1
100
0.77
[TotMetabBW34]
hed99
52
1
3
3
10 1
100
0.52
[AUCCBId]
hec99
71
1
3
3
10 1
100
0.71
[TotMetabBW34] (route-to-route)
hec99
60
1
3
3
10 1
100
0.60
[AUCCBId] (route-to-route)
a Applied dose POD adjusted to continuous exposure unless otherwise noted. POD, HEC99, and HED99 have same units as cRfC (ppm) or cRfD (mg/kg/d).
b Product of individual uncertainty factors, rounded to 3, 10, 30, 100, 300, 1000, 3000, or 10,000 [see footnote (c) below],
0 U.S. EPA's report on the RfC and RfD processes (U.S. EPA, 2002) recommends not deriving reference values with a composite UF of greater than 3000; however, composite UFs exceeding 3000 are
considered here because the derivation of the cRfCs and cRfDs is part of a screening process and the application of the PBPK model for candidate critical effects reduces the values of some of the
individual UFs for the p-cRfCs and p-cRfDs.
UFsc = subchronic-to-chronic UF; UFis = interspecies UF; UFh = human variability UF; UF|0aei = LOAEL-to-NOAEL UF; UFdb = database UF
Shaded rows represent the p-cRfC or p-cRfD using the preferred PBPK model dose metric.
6/22/2009
1085
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.13. cRfCs and cRfDs (based on applied dose) and p-cRfCs and p-cRfDs (based on PBPK modeled internal dose metrics) for
candidate critical developmental effects.
Effect type
Species
POD
POD,
UFSC
UFis
UFh
UF|oael UF db
UF b
u 1 comp
cRfC or
cRfD or
Candidate critical effect; comments [dose metric]
Candidate critical
type
hec99,
p-cRfC
p-cRfD
studies
or
(ppm)
(mg/kg/d)
HED99a
Pre- and post-natal
mortality
Healy et al. 1982
rat
LOAEL
17
1
3
10
10 1
300
0.057
resorptions
hec99
6.2
1
3
3
10 1
100
0.062
[TotMetabBW34]
hec99
14
1
3
3
10 1
100
0.14
[AUCCBId]
hed99
8.5
1
3
3
10 1
100
0.085
[TotMetabBW34] (route-to-route)
hed99
20
1
3
3
10 1
100
0.20
[AUCCBId] (route-to-route)
Narotsky et al. 1995
rat
BMDL
32.2
1
10
10
1 1
100
0.32
resorptions; BMR=1% extra risk
hed99
28
1
3
3
1 1
10
2.8
[TotMetabBW34]
hed99
29
1
3
3
1 1
10
2.9
[AUCCBId]
hec99
23
1
3
3
1 1
10
2.3
[TotMetabBW34] (route-to-route)
hec99
24
1
3
3
1 1
10
2.4
[AUCCBId] (route-to-route)
Pre- and post-natal
growth
Healy et al. 1982
rat
LOAEL
17
1
3
10
10 1
300
0.057
I fetal weight; skeletal effects
hec99
6.2
1
3
3
10 1
100
0.062
[TotMetabBW34]
hec99
14
1
3
3
10 1
100
0.14
[AUCCBId]
hed99
8.5
1
3
3
10 1
100
0.085
[TotMetabBW34] (route-to-route)
hed99
20
1
3
3
10 1
100
0.20
[AUCCBId] (route-to-route)
Conaenital defects
Johnson et al. 2003
rat
BMDL
0.0207
1
10
10
1 1
100
0.00021
heart malformations (pups); BMR=1% extra risk; highest
dose group (1000-fold higher than next highest) dropped to
improve model fit
hed99
0.0052
1
3
3
1 1
10
0.00052
[TotOxMetabBW34]
hed99
0.0017
1
3
3
1 1
10
0.00017
[AUCCBId]
hec99
0.0037
1
3
3
1 1
10
0.00037
[TotOxMetabBW34] (route-to-route)
hec99
0.00093
1
3
3
1 1
10
0.000093
[AUCCBId] (route-to-route)
Developmental
neurotoxicity
Fredricksson et al. 1993
mouse
LOAEL
50
3
10
10
10 1
3000
0.017
I rearing post-exp; pup gavage dose
hed99
4.1
3
3
3
10 1
300
0.014
[TotMetabBW34]
hed99
3.5
3
3
3
10 1
300
0.012
[AUCCBId]
hec99
3.0
3
3
3
10 1
300
0.010
[TotMetabBW34] (route-to-route)
hec99
1.8
3
3
3
10 1
300
0.0061
[AUCCBId] (route-to-route)
Taylor et al. 1985
rat
LOAEL
45
1
10
10
10 1
1000
0.045
t exploration post-exp; estimated dam dose
hed99
11
1
3
3
10 1
100
0.11
[TotMetabBW34]
hed99
4.1
1
3
3
10 1
100
0.041
[AUCCBId]
6/22/2009
1086
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
hec99
8.4
1
3
3
10
1
100
0.084
[TotMetabBW34] (route-to-route)
hec99
2.2
1
3
3
10
1
100
0.022
[AUCCBId] (route-to-route)
Isaacson&Taylor 1989 rat
LOAEL
16
1
10
10
10
1
1000
0.016 I myelination in hippocampus; estimated dam dose
Developmental
iminunotoxicitv
Peden-Adams et al. 2006 mouse
LOAEL
0.37
1
10
10
10
1
1000
0.00037 I PFC, tDTH; POD is estimated dam dose (exp thruout gest
and lactation + to 3 or 8 wks of age)
a Applied dose POD adjusted to continuous exposure unless otherwise noted. POD, HEC99, and HED99 have same units as cRfC (ppm) or cRfD (mg/kg/d).
b Product of individual uncertainty factors, rounded to 3, 10, 30, 100, 300, 1000, or 3000.
UFsc = subchronic-to-chronic UF; UFis = interspecies UF; UFh = human variability UF; UF|0aei = LOAEL-to-NOAEL UF; UFdb = database UF
Shaded rows represent the p-cRfC or p-cRfD using the preferred PBPK model dose metric or, in the cases where the PBPK model was not used, the cRfD or cRfC based on applied dose.
6/22/2009
1087
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.4 Uncertainties in cRfCs and cRfDs
5.1.4.1 Qualitative uncertainties
An underlying assumption in deriving reference values for noncancer effects is that the
dose-response relationship for these effects has a threshold. Thus, a fundamental uncertainty is
the validity of that assumption. For some effects, in particular effects on very sensitive processes
(e.g., developmental processes) or effects for which there is a nontrivial background level and
even small exposures may contribute to background disease processes in more susceptible
people, a practical threshold (i.e., a threshold within the range of environmental exposure levels
of regulatory concern) may not exist.
Nonetheless, under the assumption of a threshold, the desired exposure level to have as a
reference value is the maximum level at which there is no appreciable risk for an adverse effect
in (non-negligible) sensitive subgroups (of humans). However, because it is not possible to
know what this level is, "uncertainty factors" are used to attempt to address quantitatively
various aspects, depending on the dataset, of qualitative uncertainty.
First there is uncertainty about the "point of departure" for the application of UFs.
Conceptually, the POD should represent the maximum exposure level at which there is no
appreciable risk for an adverse effect in the study population under study conditions (i.e., the
threshold in the dose-response relationship). Then, the application of the relevant UFs is
intended to convey that exposure level to the corresponding exposure level for sensitive human
subgroups exposed continuously for a lifetime. In fact, it is again not possible to know that
exposure level even for a laboratory study because of experimental limitations (e.g. the power to
detect an effect, dose spacing, measurement errors, etc.), and crude approximations like the
NOAEL or a BMDL are used. If a LOAEL is used as the POD, the LOAEL-to-NOAEL UF is
applied as an adjustment factor to get a better approximation of the desired exposure level
(threshold), but the necessary extent of adjustment is unknown.
If a BMDL is used as the POD, there are uncertainties regarding the appropriate dose-
response model to apply to the data, but these should be minimal if the modeling is in the
observable range of the data. There are also uncertainties about what BMR to use to best
approximate the desired exposure level (threshold, see above). For continuous endpoints, in
particular, it is often difficult to identify the level of change that constitutes the "cut-point" for an
adverse effect. Sometimes, to better approximate the desired exposure level, a BMR somewhat
below the observable range of the data is selected. In such cases, the model uncertainty is
increased, but this is a trade-off to reduce the uncertainty about the POD not being a good
approximation for the desired exposure level.
6/22/2009
1088
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
For each of these types of PODs, there are additional uncertainties pertaining to
adjustments to the administered exposures (doses). Typically, administered exposures (doses)
are converted to equivalent continuous exposures (daily doses) over the study exposure period
under the assumption that the effects are related to concentration x time, independent of the daily
(or weekly) exposure regimen (i.e., a daily exposure of 6 hours to 4 ppm is considered equivalent
to 24 hours of exposure to 1 ppm). However, the validity of this assumption is generally
unknown, and, if there are dose-rate effects, the assumption of C x t equivalence would tend to
bias the POD downwards. Where there is evidence that administered exposure better correlates
to the effect than equivalent continuous exposure averaged over the study exposure period (e.g.,
visual effects), administered exposure was not adjusted. For the PBPK analyses in this
assessment, the actual administered exposures are taken into account in the PBPK modeling, and
equivalent daily values (averaged over the study exposure period) for the dose metrics are
obtained (see above, 5.1.3.2). Additional uncertainties about the PBPK-based estimates include
uncertainties about the appropriate dose metric for each effect, although for some effects there
was better information about relevant dose metrics than for others (see Section 5.1.3.1).
Second, there is uncertainty about the UFs. The human variability UF is to some extent
an adjustment factor because for more sensitive people, the dose-response relationship shifts to
lower exposures. But there is uncertainty about the extent of the adjustment required, i.e., about
the distribution of human susceptibility. Therefore, in the absence of data on a more sensitive
population(s) or on the distribution of susceptibility in the general population, an UF of 10 is
generally used, in part for pharmacokinetic variability and in part for pharmacodynamic
variability. The PBPK analyses in this assessment attempt to account for the pharmacokinetic
portion of human variability using human data on pharmacokinetic variability. A quantitative
uncertainty analysis of the PBPK-derived dose metrics used in the assessment is presented in
Section 5.1.4.2 below. There is still uncertainty regarding the susceptible subgroups for TCE
exposure and the extent of pharmacodynamic variability.
If the data used to determine a particular POD are from laboratory animals, an
interspecies extrapolation UF is used. This UF is also to some extent an adjustment factor for the
expected scaling for toxicologically-equivalent doses across species (i.e., according to body
weight to the 3/4 power for oral exposure). However, there is also uncertainty about the true
extent of interspecies differences for specific noncancer effects from specific chemical
exposures. Often, the "adjustment" component of this UF has been attributed to
pharmacokinetics, while the "uncertainty" component has been attributed to pharmacodynamics,
but as discussed above in Section 5.1.3.1, this is not the only interpretation supported. For oral
exposures, the standard value for the interspecies UF is 10, which can be viewed as breaking
down (approximately) to a factor of 3 for the "adjustment" (nominally pharmacokinetics) and a
6/22/2009
1089
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
factor of 3 for the "uncertainty" (nominally pharmacodynamics). For inhalation exposures, no
adjustment across species is generally assumed for fixed air concentrations (ppm equivalence),
and the standard value for the interspecies UF is 3 reflects "uncertainty" (nominally
pharmacodynamics only). The PBPK analyses in this assessment attempt to account for the
"adjustment" portion of interspecies extrapolation using rodent pharmacokinetic data to estimate
internal doses for various dose metrics. With respect to the "uncertainty" component,
quantitative uncertainty analyses of the PBPK-derived dose metrics used in the assessment are
presented in Section 5.1.4.2 below. However, these only address the pharmacokinetic
uncertainties in a particular dose metric, and there is still uncertainty regarding the true dose
metrics. Nor do the PBPK analyses address the uncertainty in either cross-species
pharmacodynamic differences (i.e., about the assumption that equal doses of the appropriate dose
metric convey equivalent risk across species for a particular endpoint from a specific chemical
exposure) or in cross-species pharmacokinetic differences not accounted for by the PBPK model
dose metrics (e.g., departures from the assumed inter-species scaling of clearance of the active
moiety, in the cases where only its production is estimated). A value of 3 is typically used for
the "uncertainty" about cross-species differences, and this generally represents true uncertainty
because it is usually unknown, even after adjustments have been made to account for the
expected interspecies differences, whether humans have more or less susceptibility, and to what
degree, than the laboratory species in question.
If only subchronic data are available, the sub chronic-to-chronic UF is to some extent an
adjustment factor because, if the effect becomes more severe with increasing exposure, then
chronic exposure would shift the dose-response relationship to lower exposures. However, the
true extent of the shift is unknown.
Sometimes a database UF is also applied to address limitations or uncertainties in the
database. The overall database for TCE is quite extensive, with studies for many different types
of effects, including 2-generation reproductive studies, as well as neurological, immunological,
and developmental immunological studies. In addition, there were sufficient data to develop a
reliable PBPK model to estimate route-to-route extrapolated doses for some candidate critical
effects for which data were only available for one route of exposure. Thus, there is a high degree
of confidence that the TCE database was sufficient to identify some sensitive endpoints.
5.1.4.2 Quantitative uncertainty analysis of PBPK model-based dose metrics for LOAEL or
NOAEL-based PODs
The Bayesian analysis of the PBPK model for TCE generates distributions of uncertainty
and variability in the internal dose metrics that can be readily used for characterizing the
6/22/2009
1090
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
uncertainty and variability in the PBPK model-based derivations of the HEC and HED. As
shown in Figure 5.1.4, the overall approach taken for the uncertainty analysis is similar to that
used for the point estimates except for the carrying through of distributions rather than median or
expected values at various points. Because of a lack of tested software and limitations of time
and resources, this analysis was not performed for iPODs based on BMD modeling, and was
only performed for iPODs derived from a LOAEL or NOAEL. However, for those endpoints for
which BMD modeling was performed, for the purposes of this uncertainty analysis, an
alternative iPOD was used based on the study LOAEL or NOAEL.
In brief, the distribution of rodent PBPK model parameters is carried through to a
distribution of iPODs, reflecting combined uncertainty and variability in the rodent internal
dosimetry. Separately, for each set of human population parameters, a set of individual PBPK
model parameters is generated, and the human PBPK model is run for a range of continuous
exposures from 10"1 to 2xl03 ppm or mg/kg/d to obtain the distribution of the relationship
between human exposure and internal dose. For a given set of (i) an iPOD sampled from the
rodent distribution, (ii) a human population sampled from the distribution of populations, and
(iii) an individual sampled from this population, a human equivalent exposure (HEC or HED)
corresponding to the iPOD is derived by interpolation. Within each population, a HEC or HED
corresponding to the median and 99th percentile individuals are derived, resulting in two
distributions (both reflecting uncertainty): one of "typical" individuals represented by the
distribution of population medians, and one of "sensitive" individuals represented by the
distribution of an upper percentile of the population (e.g., 99th percentile). Note that because a
distribution of rodent-derived iPODs was used, the uncertainty distribution includes the
contribution from the uncertainty in the rodent internal dose. Thus, for selected quantiles of the
population and level of confidence (e.g., Xth percentile individual at Yth percent confidence),
the interpretation is that at the resulting HEC or HED, there is Y% confidence that X% of the
population has an internal dose less than that of the rodent in the toxicity study.
As shown in Tables 5.1.14-18, the HEC99 and HED99 derived using the rodent median
dose metrics and the combined uncertainty and variability in human dose metrics is generally
near (within 1.3-fold of) the median confidence level estimate of the HEC and HED for the 99th
percentile individual. Therefore, the interpretation is that there is about 50% confidence that
human exposure at the HEC99 or HED99 will, in 99% of the human population, lead to an internal
dose less than or equal to that in the subjects (rodent or human) exposed at the POD in the
corresponding study.
In several cases, the uncertainty, as reflected in the ratio between the 95% and 50%
confidence bounds on the 99th percentile individual, was rather high (e.g., > 5-fold), and reflected
primarily uncertainty in the rodent internal dose estimates, discussed previously in Section 3.5.7.
6/22/2009
1091
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
The largest uncertainties (ratios between 95% to 50% confidence bounds of 8- to 10-fold) were
for kidney effects in mice using the AMetGSHBW34 dose metric (Kjellstrand et al., 1983; NCI,
1976). More moderate uncertainties (ratios between 95% to 50% confidence bounds of 5- to 8-
fold) were evident in some oral studies using the AUCCBld dose metric (Sanders et al., 1982;
George al. 1986; Fredricksson et al., 1993; Keil et al., 2009), as well as in studies reporting
kidney effects in rats in which the ABioactDCVCBW34 or AMetGSHBW34 dose metrics were
used (Woolhiser et al., 2006; NTP, 1988). Therefore, in these cases, a POD that is protective of
the 99th percentile individual at a confidence level higher than 50% could be as much as an order
of magnitude lower.
For comparison, Tables 5.1.14 and 5.1.18 also show the ratios of the overall 50th
percentile to the overall 99th percentile HECs and HEDs, reflecting combined human uncertainty
and variability at the median study/endpoint iPOD. The smallest ratios (up to 1.2-fold) are for
total, oxidative, and hepatic oxidative metabolism dose metrics from oral exposures, due to the
large hepatic first-pass effect resulting in virtually all of the oral intake being metabolized before
systemic circulation. Conversely, the large hepatic first-pass results in high variability in the
blood concentration of TCE following oral exposures, with ratios up to 12-fold at low exposures
(e.g., 90% versus 99% first-pass would result in amounts metabolized differing by about 10%
but TCE blood concentrations differing by about 10-fold). From inhalation exposures, there is
moderate variability in these metrics, about 2- to 3-fold. For GSH conjugation and bioactivated
DCVC, however, variability is high (8- to 10-fold) for both exposure routes, which follows from
the incorporation in the PBPK model analysis of the data from Lash et al. (1999b) showing
substantial inter-individual variability in GSH conjugation in humans.
Finally, it is important to emphasize that this analysis only addresses pharmacokinetic
uncertainty and variability, so other aspects of extrapolation addressed in the uncertainty factors
(e.g., LOAEL to NOAEL, subchronic to chronic, pharmacodynamic differences), discussed
above, are not included in the level of confidence.
6/22/2009
1092
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1
2
3
4
5
6
Rodent
non-cancer
study
experimental
paradigm
[distribution (combined
ncertainty and variability)
Rodent
model
parameters
^distribution
PBPK
model
Rodent
internal
dose
jpistribution
Model
OAEuNOAE
¦distribution
iPOD=BM
LOAEL or
NOAEL
(internal
dose unit
0.1-2000 ppm
in air or
0.1-2000
mg/kg-d
continuous
exposure
^distribution
Rodent
non-cancer
study
responses
Human
model
parameters
PBPK
model
Human
internal
dose
uman
dose or
concentration
at internal
OD
distribution (separate
iLincertainty and variability)
Uncertainty distribution
of population median
istribution of functions
f dose or concentration
Uncertainty distribution
of population 95%-ile
invert functions of dose
or concentration
Typical
human
equivalent
Sensitive
Human
equivalent
distribution
distribution
Figure 5.1.4
Flow-chart for uncertainty analysis of HECs and HEDs derived using PBPK model-based dose
metrics. Square nodes indicate point values, circle nodes indicate distributions, and the inverted
triangle indicates a (deterministic) functional relationship.
6/22/2009
1093
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.14. Comparison of "sensitive individual" HECs or HEDs for Neurological effects
based on PBPK modeled internal dose metrics at different levels of confidence and
sensitivity, at the NOAEL or LOAEL.
Candidate critical effect POD
Candidate critical study type
Ratio
HEC/D50:
HEC/D99
HECxor
HEDX
[Dose metric]
X=99
X=99,
median
X=99,
95lcb
Neurological
Trigeminal nerve effects
Ruitjen etal. 1991 (human)
HEC
2.62
5.4
5.4
2.6
[TotMetabBW34]
HEC
1.68
8.3
8.3
4.9
[AUCCBId]
HED
1.02
7.3
7.2
3.8
[TotMetabBW34] (rtr
HED
4.31
14
16
8.0
[AUCCBId] (rtr)
demyelination in hippocampus
Isaacson et al. 1990 (rat)
HED
1.02
9.21
9.20
7.39
[TotMetabBW34]
HED
7.20
4.29
5.28
2.52
[AUCCBId]
HEC
2.59
7.09
6.77
4.94
[TotMetabBW34] (rtr)
HEC
1.68
2.29
2.42
0.606
[AUCCBId] (rtr)
Changes in wakefulness
Arito et al. 1994 (rat)
HEC
2.65
4.79
4.86
2.37
[TotMetabBW34]
HEC
1.67
9
9.10
4.63
[AUCCBId]
HED
1.02
6.46
6.50
3.39
[TotMetabBW34] (rtr)
HED
4.25
15.2
18.0
8.33
[AUCCBId] (rtr)
I regeneration of sciatic nerve
Kjellstrand et al. 1987 (rat)
HEC
2.94
93.1
93.6
38.6
[TotMetabBW34]
HEC
1.90
257
266
114
[AUCCBId]
HED
1.13
97.1
96.8
43.4
[TotMetabBW34] (rtr)
HED
3.08
142
147
78.0
[AUCCBId] (rtr)
I regeneration of sciatic nerve
Kjellstrand et al. 1987 (mouse)
HEC
3.16
120
125
48.8
[TotMetabBW34]
HEC
1.84
108
111
59.7
[AUCCBId]
HED
1.21
120
121
57.0
[TotMetabBW34] (rtr)
HED
2.13
75.8
79.1
53.4
[AUCCBId] (rtr)
degeneration of dopaminergic
neurons
Gash et al. 2007 (rat)
HED
1.06
53
53.8
17.1
[TotMetabBW34]
HED
2.98
192
199
94.7
[AUCCBId]
HEC
2.70
46.8
47.9
14.2
[TotMetabBW34] (rtr)
HEC
1.87
363
380
144
[AUCCBId] (rtr)
HEC99 = the 99 percentile of the combined human uncertainty and variability distribution of continuous exposure
concentrations that lead to the (fixed) median estimate of the rodent internal dose at the POD.
HEC99,median (or HECg9,95icb) = the median (or 95th percentile lower confidence bound) estimate of the uncertainty
distribution of continuous exposure concentrations for which the 99th percentile individual has an internal dose less
than the (uncertain) rodent internal dose at the POD.
rtr = Route-to-route extrapolation using PBPK model and the specified dose metric
Shaded rows denote results for the primary dose metric
6/22/2009
1094
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 Table 5.1.15. Comparison of "sensitive individual" HECs or HEDs for Kidney and Liver
2 effects based on PBPK modeled internal dose metrics at different levels of confidence and
3 sensitivity, at the NOAEL or LOAEL.
Candidate critical effect POD Ratio HECX or [Dose metric]
Candidate critical study type HEC/D50: HEDX
(species) HEC/D99
X=99
X=99,
median
X=99,
95lcb
Kidney
Meganucleocytosis [NOAEL]3
Maltoni 1986 (rat)
HEC
7.53
0.0233
0.0260
0.00366
[ABioactDCVC BW34]
HEC
7.70
0.0364
0.0411
0.00992
[AMetGSHBW34]
HEC
2.57
8.31
7.97
4.03
[TotMetabBW34]
HED
9.86
0.0140
0.0156
0.00216
[ABioactDCVCBW34] (rtr)
HED
9.83
0.0223
0.0242
0.00597
[AMetGSHBW34] (rtr)
HED
1.02
10.6
10.7
5.75
[TotMetabBW34] (rtr)
toxic nephrosis
NCI 1976 (mouse)
HED
9.51
0.30
0.32
0.044
[AMetGSHBW34]
HED
1.05
48
48.9
16.2
[TotMetabBW34]
HEC
7.78
0.50
0.514
0.0703
[AMetGSHBW34] (rtr)
HEC
2.67
42
43.5
13.7
[TotMetabBW34] (rtr)
toxic nephropathy [LOAEL]3
NTP 1988 (rat)
HED
9.75
0.121
0.126
0.0177
[ABioactDCVC BW34]
HED
9.64
0.193
0.210
0.0379
[AMetGSHBW34]
HED
1.03
33.1
33.1
11.1
[TotMetabBW34]
HEC
7.55
0.201
0.204
0.0269
[ABioactDCVCBW34] (rtr)
HEC
7.75
0.314
0.353
0.0676
[AMetGSHBW34] (rtr)
HEC
2.59
28.2
27.2
8.77
[TotMetabBW34] (rtr)
t kidney/body weight ratio
[NOAEL]3
Kjellstrand et al. 1983b (mouse)
HEC
7.69
0.111
0.103
0.00809
[AMetGSHBW34]
HEC
2.63
34.5
33.7
13.5
[TotMetabBW34]
HED
9.78
0.068
0.00641
0.00497
[AMetGSHBW34] (rtr)
HED
1.03
39.9
39.2
17.9
[TotMetabBW34] (rtr)
t kidney/body weight ratio
[NOAEL]3
Woolhiser et al. 2006 (rat)
HEC
7.53
0.0438
0.0481
0.00737
[ABioactDCVC BW34]
HEC
7.70
0.0724
0.0827
0.0179
[AMetGSHBW34]
HEC
2.54
16.1
15.2
7.56
[TotMetabBW34]
HED
9.84
0.0264
0.0282
0.00447
[ABioactDCVCBW34] (rtr)
HED
9.81
0.0444
0.0488
0.0111
[AMetGSHBW34] (rtr)
HED
1.02
19.5
19.2
10.5
[TotMetabBW34] (rtr)
Liver
t liver/body weight ratio [LOAEL]3
Kjellstrand et al. 1983b (mouse)
HEC
2.85
16.2
16.3
6.92
[AMetLivl BW34]
HEC
3.63
40.9
38.1
15.0
[TotOxMetabBW34]
HED
1.16
14.1
14.1
5.85
[AMetLivl BW34] (rtr)
HED
1.53
40.1
39.4
17.9
[TotOxMetabBW34] (rtr)
t liver/body weight ratio [NOAEL]3
Woolhiser et al. 2006 (rat)
HEC
2.86
20.7
21.0
11.0
[AMetLivl BW34]
HEC
2.94
18.2
17.1
8.20
[TotOxMetabBW34]
HED
1.20
17.8
17.7
9.94
[AMetLivl BW34] (rtr)
HED
1.21
19.6
19.3
10.5
[TotOxMetabBW34] (rtr)
t liver/body weight ratio [LOAEL]3
Buben & O'Flaherty 1985
(mouse)
HED
1.14
8.82
8.95
4.17
[AMetLivl BW34]
HED
1.14
9.64
9.78
5.28
[TotOxMetabBW34]
HEC
2.80
10.1
9.97
4.83
[AMetLivl BW34] (rtr)
HEC 3.13 7.83 7.65 4.23 [TotOxMetabBW34] (rtr)
6/22/2009
1095
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 HECgg = the 99th percentile of the combined human uncertainty and variability distribution of continuous exposure
2 concentrations that lead to the (fixed) median estimate of the rodent internal dose at the POD.
3 HECgg,median (or HECgg.gsicb) = the median (or 95th percentile lower confidence bound) estimate of the uncertainty
4 distribution of continuous exposure concentrations for which the 99th percentile individual has an internal dose less
5 than the (uncertain) rodent internal dose at the POD.
6 rtr = Route-to-route extrapolation using PBPK model and the specified dose metric
7 Shaded rows denote results for the primary dose metric
8 a BMDL used for p-cRfC or p-cRfD, but LOAEL or NOAEL (as noted) used for uncertainty analysis
9
6/22/2009
1096
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 Table 5.1.16. Comparison of "sensitive individual" HECs or HEDs for Immunological
2 effects based on PBPK modeled internal dose metrics at different levels of confidence and
3 sensitivity, at the NOAEL or LOAEL.
Candidate critical effect POD Ratio HECX or [Dose metric]
Candidate critical study type HEC/D50: HEDX
(species) HEC/D99
X=99
X=99,
median
X=99,
95lcb
Immunological
Changes in immunoreactive
organs - liver (including sporatic
necrosis in hepatic lobules),
spleen
Kaneko et al. 2000 (mouse)
HEC
2.65
36.7
38.3
16.0
[TotMetabBW34]
HEC
1.75
68.9
70.0
37.1
[AUCCBId]
HED
1.04
42.3
43.3
21.3
[TotMetabBW34] (rtr)
HED
3.21
56.5
59.0
39.8
[AUCCBId] (rtr)
t anti-dsDNA & anti-ssDNA Abs
(early markers for SLE); J, thymus
weight
Keil et al. 2009 (mouse)
HED
1.02
0.0482
0.0483
0.0380
[TotMetabBW34]
HED
12.1
0.0161
0.0189
0.00363
[AUCCBId]
HEC
2.77
0.0332
0.0337
0.0246
[TotMetabBW34] (rtr)
HEC
1.69
0.00821
0.00787
0.00199
[AUCCBId] (rtr)
I PFC response [NOAEL]3
Woolhiser et al. 2006 (rat)
HEC
2.54
16.1
15.2
7.56
[TotMetabBW34]
HEC
1.73
59.6
60.1
26.2
[AUCCBId]
HED
1.02
19.5
19.2
10.5
[TotMetabBW34] (rtr)
HED
3.21
52
55.9
33.0
[AUCCBId] (rtr)
I stem cell bone marrow
recolonization; J, cell-mediated
response to sRBC
Sanders et al. 1982 (mouse)
HED
1.02
2.48
2.48
1.94
[TotMetabBW34]
HED
10.5
0.838
0.967
0.187
[AUCCBId]
HEC
2.77
1.72
1.75
1.28
[TotMetabBW34] (rtr)
HEC 1.68 0.43 0.412 0.103 [AUCCBId] (rtr)
4 HEC99 = the 99th percentile of the combined human uncertainty and variability distribution of continuous exposure
5 concentrations that lead to the (fixed) median estimate of the rodent internal dose at the POD.
6 HEC99,median (or HECg9,95icb) = the median (or 95th percentile lower confidence bound) estimate of the uncertainty
7 distribution of continuous exposure concentrations for which the 99th percentile individual has an internal dose less
8 than the (uncertain) rodent internal dose at the POD.
9 rtr = Route-to-route extrapolation using PBPK model and the specified dose metric
10 Shaded rows denote results for the primary dose metric
11 a BMDL used for p-cRfC or p-cRfD, but LOAEL or NOAEL (as noted) used for uncertainty analysis
12
6/22/2009
1097
-------
1
2
3
4
5
6
7
8
9
10
11
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.17. Comparison of "sensitive individual" HECs or HEDs for Reproductive
effects based on PBPK modeled internal dose metrics at different levels of confidence and
sensitivity, at the NOAEL or LOAEL.
Candidate critical effect POD
Candidate critical study type
Ratio
HEC/D50:
HEC/D99
HECxor
HEDX
[Dose metric]
X=99
X=99,
median
X=99,
95lcb
Reproductive
hyperzoospermia
Chia et al. 1996 (human)
HEC
2.78
0.50
0.53
0.25
[TotMetabBW34]
HEC
1.68
0.83
0.83
0.49
[AUCCBId]
HED
1.02
0.73
0.71
0.37
[TotMetabBW34] (rtr)
HED
9.69
1.6
2.0
0.92
[AUCCBId] (rtr)
I fertilization
Xu et al. 2004 (mouse)
HEC
2.85
66.6
72.3
26.6
[TotMetabBW34]
HEC
1.89
170
171
97.1
[AUCCBId]
HED
1.09
73.3
76.9
32.9
[TotMetabBW34] (rtr)
HED
3.11
104
109
67.9
[AUCCBId] (rtr)
multiple sperm effects, testicular
enzyme markers
Kumar et al. 2000a 2001 b (rat)
HEC
2.53
12.8
12.2
6.20
[TotMetabBW34]
HEC
1.72
53.2
54.4
23.2
[AUCCBId]
HED
1.02
15.8
15.7
8.60
[TotMetabBW34] (rtr)
HED
3.21
48.8
52.6
30.6
[AUCCBId] (rtr)
I ability of sperm to fertilize in vitro
DuTeaux et al. 2004 (rat)
HED
4.20
15.6
18.1
4.07
[AUCCBId]
HED
1.57
41.7
41.9
32.0
[TotOxMetabBW34]
HEC
1.67
9.3
10.1
2.09
[AUCCBId] (rtr)
HEC
3.75
42.5
55.6
39.1
[TotOxMetabBW34] (rtr)
effects on epididymis epithelium
Forkert et al. 2002, Kan et al.
2007 (mouse)
HEC
2.85
66.6
72.3
26.6
[TotMetabBW34]
HEC
1.89
170
171
97.1
[AUCCBId]
HED
1.09
73.3
76.9
32.9
[TotMetabBW34] (rtr)
HED
3.11
104
109
67.9
[AUCCBId] (rtr)
testes effects
Kumar et al. 2000a 2001 b (rat)
HEC
2.53
12.8
12.2
6.20
[TotMetabBW34]
HEC
1.72
53.2
54.4
23.2
[AUCCBId]
HED
1.02
15.8
15.7
8.60
[TotMetabBW34] (rtr)
HED
3.21
48.8
52.6
30.6
[AUCCBId] (rtr)
delayed parturition
Narotsky et al. 1995 (rat)
HED
1.06
44.3
43.9
15.1
[TotMetabBW34]
HED
3.07
114
119
47.7
[AUCCBId]
HEC
2.66
36.9
35.3
11.6
[TotMetabBW34] (rtr)
HEC
1.91
190
197
48.1
[AUCCBId] (rtr)
I mating (both sexes exposed)
George et al. 1986 (rat)
HED
1.10
77.4
77.1
34.2
[TotMetabBW34]
HED
3.21
51.9
55.8
14.7
[AUCCBId]
HEC
2.86
71.1
70.0
29.5
[TotMetabBW34] (rtr)
HEC 1.73 59.5 63.3 8.14 [AUCCBId] (rtr)
HECgg = the 99th percentile of the combined human uncertainty and variability distribution of continuous exposure
concentrations that lead to the (fixed) median estimate of the rodent internal dose at the POD.
HECg
(or HECg
i = the median (or 95 percentile lower confidence bound) estimate of the uncertainty
distribution of continuous exposure concentrations for which the 99 percentile individual has an internal dose less
than the (uncertain) rodent internal dose at the POD.
rtr = Route-to-route extrapolation using PBPK model and the specified dose metric
Shaded rows denote results for the primary dose metric
6/22/2009
1098
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.1.18. Comparison of "sensitive individual" HECs or HEDs for Developmental
effects based on PBPK modeled internal dose metrics at different levels of confidence and
sensitivity, at the NOAEL or LOAEL.
Candidate critical effect POD Ratio HECX or [Dose metric]
Candidate critical study type HEC/D50: HEDX
(species) HEC/D99
X=99 X=95, X=95,
median 95lcb
Developmental
resorptions
Healy et al. 1982 (rat)
HED 1.02 8.5 8.50 4.61 [TotMetabBW34] (rtr)
HED 3.68 19.7 22.4 11.5 [AUCCBId] (rtr)
resorptions [LOAEL]3
Narotsky et al. 1995 (rat)
I fetal weight; skeletal effects
Healy et al. 1982 (rat)
heart malformations (pups)
[LOAEL]3
Johnson et al. 2003 (rat)
HED
11.6
0.00382
0.00476
0.00112
[AUCCBId]
HEC
2.75
0.00848
0.00866
0.00632
[TotOxMetabBW34]
(rtr)
HEC
1.70
0.00216
0.00221
0.000578
[AUCCBId] (rtr)
I rearing post-exp
Fredricksson et al. 1993
(mouse)
t exploration post-exp
Taylor etal. 1985 (rat)
HEC 2.57 8.36 7.94 5.95 [TotMetabBW34] (rtr)
HEC 1.68 2.19 2.31 0.580 [AUCCBId] (rtr)
HECgg = the 99th percentile of the combined human uncertainty and variability distribution of continuous exposure
concentrations that lead to the (fixed) median estimate of the rodent internal dose at the POD.
HECgg,median (or HECgg.gsicb) = the median (or 95th percentile lower confidence bound) estimate of the uncertainty
distribution of continuous exposure concentrations for which the 99th percentile individual has an internal dose less
than the (uncertain) rodent internal dose at the POD.
rtr = Route-to-route extrapolation using PBPK model and the specified dose metric
Shaded rows denote results for the primary dose metric
a BMDL used for p-cRfC or p-cRfD, but LOAEL or NOAEL (as noted) used for uncertainty analysis
HEC 2.58 6.19 6.02 3.13 [TotMetabBW34]
HEC 1.69 13.7 13.9 7.27 [AUCCBId]
HED
1.06
44.3
43.9
15.1
[TotMetabBW34]
HED
3.07
114
119
47.7
[AUCCBId]
HEC
2.66
36.9
35.3
11.6
[TotMetabBW34] (rtr)
HEC
1.91
190
197
48.1
[AUCCBId] (rtr)
HEC
2.58
6.19
6.02
3.13
[TotMetabBW34]
HEC 1.69 13.7 13.9 7.27 [AUCCBId]
HED
1.02
8.5
8.50
4.61
[TotMetabBW34] (rtr)
HED
3.68
19.7
22.4
11.5
[AUCCBId] (rtr)
HED
1.02
0.012
0.012
0.0102
[TotOxMetabBW34]
HED
1.02
4.13
4.19
2.22
[TotMetabBW34]
HED
7.69
3.46
4.21
0.592
[AUCCBId]
HEC
2.71
2.96
2.96
1.48
[TotMetabBW34] (rtr)
HEC
1.68
1.84
1.81
0.302
[AUCCBId] (rtr)
HED
1.02
10.7
10.7
8.86
[TotMetabBW34]
HED 7.29 4.11 5.08 1.16 [AUCCBId]
6/22/2009
1099
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.5 Summary of non-cancer reference values
5.1.5.1 Preferred candidate reference values (cRfCs, cRfD, p-cRfCs andp-cRfDs) for
candidate critical effects
The candidate critical effects which yielded the lowest p-cRfC or p-cRfD, based on the
primary dose metric, for each type of effect are summarized in Tables 5.1.19 (p-cRfCs) and
5.1.20 (p-cRfDs). These results are extracted from Tables 5.1.8-5.1.13. In cases where a route-
to-route extrapolated p-cRfC (p-cRfD) is lower than the lowest p-cRfC (p-cRfD) from an
inhalation (oral) study, both values are presented in the Table. In addition, if there is greater than
usual uncertainty associated with the lowest p-cRfC or p-cRfD for a type of effect, then the
endpoint with the next lowest value is also presented. Furthermore, given those selections, the
same sets of critical effects and studies are displayed across both Tables, with the exception of
two oral studies for which route-to-route extrapolation was not performed. Tables 5.1.19 and
5.1.20 are further summarized in Tables 5.1.21 and 5.1.22 to present the overall preferred p-cRfC
and p-cRfD for each type of non-cancer effect. The purpose of these summary Tables is to show
the most sensitive endpoints for each type of effect and the apparent relative sensitivities (based
on reference value estimates) of the different types of effects.
For neurological, kidney, immunological, and developmental effects, the lowest p-cRfCs
were derived from oral studies by route-to-route extrapolation. This appears to be a function of
the lack of comparable inhalation studies for many effects studied via the oral exposure route, for
which there is a larger database of studies. For the liver and reproductive effects, inhalation
studies yielded a p-cRfC lower than the lowest route-to-route extrapolated p-cRfC for that type
of effect. Conversely, the lowest p-cRfDs were derived from oral studies with the exception of
reproductive effects, for which route-to-route extrapolation from an inhalation study in humans
also yielded among the lowest p-cRfDs. The only effect for which there were comparable
studies for comparing a p-cRfC from an inhalation study with a p-cRfC estimated by route-to-
route extrapolation from an oral study was increased liver weight in the mouse. The primary
dose metric of amount of TCE oxidized in the liver yielded similar p-cRfCs of 1.0 ppm and 1.1
ppm for the inhalation result and the route-to-route extrapolated result, respectively (see Table
5.1.10).
As can be seen in these Tables, the most sensitive types of effects (the types with the
lowest p-cRfCs and p-cRfDs) appear to be developmental, kidney, and immunological (adult and
developmental) effects, and then neurological and reproductive effects, in that order. Lastly, the
liver effects have p-cRfC and p-cRfD values that are about 3 V2 orders of magnitude higher than
those for developmental, kidney, and immunological effects.
6/22/2009
1100
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 Table 5.1.19. Lowest p-cRfCs or cRfCs for different effect domains
Effect domain Candidate critical effect p-cRfC or cRfC in ppm
(Species/Critical Study) (composite uncertainty factor)
Effect type
Preferred
Default
Alternative dose
dose
methodology
metrics/studies
metric3
(Tables 5.1.8-13)
Neurologic
Trigeminal
Trigeminal nerve effects
0.53
0.47
0.83
nerve effects
(human/Ruitjen et al. 1991)
(10)
(30)
(10)
Cognitive effects
Demyelination in hippocampus
0.0071
-
0.0023
(rat/Isaacson etal. 1990)
(1000)
[rtr]
(1000)
Mood/sleep
Changes in wakefulness
0.016
0.012
0.030
changes
(rat/Arito et al. 1994)
(300)
(1000)
(300)
Kidney
Histological
Toxic nephropathy
0.00056
-
changes
(rat/NTP 1988)
(10)
[rtr]
0.00087-1.3
Toxic nephrosis
0.0017
-
(10-300)
(mouse/NC11976)
(300)
[rtr]
1 kidney weight
f kidney weight
0.0013
0.52
0.0022-2.1
(rat/Woolhiser et al. 2006)
(10)
(30)
(10-30)
Liver
f liver weight
f liver weight
0.91
0.72
0.83-2.5
(mouse/Kjellstrand et al. 1983b)
(10)
(30)
(10-30)
Immunologic
| thymus weight
| thymus weight
0.00033
-
0.000082
(mouse/Keil et al. 2009)
(100)
[rtr]
(100)
Immuno-
| stem cell recolonization
0.057
-
suppression
(mouse/Sanders et al. 1982)
(30)
[rtr]
0.014-1.4
Decreased PFC response
0.11
0.083
(30-100)
(rat/Woolhiser et al. 2006)
(100)
(300)
Autoimmunity
f anti-dsDNA & anti-ssDNA Abs
0.0033
-
(mouse/Keil et al. 2009)
(10)
[rtr]
0.00082-0.23
Autoimmune organ changes
0.12
0.070
(10-300)
(mouse/Kaneko et al. 2000)
(300)
(1000)
Reproductive
Effects on
| ability of sperm to fertilize
0.0093
-
sperm & testes
(rat/DuTeaux et al. 2004)
(1000)
[rtr]
Multiple effects
0.013
0.015
0.028-0.17
(rat/Kumar et al. 2000a, 2001b)
(1000)
(3000)
(30-1000)
Hyperzoospermia
0.017
0.014
(human/Chia etal. 1996)b
(30)
(100)
Developmental
Congenital
Heart malformations
0.00037
-
0.000093
defects
(rat/Johnson et al. 2003)
(10)
[rtr]
(10)
Develop.
| rearing post-exposure
0.028
-
0.0077-0.084
neurotox
(rat/Fredricksson et al. 1993)
(300)
[rtr]
(100-300)
Pre-/post-
Resorptions/^ fetal weight/
0.062
0.057
0.14-2.4
natal mortality/
skeletal effects
(100)
(300)
(10-100)
growth
(rat/Healy etal. 1982)
2 a The critical effects/studies and p-cRfCs supporting the RfC are in bold.
3 b greater than usual degree of uncertainty (see Section 5.1.2)
4 rtr: route-to-route extrapolated result
6/22/2009
1101
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 Table 5.1.20. Lowest p-cRfDs or cRfDs for different effect domains
Effect domain Candidate critical effect p-cRfD or cRfD in mg/kg/d
(Species/Critical Study) (composite uncertainty factor)
Effect type
Preferred
Default
Alternative dose
dose
methodology
metrics/studies
metric3
(Tables 5.1.8-13)
Neurologic
Trigeminal
Trigeminal nerve effects
0.73
-
1.4
nerve effects
(human/Ruitjen et al. 1991)
(10)
[rtr]
(10)
Cognitive effects
Demyelination in hippocampus
0.0092
0.0047
0.0043
(rat/Isaacson etal. 1990)
(1000)
(10,000b)
(1000)
Mood/sleep
Changes in wakefulness
0.022
-
0.051
changes
(rat/Arito et al. 1994)
(300)
[rtr]
(300)
Kidney
Histological
Toxic nephropathy
0.00034
0.0945
changes
(rat/NTP 1988)
(10)
(100)
0.00053-1.9
Toxic nephrosis
0.0010
(10-300)
(mouse/NCI 1976)
(300)
1 kidney weight
f kidney weight
0.00079
-
0.0013-2.5
(rat/Woolhiser et al. 2006)
(10)
[rtr]
(10)
Liver
f liver weight
f liver weight
0.79
-
0.82-2.6
(mouse/Kjellstrand et al. 1983b)
(10)
[rtr]
(10-100)
Immunologic
| thymus weight
| thymus weight
0.00048
0.00035
0.00016
(mouse/Keil et al. 2009)
(100)
(1000)
(100)
Immuno-
| stem cell recolonization
0.083
0.060
suppression
(mouse/Sanders et al. 1982)
(30)
(300)
0.028-0.91
Decreased PFC response
0.14
-
(30-100)
(rat/Woolhiser et al. 2006)
(100)
[rtr]
Autoimmunity
f anti-dsDNA & anti-ssDNA Abs
0.0048
0.0035
(mouse/Keil et al. 2009)
(10)
(100)
0.0016-0.19
Autoimmune organ changes
0.14
-
(10-300)
(mouse/Kaneko et al. 2000)
(300)
[rtr]
Reproductive
Effects on
| ability of sperm to fertilize
0.016
0.014
sperm & testes
(rat/DuTeaux et al. 2004)
(1000)
(10,000b)
Multiple effects
0.016
-
0.042-0.10
(rat/Kumar et al. 2000a, 2001b)
(1000)
[rtr]
(30-1000)
Hyperzoospermia
(human/Chia etal. 1996)°
0.024
(30)
[rtr]
Developmental
Develop.
| PFC, T DTH
0.00037
Same as
-
Immunotox
(rat/Peden-Adams et al. 2006)d
(1000)
preferred
Congenital
Heart malformations
0.00052
0.00021
0.00017
defects
(rat/Johnson et al. 2003)
(10)
(100)
(10)
Develop.
| rearing post-exposure
0.016
Same as
0.017-0.11
Neurotox
(rat/Fredricksson et al. 1993)d
(1000)
preferred
(100-3000)
Pre-/post-
Resorptions/^ fetal weight/
0.085
[rtr]
0.70-2.9
natal mortality/
skeletal effects
(100)
(10-100)
growth
(rat/Healy etal. 1982)
2 a The critical effects/studies and p-cRfDs or cRfDs supporting the RfD are in bold.
3 b U.S. EPA's report on the RfC and RfD processes (U.S. EPA, 2002) recommends not deriving reference values with
4 a composite UF of greater than 3000; however, composite UFs exceeding 3000 are considered here because the
6/22/2009
1102
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
derivation of the cRfCs and cRfDs is part of a screening process and the application of the PBPK model for candidate
critical effects reduces the values of some of the individual UFs for the p-cRfCs and p-cRfDs.
c greater than usual degree of uncertainty (see Section 5.1.2)
d No PBPK model based analyses were done, so cRfD on the basis of applied dose only,
rtr: route-to-route extrapolated result (no value for default methodology)
Table 5.1.21. Lowest p-cRfCs for candidate critical effects for different types of effect
based on primary t
ose metric
Type of effect
effect
(primary dose metric)
p-cRfC (ppm)
Neurological
demyelination in hippocampus in rats
(T otMetabB W3 4)
0.007 (rtr)
Kidney
toxic nephropathy in rats
(ABioactDCVCBW34)
0.0006 (rtr)
Liver
increased liver weight in mice
(AMetLivlBW34)
0.9
Immunological
decreased thymus weight in mice
(T otMetabB W3 4)
0.0003 (rtr)
Reproductive
decreased ability of rat sperm to fertilize
(AUCCBld)
0.009 (rtr)a
Developmental
heart malformations in rats
(T otOxMetabB W 34)
0.0004 (rtr)
rtr: route-to-route extrapolated result
a this value is supported by the p-cRfC value of 0.01 ppm for multiple testes and sperm effects
from an inhalation study in rats.
Table 5.1.22. Lowe
based on primary t
st p-cRfDs for candidate critical effects for different types of effect
ose metric
Type of effect
effect
(primary dose metric)
p-cRfD
(mg/kg/d)
Neurological
demyelination in hippocampus in rats
(T otMetabB W3 4)
0.009
Kidney
toxic nephropathy in rats
(ABioactDCVCBW34)
0.0003
Liver
increased liver weight in mice
(AMetLivlBW34)
0.8 (rtr)
Immunological
decreased thymus weight in mice
(T otMetabB W3 4)
0.0005
Reproductive
decreased ability of rat sperm to fertilize (AUCCBld) &
multiple testes and sperm effects (TotMetabBW34)a
0.02
Developmental
heart malformations in rats
(T otOxMetabB W 34)
0.0005b
a endpoints from 2 different studies yielded the same p-cRfD value
b this value is supported by the cRfD value of 0.0004 mg/kg/day derived for developmental
immuntoxicity effects in mice (Peden-Adams et al., 2006); however, no PBPK analyses were
done for this latter effect, so the value of 0.0004 mg/kg/d is based on applied dose
6/22/2009
1103
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
rtr: route-to-route extrapolated result
5.1.5.2 Reference Concentration
The goal is to select an overall RfC that is well supported by the available data (i.e.,
without excessive uncertainty given the extensive database) and protective for all the candidate
critical effects, recognizing that individual candidate RfC values are by nature somewhat
imprecise. The lowest candidate RfC values within each health effect category span a 3000-fold
range from 0.0003-0.9 ppm (Table 5.1.21). One approach to selecting a RfC would be to select
the lowest calculated value of 0.0003 ppm for decreased thymus weight in mice. However, as
can be seen in Table 5.1.19, six p-cRfCs from both oral and inhalation studies are in the
relatively narrow range of 0.0003-0.003 ppm at the low end of the overall range. Given the
somewhat imprecise nature of the individual candidate RfC values, and the fact that multiple
effects/studies lead to similar candidate RfC values, the approach taken in this assessment is to
select a RfC supported by multiple effects/studies. The advantages of this approach, which is
only possible when there is a relatively large database of studies/effects and when multiple
candidate values happen to fall within a narrow range at the low end of the overall range, are that
it leads to a more robust RfC (less sensitive to limitations of individual studies) and that it
provides the important characterization that the RfC exposure level is similar for multiple non-
cancer effects rather than being based on a sole explicit critical effect.
Table 5.1.23 summarizes the PODs and UFs for the six critical studies/effects
corresponding to the p-cRfCs that have been chosen to support the RfC for TCE non-cancer
effects. Five of the lowest candidate p-cRfCs, ranging from 0.0003-0.003 ppm, for
developmental, kidney, and immunologic effects, are values derived from route-to-route
extrapolation using the PBPK model. The lowest p-cRfC estimate (for a primary dose metric)
from an inhalation studies is 0.001 ppm for kidney effects. For all six candidate RfCs, the PBPK
model was used for inter- and intra-species extrapolation, based on the preferred dose metric for
each endpoints. There is high confidence in the p-cRfCs for kidney effects (see Section 5.1.2.2)
for the following reasons: they are based on clearly adverse effects, two of the values are derived
from chronic studies, and the extrapolation to humans is based on dose metrics clearly related to
toxicity estimated with high confidence with the PBPK model developed in Section 3.5. There is
somewhat less confidence in the lowest p-cRfC for developmental effects (heart malformations)
(see Section 5.1.2.8) and the lowest p-cRfC estimates for immunological effects (see Section
5.1.2.5). Thus, we do not rely on any single estimate alone; however, each estimate is supported
by estimates of similar magnitude from other effects.
As a whole, the estimates support a preferred RfC estimate of 0.001 ppm (1 ppb or 5
(j,g/m3). This estimate is within approximately a factor of 3 of the lowest estimates of 0.0003
6/22/2009
1104
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
ppm for decreased thymus weight in mice, 0.0004 ppm for heart malformations in rats, 0.0006
ppm for toxic nephropathy in rats, 0.001 ppm for increased kidney weight in rats, 0.002 ppm for
toxic nephrosis in mice, and 0.003 ppm for increased anti-dsDNA antibodies in mice. Thus,
there is robust support for a RfC of 0.001 ppm provided by estimates for multiple effects from
multiple studies. The estimates are based on PBPK model-based estimates of internal dose for
inter-species, intra-species, and/or route-to-route extrapolation, and there is sufficient confidence
in the PBPK model, as well as support from mechanistic data for some of the dose metrics
(specifically TotOxMetabBW34 for the heart malformations and ABioactDCVCBW34 and
AMetGSHBW34 for toxic nephropathy) (see Section 5.1.3.1). Note that there is some human
evidence of developmental heart defects from TCE exposure in community studies (see Section
4.7.3.1.1) and of kidney toxicity in TCE-exposed workers (Section 4.3.1).
In summary, the preferred RfC estimate is 0.001 ppm (1 ppb or 5 |ig/m3) based on route-
to-route extrapolated results from oral studies for the critical effects of heart malformations
(rats), immunotoxicity (mice), and toxic nephropathy (rats, mice), and an inhalation study for the
critical effect of increased kidney weight (rats).
Table 5.1.23. Summary of Critical Studies, Effects, POPs, and UFs supporting the RfC
NTP (1988) - Toxic nephropathy in female Marshall rats exposed for 104 weeks by oral gavage (5 d/wk)
• iPOD = 0.0132 mg DCVC bioactivated/kg3/Vd, which is the BMDL from BMD modeling using
PBPK model-predicted internal doses, BMR=5% (clearly toxic effect), and loglogistic model
(See Appendix F, Section F.7.1).
• HEC99 = 0.0056 ppm (lifetime continuous exposure) derived from combined inter-species, intra-
species, and route-to-route extrapolation using PBPK model.
• UF1S = 3.16 because the PBPK model was used for inter-species extrapolation
• UFh = 3.16 because the PBPK model was used to characterize human toxicokinetic variability
• p-cRfC = 0.0056 / 10 = 0.00056 ppm (3 ^ig/m3)
NCI (1976) - Toxic nephrosis in female B3C3F1 mice exposed for 78 weeks by oral gavage (5 d/wk)
• iPOD = 0.735 mg TCE conjugated with GSH/kg Vd. which is the PBPK model-predicted internal
dose at the applied dose LOAEL of 869 mg/kg/d (5 d/wk) (BMD modeling failed due to almost
maximal response at lowest dose) (See Appendix F, Section F.7.2).
• HEC99 = 0.50 ppm (lifetime continuous exposure) derived from combined inter-species, intra-
species, and route-to-route extrapolation using PBPK model.
• UFioaei = 30 because POD is a LOAEL for an adverse effect with a response > 90%.
• UF1S = 3.16 because the PBPK model was used for inter-species extrapolation
• UFh = 3.16 because the PBPK model was used to characterize human toxicokinetic variability
• p-cRfC = 0.50 / 300 = 0.0017 ppm (0.9 ^ig/m3)
6/22/2009
1105
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Woolhiser et al. (2006) - Increased kidney weight in female SD rats exposed for 4 weeks by inhalation
(6 hr/d, 5 d/wk)
• iPOD = 0.0309 mg DCVC bioactivated/kg /d. which is the BMDL from BMD modeling using
PBPK model-predicted internal doses, BMR=10%, and Hill model with constant variance (See
Appendix F, Section F.7.3).
• HEC99 = 0.013 ppm (lifetime continuous exposure) derived from combined inter-species and
intra-species extrapolation using PBPK model.
• UFSC = 1 because Kjellstrand et al. (1983b) reported that in mice, kidney effects after exposure for
120 days was no more severe than those after 30 days exposure.
• UF1S = 3.16 because the PBPK model was used for inter-species extrapolation
• UFh = 3.16 because the PBPK model was used to characterize human toxicokinetic variability
• P-cRfC = 0.013 / 10 = 0.0013 ppm (7 ^ig/m3)
Keil et al. (2009) - Decreased thymus weight in female B6C3F1 mice exposed for 30 weeks by drinking
water
• iPOD = 0.139 mg TCE metabolized/kg Vd. which is the PBPK model-predicted internal dose at
the applied dose LOAEL of 0.35 mg/kg/d (continuous) (no BMD modeling due to inadequate
model fit caused by supralinear dose-response shape) (See Appendix F, Section F.7.4).
• HEC99 = 0.033 ppm (lifetime continuous exposure) derived from combined inter-species, intra-
species, and route-to-route extrapolation using PBPK model.
• UFioaei =10 because POD is a LOAEL for an adverse effect.
• UF1S = 3.16 because the PBPK model was used for inter-species extrapolation
• UFh = 3.16 because the PBPK model was used to characterize human toxicokinetic variability
• p-cRfC = 0.033 / 100 = 0.00033 ppm (2 ^ig/m3)
Keil et al. (2009) - Increased anti-dsDNA and anti-ssDNA antibodies in female B6C3F1 mice exposed
for 30 weeks by drinking water
• iPOD = 0.139 mg TCE metabolized/kg Vd, which is the PBPK model-predicted internal dose at
the applied dose LOAEL of 0.35 mg/kg/d (continuous) (no BMD modeling due to inadequate
model fit caused by supralinear dose-response shape) (See Appendix F, Section F.7.4).
• HEC99 = 0.033 ppm (lifetime continuous exposure) derived from combined inter-species, intra-
species, and route-to-route extrapolation using PBPK model.
• UFioaei = 1 because POD is a LOAEL for an early marker for an adverse effect.
• UF1S = 3.16 because the PBPK model was used for inter-species extrapolation
• UFh = 3.16 because the PBPK model was used to characterize human toxicokinetic variability
• p-cRfC = 0.033 / 10 = 0.0033 ppm (18 ^ig/m3)
lohnson et al. (2003) - fetal heart malformations in SD rats exposed from GD 1-22 by drinking water
• iPOD = 0.0142 mg TCE metabolized by oxidation/kg'Vd, which is the BMDL from BMD
modeling using PBPK model-predicted internal doses, with highest dose group (1000-fold higher
than next highest dose group) dropped, pup as unit of analysis, BMR=1% (due to severity of
defects, some of which could have been fatal), and a nested loglogistic model to account for intra-
litter correlation (See Appendix F, Section F.7.5).
• HEC99 = 0.0037 ppm (lifetime continuous exposure) derived from combined inter-species, intra-
species, and route-to-route extrapolation using PBPK model.
• UF1S = 3.16 because the PBPK model was used for inter-species extrapolation
• UFh = 3.16 because the PBPK model was used to characterize human toxicokinetic variability
» P-cRfC = 0.0037 / 10 = 0.00037 ppm (2 ng/m3)
6/22/2009
1106
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.1.5.3 Reference Dose
As with the RfC determination above, the goal is to select an overall RfD that is well
supported by the available data (i.e., without excessive uncertainty given the extensive database)
and protective for all the candidate critical effects, recognizing that individual candidate RfD
values are by nature somewhat imprecise. The lowest candidate RfD values within each health
effect category span a nearly 3000-fold range from 0.0003-0.8 mg/kg/d (Table 5.1.21). One
approach to selecting a RfC would be to select the lowest calculated value of 0.0003 ppm for
toxic nephropathy in rats. However, as can be seen in Table 5.1.20, multiple p-cRfDs or cRfDs
from oral studies are in the relatively narrow range of 0.0003-0.0005 mg/kg/d at the low end of
the overall range. Given the somewhat imprecise nature of the individual candidate RfD values,
and the fact that multiple effects/studies lead to similar candidate RfD values, the approach taken
in this assessment is to select a RfD supported by multiple effects/studies. The advantages of
this approach, which is only possible when there is a relatively large database of studies/effects
and when multiple candidate values happen to fall within a narrow range at the low end of the
overall range, are that it leads to a more robust RfD (less sensitive to limitations of individual
studies) and that it provides the important characterization that the RfD exposure level is similar
for multiple non-cancer effects rather than being based on a sole explicit critical effect.
Table 5.1.24 summarizes the PODs and UFs for the four critical studies/effects
corresponding to the p-cRfDs or cRfDs that have been chosen to support the RfD for TCE non-
cancer effects. Three of the lowest p-cRfDs for the primary dose metrics - 0.0003 mg/kg/d for
toxic nephropathy in rats, and 0.0005 mg/kg/d for heart malformations in rats and decreased
thymus weights in mice - are derived using the PBPK model for inter- and intra-species
extrapolation. The other of these lowest values - 0.0004 mg/kg/d for developmental
immunotoxicity (decreased PFC response and increased delayed-type hypersensitivity) in mice -
is based on applied dose. There is high confidence in the p-cRfD for kidney effects (see Section
5.1.2.2), which is based on clearly adverse effects, derived from a chronic study, and
extrapolated to humans based on a dose metric clearly related to toxicity estimated with high
confidence with the PBPK model developed in Section 3.5. There is somewhat less confidence
in the p-cRfDs for decreased thymus weights (see Section 5.1.2.5) and heart malformations and
developmental immunological effects (see Section 5.1.2.8). Thus, we do not rely on any single
estimate alone; however, each estimate is supported by estimates of similar magnitude from
other effects.
As a whole, the estimates support a preferred RfD of 0.0004 mg/kg/d. This estimate is
within 25% of the lowest estimates of 0.0003 for toxic nephropathy in rats, 0.0004 mg/kg/d for
developmental immunotoxicity (decreased PFC and increased delay ed-type hypersensitivity) in
mice, and 0.0005 mg/kg/d for heart malformations in rats and decreased thymus weights in mice.
6/22/2009
1107
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Thus, there is strong, robust support for a RfD of 0.0004 mg/kg/d provided by the concordance
of estimates derived from multiple effects from multiple studies. The estimates for kidney
effects, thymus effects, and developmental heart malformations are based on PBPK model-based
estimates of internal dose for inter-species and intra-species extrapolation, and there is sufficient
confidence in the PBPK model, as well as support from mechanistic data for some of the dose
metrics (specifically TotOxMetabBW34 for the heart malformations and ABioactDCVCBW34
for toxic nephropathy) (see Section 5.1.3.1). Note that there is some human evidence of
developmental heart defects from TCE exposure in community studies (see Section 4.7.3.1.1)
and of kidney toxicity in TCE-exposed workers (Section 4.3.1).
In summary, the preferred RfD estimate is 0.0004 mg/kg/d based on the critical effects of
heart malformations (rats), adult immunological effects (mice), developmental immunotoxicity
(mice), and toxic nephropathy (rats).
6/22/2009
1108
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 Table 5.1.24. Summary of Critical Studies, Effects, POPs, and UFs supporting the RfD
NTP (1988) - Toxic nephropathy in female Marshall rats exposed for 104 weeks by oral gavage (5 d/wk)
• iPOD = 0.0132 mg DCVC bioacti v ated/kg /d. which is the BMDL from BMD modeling using
PBPK model-predicted internal doses, BMR=5% (clearly toxic effect), and loglogistic model
(See Appendix F, Section F.7.1).
• HED99 = 0.0034 mg/kg/d (lifetime continuous exposure) derived from combined inter-species and
intra-species extrapolation using PBPK model.
• UF1S = 3.16 because the PBPK model was used for inter-species extrapolation
• UFh = 3.16 because the PBPK model was used to characterize human toxicokinetic variability
• p-cRfD = 0.0034/ 10 = 0.00034 mg/kg/d
Keil et al. (2009) - Decreased thymus weight in female B6C3F1 mice exposed for 30 weeks by drinking
water
• iPOD = 0.139 mg TCE metabolized/kg Yd, which is the PBPK model-predicted internal dose at
the applied dose LOAEL of 0.35 mg/kg/d (continuous) (no BMD modeling due to inadequate
model fit caused by supralinear dose-response shape) (See Appendix F, Section F.7.4).
• HED99 = 0.048 mg/kg/d (lifetime continuous exposure) derived from combined inter-species and
intra-species extrapolation using PBPK model.
• UFioaei =10 because POD is a LOAEL for an adverse effect.
• UF1S = 3.16 because the PBPK model was used for inter-species extrapolation
• UFh = 3.16 because the PBPK model was used to characterize human toxicokinetic variability
• p-cRfD = 0.048 / 100 = 0.00048 mg/kg/d
Peden-Adams et al. (2006) - Decreased PFC response (3 and 8 weeks), increased delayed-type
hypersensitivity (8 weeks) in pups exposed from GD0 to 3 or 8 weeks of age through drinking water
(placental and lactational transfer, and pup ingestion)
• POD = 0.37 mg/kg/d is the applied dose LOAEL (estimated daily dam dose) (no BMD modeling
due to inadequate model fit caused by supralinear dose-response shape). No PBPK modeling was
attempted due to lack of appropriate models/parameters to account for complicated fetal/pup
exposure pattern (See Appendix F, Section F.7.6).
• UFioaei =10 because POD is a LOAEL for multiple adverse effects.
• UF1S = 10 for inter-species extrapolation because PBPK model was not used
• UFh = 10 for human variability because PBPK model was not used
• cRfD = 0.37/ 1000 = 0.00037 mg/kg/d
Johnson et al. (2003) - fetal heart malformations in SD rats exposed from GD 1-22 by drinking water
• iPOD = 0.0142 mg TCE metabolized by oxidation/kg'Vd, which is the BMDL from BMD
modeling using PBPK model-predicted internal doses, with highest dose group (1000-fold higher
than next highest dose group) dropped, pup as unit of analysis, BMR=1% (due to severity of
defects, some of which could have been fatal), and a nested loglogistic model to account for intra-
litter correlation (See Appendix F, Section F.7.5).
• HED99 = 0.0051 mg/kg/d (lifetime continuous exposure) derived from combined inter-species and
intra-species extrapolation using PBPK model.
• UF1S = 3.16 because the PBPK model was used for inter-species extrapolation
• UFh = 3.16 because the PBPK model was used to characterize human toxicokinetic variability
» p-cRfD = 0.0051 / 10 = 0.00051 mg/kg/d
6/22/2009
1109
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.2 Dose-Response Analysis for Cancer Endpoints
This section describes the dose-response analysis for cancer endpoints. Section 5.2.1
discusses the analyses of data from chronic rodent bioassays. Section 5.2.2 discusses the
analyses of human epidemiologic data. Section 5.2.3 discusses the choice of the preferred
inhalation unit risk and oral unit risk estimates, as well as the application of age-dependent
adjustment factors to the unit risk estimates.
5.2.1 Dose-Response Analyses: Rodent Bioassays
This section describes the estimation of cancer unit risks based on rodent bioassays.
First, all the available studies (i.e., chronic rodent bioassays) were considered, and those suitable
for dose-response modeling were selected for analysis (Section 5.2.1.1). Then dose-response
modeling using the linearized multi-stage model was performed using applied doses (default
dosimetry) as well as PBPK model-based internal doses (Section 5.2.1.2). Bioassays for which
time-to-tumor data were available were analyzed using poly-3 adjustment techniques and using a
the multi-stage Weibull model. In addition, a cancer potency estimate for different tumor types
combined was derived from bioassays in which there was more than one type of tumor response
in the same sex and species. Unit risk estimates based on PBPK model-estimated internal doses
were then extrapolated to human population unit risk estimates using the human PBPK model.
From these results (Section 5.2.1.3), estimates from the most sensitive bioassay (i.e., that with
the greatest unit risk estimate) for each combination of administration route, sex, and species,
based on the PBPK model-estimated internal doses, were considered as candidate unit risk
estimates for TCE. Uncertainties in the rodent-based dose-response analyses are described in
Section 5.2.1.4.
5.2.1.1 Rodent dose-response analyses: Studies and Modeling Approaches
The rodent cancer bioassays that were identified for consideration for dose-response
analysis are listed in Tables 5.2.1 (inhalation bioassays) and 5.2.2 (oral bioassays) for each
sex/species combination. The bioassays selected for dose-response analysis are marked with an
asterisk; rationales for rejecting the bioassays that were not selected are provided in the
"Comments" columns of the Tables. For the selected bioassays, the tissues/organs that exhibited
a TCE-associated carcinogenic response and for which dose-response modeling was performed
are listed in the "Tissue/Organ" columns.
6/22/2009
1110
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 Table 5.2.1. Inhalation bioassays
Study
Strain
Tissue/Organ Comments
Female mice
*Fukuda et al. (1983)
*Henschler et al. (1980)
*Maltoni et al. (1986)
Maltoni et al. (1986)
Cij:CD-l
(ICR)
Han:NMRI
B6C3F1
Swiss
Lung
Lymphoma
Liver, Lung
No dose-response
Male mice
Henschler et al. (1980)
Maltoni et al. (1986)
Maltoni et al. (1986)
Han:NMRI
B6C3F1
B6C3F1
"Maltoni et al. (1986) Swiss
Liver
Liver
Liver
No dose-response
Exp #BT306: excessive fighting
Exp #BT306bis. Results
similar to Swiss mice
Female rats
Fukuda et al. (1983)
Henschler et al. (1980)
Maltoni et al. (1986)
Sprague-
Dawley
Wistar
Sprague-
Dawley
No dose-response
No dose-response
No dose-response
2
3
4
5
Male rats
Henschler et al. (1980)
*Maltoni et al. (1986)
Wistar
Sprague-
Dawley
Kidney, Leydig
cell, Leukemia
No dose-response
* Selected for dose-response analysis
'No dose-response"= no tumor incidence data suitable for dose-response modeling.
Table 5.2.2 Oral bioassays
Study Strain
Tissue/Organ Comments
Female mice
Henschler et al. (1984)
*NCI (1976)
NTP (1990)
HanNMRI
B6C3F1
B6C3F1
Toxicity, no dose-response
VanDuren et al. (1979) Swiss
Liver, Lung,
sarcomas and
lymphomas
Liver, Lung,
Lymphomas
Liver
Single dose
Single dose, no dose-response
Male mice
Anna et al. (1994)
Bull et al. (2002)
Henschler et al. (1984)
*NCI (1976)
NTP (1990)
B6C3F1
B6C3F1
HanNMRI
B6C3F1
B6C3F1
Liver
Liver
Liver
Liver
Single dose
Single dose
Toxicity, no dose-response
Single dose
6/22/2009
1111
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
VanDuren et al. (1979)
Swiss
-
Single dose, no dose-response
Female rats
NCI (1976)
Osborne-
Mendel
—
Toxicity, no dose-response
NTP (1988)
ACI
-
No dose-response
*NTP (1988)
August
Leukemia
NTP (1988)
Marshall
-
No dose-response
NTP (1988)
Osborne-
Mendel
Adrenal cortex
Adenomas only
NTP(1990)
F344/N
-
No dose-response
Male rats
NCI (1976)
Osborne-
Mendel
—
Toxicity, no dose-response
NTP (1988)
ACI
-
No dose-response
*NTP (1988)
August
Subcutaneous
tissue sarcomas
*NTP (1988)
Marshall
Testes
*NTP (1988)
Osborne-
Mendel
Kidney
*NTP(1990)
F344/N
Kidney
* Selected for dose-response analysis
"No dose-response"= no tumor incidence data suitable for dose-response modeling.
The general approach used was to model each sex/species/bioassay tumor response to
determine the most sensitive bioassay response (in terms of human equivalent exposure or dose)
for each sex/species combination. The various modeling approaches, model selection, and unit
risk derivation are discussed below. Modeling was done using the applied dose or exposure
(default dosimetry) and several internal dose metrics. The dose metrics used in the dose-
response modeling are discussed in Section 5.2.1.2. Because of the large volume of analyses and
results, detailed discussions about how the data were modeled using the various dosimetry and
modeling approaches and results for individual datasets are provided in Appendix G. The
overall results are summarized and discussed in Section 5.2.1.3.
Most tumor responses were modeled using the multistage model in EPA's Benchmark
Dose Software (BMDS) (www.epa.gov/ncea/bmds). The multistage model is a flexible model,
capable of fitting most cancer bioassay data, and it is EPA's long-standing model for the
modeling of such cancer data. The multistage model has the general form
P(d) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk)],
6/22/2009
1112
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
where P(d) represents the lifetime risk (probability) of cancer at dose d, and parameters q, > 0,
for i = 0, 1, ..., k. For each dataset, the multistage model was evaluated for one stage and (n - 1)
stages, where n is the number of dose groups in the bioassay. A detailed description of how the
data were modeled, as well as tables of the dose-response input data and figures of the multistage
modeling results, is provided in Appendix G.
Only models with acceptable fit (p > 0.05) were considered. If 1-parameter and 2-
parameter models were both acceptable (in no case was there a 3-parameter model), the more
parsimonious model (i.e., the 1-parameter model) was selected unless the inclusion of the 2nd
parameter resulted in a statistically significant42 improvement in fit. If two different 1-parameter
models were available (e.g., a 1-stage model and a 3-stage model with Pi and p2 both equal to 0),
the one with the best fit, as indicated by the lowest AIC value, was selected. If the AIC values
were the same (to 3 significant figures), then the lower-stage model was selected. Visual fit and
scaled Chi-square residuals were also considered for confirmation in model selection. For two
datasets, the highest dose group was dropped to improve the fit in the lower dose range.
From the selected model for each dataset, the maximum likelihood estimate (MLE) for
the dose corresponding to a specified level of risk (i.e., the benchmark dose, or BMD) and its
95% lower confidence bound (BMDL) were estimated43. In most cases, the risk level, or
benchmark response (BMR), was 10% extra risk44; however, in a few cases with low response
rates, a BMR of 5%, or even 1%, extra risk was used to avoid extrapolation above the range of
the data. As discussed in Section 4.3, there is sufficient evidence to conclude that a mutagenic
MOA is operative for TCE-induced kidney tumors, so linear extrapolation from the BMDL to
the origin was used to derive unit risk estimates (or "slope factors" for oral exposures) for this
site. For all other tumor types, the available evidence supports the conclusion that the MOA(s)
for TCE-induced rodent tumors is unknown, as discussed in Sections 4.4 - 4.9 and summarized
in Section 4.10.2.3. Therefore, linear extrapolation was also used based on the general principles
outlined in EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) and reviewed
below in Section 5.2.1.4.1. Thus, for all TCE-associated rodent tumors, unit risk estimates are
equal to BMR/BMDL, e.g., 0.10/BMDLio for a BMR of 10%). See Section 5.2.1.3 for a
summary of the unit risk estimates for each sex/species/bioassay/tumor type.
Some of the bioassays exhibited differential early mortality across the dose groups, and,
for three such male rat studies (identified with checkmarks in the "Time-to-tumor" column of
42 Using a standard criterion for nested models, that the difference in -2*log-likelihood exceeds 3.84 (the 95th
percentile of x2 (1)).
43 BMDS estimates confidence intervals using the profile likelihood method.
44 Extra risk over the background tumor rate is defined as [P(d) - P(0)]/[1 - P(0)], where P(d) represents the lifetime
risk (probability) of cancer at dose d.
6/22/2009
1113
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.2.3), analyses which take individual animal survival times into account were performed.
(For bioassays with differential early mortality occurring primarily before the time of the 1st
tumor [or 52 weeks, whichever came first], the effects of early mortality were largely accounted
for by adjusting the tumor incidence for animals at risk, as described in Appendix G, and the
dose-response data were modeled using the regular multistage model, as discussed above, rather
than approaches which account for individual animal survival times.) Two approaches were
used to take individual survival times into account. First, EPA's Multistage Weibull (MSW)
software45 was used for time-to-tumor modeling. The multistage Weibull time-to-tumor model
has the general form
P(d,t) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk) * (t - to)z],
where P(d,t) represents the probability of a tumor by age t for dose d, and parameters z >1, to > 0,
and q; > 0 for i = 0,1,...,k, where k = the number of dose groups; the parameter to represents the
time between when a potentially fatal tumor becomes observable and when it causes death. (All
of our analyses used the model for incidental tumors, which has no to term.) Although the fit of
the MSW model can be assessed visually using the plot feature of the MSW software, because
there is no applicable goodness-of-fit statistic with a well-defined asymptotic distribution, an
alternative survival-adjustment technique, "poly-3 adjustment", was also applied (Portier and
Bailer, 1989). This technique was used to adjust the tumor incidence denominators based on the
individual animal survival times.46 The adjusted incidence data then served as inputs for EPA's
BMDS multistage model, and model (i.e., stage) selection was conducted as already described
above. Under both survival-adjustment approaches, BMDs and BMDLs were obtained and unit
risks derived as discussed above for the standard multistage model approach. See Appendix G
for a more detailed description of the MSW modeling and for the results of both the MSW and
poly-3 approaches for the individual datasets. A comparison of the results for the three different
datasets and the various dose metrics used is presented in Section 5.2.1.3.
For bioassays that exhibited more than one type of tumor response in the same sex and
species (these studies have a row for "combined risk" in the "Endpoint" column of Table 5.2.3),
the cancer potency for the different tumor types combined was estimated. The combined tumor
risk estimate describes the risk of developing tumors for any (not all together) of the tumor types
45 This software has been thoroughly tested and externally reviewed. In February 2009 it will become available on
EPA's website.
46 Each tumorless animal is weighted by its fractional survival time (number of days on study divided by 728 days,
the typical number of days in a 2-year bioassay) raised to the power of 3 to reflect the fact that animals are at greater
risk of cancer at older ages. Animals with tumors are given a weight of 1. The sum of the weights of all the animals
in an exposure group yields the effective survival-adjusted denominator.
6/22/2009
1114
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
that exhibited a TCE-associated tumor response; this estimate then represents the total excess
cancer risk. The model for the combined tumor risk is also multistage, with the sum of the stage-
specific multistage coefficients from the individual tumor models serving as the stage-specific
coefficients for the combined risk model (i.e., for each qH qi[COmbined] = qii + qi2 + ••• + qik, where
the q;s are the coefficients for the powers of dose and k is the number of tumor types being
combined) (Bogen, 1990; NRC, 1994). This model assumes that the occurrences of two or more
tumor types are independent. Although the resulting model equation can be readily solved for a
given BMR to obtain an MLE (BMD) for the combined risk, the confidence bounds for the
combined risk estimate are not calculated by available modeling software. Therefore, the
confidence bounds on the combined BMD were estimated using a Bayesian approach, computed
using Markov chain Monte Carlo techniques and implemented using the freely available
WinBugs software (Spiegelhalter et al., 2003). Use of WinBugs for derivation of a distribution
of BMDs for a single multistage model has been demonstrated by Kopylev et al. (2007), and this
approach can be straightforwardly generalized to derive the distribution of BMDs for the
combined tumor load. For further details on the implementation of this approach and for the
results of the analyses, see Appendix G.
5.2.1.2 Rodent Dose-Response Analyses: Dosimetry
In modeling the applied doses (or exposures), default dosimetry procedures were applied
to convert applied rodent doses to human equivalent doses. Essentially, for inhalation exposures,
"ppm equivalence" across species was assumed. For oral doses, 3/4-power body-weight scaling
was used, with a default average human body weight of 70 kg. See Appendix G for more details
on the default dosimetry procedures.
In addition to applied doses, several internal dose metrics were used in the dose-response
modeling for each tumor type. Use of internal dose metrics in dose-response modeling is
described here briefly. For more details on the PBPK modeling used to estimate the levels of the
dose metrics corresponding to different exposure scenarios in rodents and humans, see Section
3.5; for a more detailed discussion of how the dose metrics were used in dose-response
modeling, see Appendix G.
5.2.1.2.1 Selection of dose metrics for different tumor types
One area of scientific uncertainty in cancer dose-response assessment is the appropriate
scaling between rodent and human doses for equivalent responses. As discussed above, for
applied dose, the standard dosimetry assumptions for equal lifetime carcinogenic risk are, for
inhalation exposure, the same lifetime exposure concentration in air, and, for oral exposure, the
same lifetime daily dose scaled by body weight to the 3/4 power. For scaling internal doses, it is
useful to consider two possible interpretations of these standard dosimetry assumptions. The
6/22/2009
1115
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
first (denoted "empirical dosimetry") interpretation is that standard dosimetry is based on the
empirical finding that scaling the delivered dose rate by body weight to the 3/4 power results in
equivalent toxicity (e.g., Travis and White, 1988; USEPA, 1992). This is supported biologically
by data showing that rates of both kinetic and dynamic physiologic processes are generally
consistent with 3/4 power of body weight scaling across species (USEPA 1992). Note also that
this applies to inhalation exposure because the delivered dose rate in that case is the air
concentration multiplied by the ventilation rate, which scales by body weight to the 3/4 power.
Applying this interpretation to internal doses would imply that the dose rate of the active moiety
delivered to the target tissue, scaled by body weight to the 3/4 power, would be assumed to result
in equivalent responses. The second (denoted "concentration equivalence dosimetry")
interpretation hypothesizes that the empirical finding is pharmacokinetically-driven, due to the
body weight to the 3/4 scaling of physiologic flows (cardiac output, ventilation rate, glomerular
filtration, etc.) and metabolic rates (enzyme-mediated biotransformation). Therefore, the
standard dosimetry assumptions yield equivalent average internal concentrations, which in turn
yield equivalent carcinogenic risk (NRC, 1986; NRC, 1987). Applying this dosimetry
interpretation to internal doses would imply that equivalent carcinogenic risk should be based on
equal (average) concentrations of the active moiety or moieties at the target tissue.
To the extent that production and clearance of the active moiety or moieties all scale by
body weight to the 3/4 power, these two dosimetry interpretations both lead to the same
quantitative results. However, these interpretations may lead to different quantitative results
when there are deviations of the underlying physiologic or metabolic processes from body
weight to the 3/4 power scaling. For instance, as discussed in Section 3.5, the PBPK model
predictions for AUC of TCE in blood deviate from the body weight to the 3/4 scaling (the scaling
is closer to mg/kg/d than mg/kgy4/d), so use of this dose metric when TCE is the active moiety
implicitly assumes the "concentration equivalence dosimetry." In addition, as discussed below,
in most cases involving TCE metabolites, only the rate of production of the active moiety(ies) or
the rate of transformation through a particular metabolic pathway can be estimated using the
PBPK model, and the actual concentration of the active moiety(ies) cannot be estimated due to
data limitations. Under "empirical dosimetry," these metabolism rates, which are estimates of
the systemic or tissue-specific delivery of the active moiety(ies), would be scaled by body weight
to the 3/4 power to yield equivalent carcinogenic risk. Under "concentration equivalence
dosimetry," additional assumptions about the rate of clearance are necessary to specify the
scaling that would yield concentration equivalence. In the absence of data, active metabolites are
assumed to be sufficiently stable so that clearance is via enzyme-catalyzed transformation or
systemic excretion (e.g., blood flow, glomerular filtration), which scale approximately by body
6/22/2009
1116
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
weight to the 3/4 power. Therefore, under "concentration equivalence dosimetry," the metabolism
rates would also be scaled by body weight to the 3/4 power in the absence of additional data.
For toxicity that is associated with local (in situ) production of "reactive" metabolites
whose concentrations cannot be directly measured in the target tissue, an alternative approach,
under "concentration equivalence dosimetry," of scaling by unit tissue mass has been proposed
(e.g., Andersen et al. 1987). As discussed by Travis (1990), in this situation, scaling the rate of
local metabolism across species and individuals by tissue mass is appropriate if the metabolites
are sufficiently reactive and are cleared by "spontaneous" deactivation (i.e., changes in chemical
structure without the need of biological influences). Thus, use of this alternative scaling
approach requires that (i) the active moiety or moieties do not leave the target tissue in
appreciable quantities (i.e., are cleared primarily by in situ transformation to other chemical
species and/or binding to/reactions with cellular components); and (ii) the clearance of the active
moieties from the target tissue is governed by biochemical reactions whose rates are independent
of body weight (e.g., purely chemical reactions). If these conditions are met, then under the
"concentration equivalence dosimetry," the relevant metabolism rates estimated by the PBPK
model would be scaled by tissue mass, rather than by body weight to the 3/4 power.
To summarize, the appropriate internal dose metric for equivalent carcinogenic responses
can be specified by invoking one of two alternative interpretations of the standard dosimetry for
applied dose: "empirical dosimetry" based on the rate at which the active moiety(ies) is(are)
delivered to the target tissue scaled by body weight to the 3/4 power or "concentration equivalence
dosimetry" based on matching internal concentrations of the active moiety(ies) in the target
tissue. If the active moiety(ies) is TCE itself or a putatively reactive metabolite, the choice of
interpretation will affect the choice of internal dose metric. In the discussions of dose metric
selections for the individual tumors sites below, the implications of both "empirical dosimetry"
and "concentration equivalence dosimetry" are discussed. Additionally, an attempt was made to
use tissue-specific dose metrics representing particular pathways or metabolites identified from
available data as having a likely role in the induction of a tissue-specific cancer. Where
insufficient information was available to establish particular metabolites or pathways of likely
relevance to a tissue-specific cancer, more general "upstream" metrics representing either parent
compound or total metabolism had to be used. In addition, the selection of dose metrics was
limited to metrics that could be adequately estimated by the PBPK model (see Section 3.5). The
(PBPK-based) dose metrics used for the different tumor types are listed in Table 5.2.3. For each
tumor type, the "primary" dose metric referred to in Table 5.2.3 is the metric representing the
particular metabolite or pathway whose involvement in carcinogenicity has the greatest
biological support, whereas "alternative" dose metrics represent upstream metabolic pathways
(or TCE distribution, in the case of AUCCBld) which may be more generally involved.
6/22/2009
1117
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.2.1.2.1.1 Kidney
As discussed in Sections 4.3.6 - 4.3.7, there is sufficient evidence to conclude that TCE-
induced kidney tumors in rats are primarily caused by GSH-conjugation metabolites either
produced in situ in or delivered systemically to the kidney. As discussed in Section 3.3.3.2,
bioactivation of these metabolites within the kidney, either by beta-lyase, FMO, or P450s,
produces reactive species. Therefore, multiple lines of evidence support the conclusion that
renal bioactivation of DCVC is the preferred basis for internal dose extrapolations of TCE-
induced kidney tumors. However, uncertainties remain as to the relative contributions from each
bioactivation pathway, and quantitative clearance data necessary to calculate the concentration of
each species are lacking.
Under "empirical dosimetry," the rate of renal bioactivation of DCVC would be scaled by
body weight to the 3/4 power. As discussed above, under "concentration equivalence dosimetry,"
when the concentration of the active moiety cannot be estimated, qualitative data on the nature of
clearance of the active moiety or moieties can be used to inform whether to scale the rate of
metabolism by body weight to the 3/4 power or by the target tissue weight. For the beta-lyase
pathway, Dekant et al. (1988) reported in trapping experiments that the postulated reactive
metabolites decompose to stable (unreactive) metabolites in the presence of water. Moreover,
the necessity of a chemical trapping mechanism to detect the reactive metabolites suggests a very
rapid reaction such that it is unlikely that the reactive metabolites leave the site of production.
Therefore, these data support the conclusion that, for this bioactivation pathway, clearance is
chemical in nature and hence species-independent. If this were the only bioactivation pathway,
then the scaling by kidney weight would be supported. With respect to the FMO bioactivation
pathway, Sausen and Elfarra (1991) reported that after direct dosing of the postulated reactive
sulfoxide, the sulfoxide was detected as an excretion product in bile. These data suggest that
reactivity in the tissue to which the sulfoxide was delivered (the liver, in this case) is insufficient
to rule out a significant role for enzymatic or systemic clearance. Therefore, according to the
criteria outlined above, for this bioactivation pathway, the data support scaling the rate of
metabolism by body weight to the 3/4 power. For P450-mediated bioactivation producing N-
acetyl DCVC (mercapturic acid) sulfoxide, the only relevant data on clearance are from a study
of the structural analogue to DCVC, FDVE (Sheffels et al. 2004), which reported that the
postulated reactive sulfoxide was detected in urine. This suggests that the sulfoxide is
sufficiently stable to be excreted by the kidney and supports the scaling of the rate of metabolism
by body weight to the % power.
Therefore, because the contributions to TCE-induced nephrocarcinogenicity from each
possible bioactivation pathway are not clear, and, even under "concentration equivalence
dosimetry," the scaling by body weight to the % power is supported for two of the three
6/22/2009
1118
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
bioactivation pathways, it is decided here to scale the DCVC bioactivation rate by body weight
to the 3/4 power. The primary internal dose metric for TCE-induced kidney tumors is thus the
weekly rate of DCVC bioactivation per unit body weight to the 3/4 power (ABioactDCVCBW34
[mg/kgyYwk]). However, it should be noted that due to the larger relative kidney weight in rats
as compared to humans, scaling by kidney weight instead of body weight to the 3/4 power would
only change the quantitative inter-species extrapolation by about 2-fold,47 so the sensitivity of
the results to the scaling choice is relatively small.
To summarize, under the "empirical dosimetry" approach, the underlying assumption for
the ABioactDCVCBW34 dose metric is that equalizing the rate of renal bioactivation of DCVC
(i.e., local production of active moiety(ies) in the target tissue), scaled by the 3/4 power of body
weight, yields equivalent lifetime cancer risk across species. Under "concentration equivalence
dosimetry," the underlying assumptions for the ABioactDCVCBW34 dose metric are that (i) the
same average concentration of reactive species produced from DCVC in the kidney leads to a
similar lifetime cancer risk across species; and (ii) the rate of clearance of these reactive species
scales by the 3/4 power of body weight (e.g., assumed for enzyme-activity or blood-flow).
An alternative dose metric that also involves the GSH conjugation pathway is the amount
of GSH conjugation scaled by the 3/4 power of body weight (AMetGSHBW34 [mg/kgyYwk]).
This dose metric uses the total flux of GSH conjugation as the toxicologically-relevant dose, and
thus incorporates any direct contributions from DCVG and DCVC, which are not addressed in
the DCVC bioactivation metric. Under the "empirical dosimetry" approach, the underlying
assumption for the AMetGSHBW34 dose metric is that equalizing the (whole body) rate of
production of GSH conjugation metabolites (i.e., systemic production of active moiety(ies)),
scaled by the 3/4 power of body weight, yields equivalent lifetime cancer risk across species.
Under "concentration equivalence dosimetry," the AMetGSHBW34 dose metric is consistent
with the assumptions that (i) the same average concentration of the (relatively) stable upstream
metabolites DCVG and (subsequently) DCVC in the kidney (the PBPK model assumes all
DCVG and DCVC produced translocates to the kidney) leads to the same lifetime cancer risk
across species; and (ii) the rates of clearance of DCVG and (subsequently) DCVC scale by the 3/4
power of body weight (as is assumed for enzyme activity or blood flow).
Another alternative dose metric is the total amount of TCE metabolism (oxidation and
GSH conjugation together) scaled by the 3/4 power of body weight (TotMetabBW34
[mg/kgyYwk]). This dose metric uses the total flux of TCE metabolism as the toxicologically-
relevant dose, and thus incorporates the possible involvement of oxidative metabolites, acting
47 The range of the difference is 2.1-2.4-fold using the posterior medians for the relative kidney weight in rats and
humans from the PBPK model described in Section 3.5 (Table 3.5.7) and body weights of 0.3-0.4 kg for rats and
60-70 kg for humans.
6/22/2009
1119
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
either additively or interactively, in addition to GSH conjugation metabolites in
nephrocarcinogenicity (see Section 4.3.6). While there is no evidence that TCE oxidative
metabolites can on their own induce kidney cancer, some nephrotoxic effects attributable to
oxidative metabolites (e.g., peroxisome proliferation) may modulate the nephrocarcinogenic
potency of GSH metabolites. However, this dose metric is given less weight than those
involving GSH conjugation because, as discussed in Sections 4.3.6 and 4.3.7, the weight of
evidence supports the conclusion that GSH conjugation metabolites play a predominant role in
nephrocarciongenicity. Under the "empirical dosimetry" approach, the underlying assumption
for the TotMetabBW34 dose metric is that equalizing the (whole body) rate of production of all
metabolites (i.e., systemic production and distribution of active moiety(ies)), scaled by the 3/4
power of body weight, yields equivalent lifetime cancer risk across species. Under
"concentration equivalence dosimetry," the TotMetabBW34 dose metric is consistent with the
assumptions that (i) the relative proportions and blood:tissue partitioning of the active
metabolites is similar across species; (ii) the same average concentration of one or more active
metabolites in the kidney leads to a similar lifetime cancer risk across species; and (iii) the rates
of clearance of active metabolites scale by the 3/4 power of body weight (e.g., as is assumed for
enzyme activity or blood flow).
5.2.1.2.1.2 Liver
As discussed in Section 4.4.6, there is substantial evidence that oxidative metabolism is
involved in TCE hepatocarcinogenicity, based primarily on non-cancer and cancer effects similar
to those observed with TCE being observed with a number of oxidative metabolites of TCE (e.g.,
CH, TCA, and DCA). While TCA is a stable, circulating metabolite, CH and DCA are relatively
short-lived, although enzymatically cleared (see Section 3.3.3.1). As discussed in Sections 4.4.6
and 4.4.7, there is now substantial evidence that TCA does not adequately account for the
hepatocarcinogenicity of TCE; therefore, unlike in previous dose-response analyses (Rhomberg
2000, Clewell and Andersen 2004), the area-under-the-curve (AUC) of TCA in plasma and in
liver were not considered as dose metrics. However, there are inadequate data across species to
quantify the dosimetry of CH and DCA, and other intermediates of oxidative metabolism (such
as TCE-oxide or dichloroacetylchloride) also may be involved in carcinogenicity. Thus, due to
uncertainties as to the active moiety(ies), but the strong evidence associating TCE liver effects
with oxidative metabolism in the liver, hepatic oxidative metabolism is the preferred basis for
internal dose extrapolations of TCE-induced liver tumors. Under "empirical dosimetry," the rate
of hepatic oxidative metabolism would be scaled by body weight to the 3/4 power. As discussed
above, under "concentration equivalence dosimetry," when the concentration of the active
moiety cannot be estimated, qualitative data on the nature of clearance of the active moiety or
6/22/2009
1120
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
moieties can be used to inform whether to scale the rate of metabolism by body weight to the 3/4
power or by the target tissue weight. However, several of the oxidative metabolites are stable
and systemically available, and several of those that are cleared rapidly are metabolized
enzymatically, so, according to the criteria discussed above, there are insufficient data to support
the conclusions that the active moiety or moieties do not leave the target tissue in appreciable
quantities and are cleared by mechanisms whose rates are independent of body weight. Thus,
scaling the rate of oxidative metabolism by body weight to the 3/4 power would also be supported
under "concentration equivalence dosimetry." Therefore, the primary internal dose metric for
TCE-induced liver tumors is selected to be the weekly rate of hepatic oxidation per unit body
weight to the 3/4 power (AMetLivlBW34 [mg/kgyYwk]). It should be noted that due to the
larger relative liver weight in mice as compared to humans, scaling by liver weight instead of
body weight to the 3/4 power would only change the quantitative inter-species extrapolation by
about 4-fold,48 so the sensitivity of the results to the scaling choice is relatively modest.
To summarize, under the "empirical dosimetry" approach, the underlying assumption for
the AMetLivlBW34 dose metric is that equalizing the rate of hepatic oxidation of TCE (i.e.,
local production of active moiety(ies) in the target tissue), scaled by the 3/4 power of body weight,
yields equivalent lifetime cancer risk across species. Under "concentration equivalence
dosimetry," the AMetLivlBW34 dose metric is consistent with the assumptions that (i) the same
average concentrations of the active oxidative metabolites in the liver leads to a similar lifetime
cancer risk across species; (ii) active metabolites are primarily generated in situ in the liver; (iii)
the relative proportions of the active oxidative metabolites are similar across species; and (iv) the
rates of clearance of the active oxidative metabolites scale by the 3/4 power of body weight (e.g.,
enzyme-activity or blood-flow).
It is also known that the lung has substantial capacity for oxidative metabolism, with
some proportion of the oxidative metabolites produced there entering systemic circulation. Thus,
it is possible that extra-hepatic oxidative metabolism can contribute to TCE
hepatocarcinogenicity. Therefore, the total amount of oxidative metabolism of TCE scaled by
the 3/4 power of body weight (TotOxMetabBW34 [mg/kgyYwk]) was selected as an alternative
dose metric (the justification for the body weight to the 3/4 power scaling is analogous to that for
hepatic oxidative metabolism, above). Under the "empirical dosimetry" approach, the
underlying assumption for the TotOxMetabBW34 dose metric is that equalizing the rate of total
oxidation of TCE (i.e., systemic production of active moiety(ies)), scaled by the 3/4 power of
body weight, yields equivalent lifetime cancer risk across species. Under "concentration
48 The range of the difference is 3.5-3.9-fold using the posterior medians for the relative liver weight in mice and
humans from the PBPK model described in Section 3.5 (Table 3.5.7), and body weights of 0.03-0.04 kg for mice
and 60-70 kg for humans.
6/22/2009
1121
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
equivalence dosimetry," this dose metric is consistent with the assumptions that (i) active
metabolites may be generated in situ in the liver or delivered to the liver via systemic circulation;
(ii) the relative proportions and blood:tissue partitioning of the active oxidative metabolites are
similar across species; (iii) the same average concentrations of the active oxidative metabolites in
the liver leads to a similar lifetime cancer risk across species; and (iv) the rates of clearance of
the active oxidative metabolites scale by the 3/4 power of body weight (e.g., as is assumed for
enzyme activity or blood flow).
5.2.1.2.1.3 Lung
As discussed in Section 4.6.3, in situ oxidative metabolism in the respiratory tract may be
more important to lung toxicity than systemically delivered metabolites, at least as evidenced by
acute pulmonary toxicity. While chloral was originally implicated as the active metabolite,
based on either acute toxicity or mutagenicity of chloral and/or chloral hydrate, more recent
evidence suggests that other oxidative metabolites may also contribute to lung toxicity. These
data include the identification of DAL adducts in Clara cells (Forkert et al. 2006), and the
induction of pulmonary toxicity by TCE in CYP2El-null mice, which may generate a different
spectrum of oxidative metabolites as compared to wild-type mice (respiratory tract tissue also
contains P450s from the CYP2F family). Overall, the weight of evidence supports the selection
of respiratory tract oxidation of TCE as the preferred basis for internal dose extrapolations of
TCE-induced lung tumors. However, uncertainties remain as to the relative contributions from
different oxidative metabolites, and quantitative clearance data necessary to calculate the
concentration of each species are lacking.
Under "empirical dosimetry," the rate of respiratory tract oxidation would be scaled by
body weight to the 3/4 power. As discussed above, under "concentration equivalence dosimetry,"
when the concentration of the active moiety cannot be estimated, qualitative data on the nature of
clearance of the active moiety or moieties can be used to inform whether to scale the rate of
metabolism by body weight to the 3/4 power or by the target tissue weight. For chloral, as
discussed in Section 4.6.3, the reporting of substantial TCOH but no detectable chloral hydrate in
blood following TCE exposure from experiments in isolated perfused lungs (Dalby and
Bingham, 1978) support the conclusion that chloral does not leave the target tissue in substantial
quantities, but that there is substantial clearance by enzyme-mediated biotransformation.
Dichloroacetyl chloride is a relatively-short-lived intermediate from aqueous (non-enzymatic)
decomposition of TCE-oxide that can be trapped with lysine or degrade further to form DCA,
among other products (Cai and Guengerich, 1999). Cai and Guengerich (1999) reported a half-
life of TCE-oxide under aqueous conditions of 12 s at 23 C, a time-scale that would be shorter at
physiological conditions (37 C) and that includes formation of dichloroacetyl chloride as well as
6/22/2009
1122
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
its decomposition. Therefore, evidence for this metabolite suggests its clearance both is
sufficiently rapid so that it would remain at the site of formation and is non-enzymatically
mediated so that its rate would be independent of body weight. Other oxidative metabolites may
also play a role, but, because they have not been identified, no inferences can be made as to their
clearance.
Therefore, because it is not clear what the contributions to TCE-induced lung tumors are
from different oxidative metabolites produced in situ and, even under "concentration equivalence
dosimetry," the scaling by body weight to the 3/4 power is supported for at least one of the
possible active moieties, it was decided here to scale the rate of respiratory tract tissue oxidation
of TCE by body weight to the 3/4 power. The primary internal dose metric for TCE-induced lung
tumors is thus the weekly rate of respiratory tract oxidation per unit body weight to the 3/4 power
(AMetLngBW34 [mg/kgyYwk]). It should be noted that, due to the larger relative respiratory
tract tissue weight in mice as compared to humans, scaling by tissue weight instead of body
weight to the 3/4 power would change the quantitative inter-species extrapolation by less than 2-
fold,49 so the sensitivity of the results to the scaling choice is relatively small.
To summarize, under the "empirical dosimetry" approach, the underlying assumption for
the AMetLngBW34 dose metric is that equalizing the rate of respiratory tract oxidation of TCE
(i.e., local production of active moiety(ies) in the target tissue), scaled by the 3/4 power of body
weight, yields equivalent lifetime cancer risk across species. Under "concentration equivalence
dosimetry," the use of the AMetLngBW34 dose metric is consistent with the assumptions that (i)
the proportion of respiratory tract oxidative metabolism to active metabolites are similar across
species (ii) the same average concentration of the active moiety(ies) in the metabolizing
respiratory tract tissue leads to a similar lifetime cancer risk across species; and (iii) the rates of
clearance of these reactive species scale by the 3/4 power of body weight (e.g., enzyme-activity or
blood-flow).
While there is substantial evidence that acute pulmonary toxicity is related to pulmonary
oxidative metabolism, for carcinogenicity, it is possible that, in addition to locally produced
metabolites, systemically-delivered oxidative metabolites also play a role. Therefore, total
oxidative metabolism scaled by the 3/4 power of body weight (TotOxMetabBW34 [mg/kgy7wk])
was selected as an alternative dose metric (the justification for the body weight to the 3/4 power
scaling is analogous to that for respiratory tract oxidative metabolism, above). Under the
"empirical dosimetry" approach, the underlying assumption for the TotOxMetabBW34 dose
metric is that equalizing the rate of total oxidation of TCE (i.e., systemic production of oxidative
49 The range of the difference is 1.6-1.8-fold using the posterior medians for the relative respiratory tract tissue
weight in mice and humans from the PBPK model described in Section 3.5 (Table 3.5.7), and body weights of 0.03-
0.04 kg for mice and 60-70 kg for humans.
6/22/2009
1123
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
metabolites), scaled by the 3/4 power of body weight, yields equivalent lifetime cancer risk across
species. Under "concentration equivalence dosimetry," this dose metric is consistent with the
assumptions that (i) active oxidative metabolites may be generated in situ in the lung or delivered
to the lung via systemic circulation; (ii) the relative proportions and blood:tissue partitioning of
the active oxidative metabolites are similar across species; (iii) the same average concentrations
of the active oxidative metabolites in the lung leads to a similar lifetime cancer risk across
species; and (iv) the rates of clearance of the active oxidative metabolites scale by the 3/4 power
of body weight (e.g., as is assumed for enzyme activity or blood flow).
Another alternative dose metric considered here is the AUC of TCE in blood (AUCCBld
[mg h/l/wk]). Under either the "empirical dosimetry" or "concentration equivalence" approach,
this dose metric would account for the possibility that local metabolism is determined primarily
by TCE delivered in blood via systemic circulation to pulmonary tissue (the flow rate of which
scales as body weight to the 3/4 power), as assumed in previous PBPK models, rather than TCE
delivered in air via diffusion to the respiratory tract, as is assumed in the PBPK model described
in Section 3.5. However, as discussed in Section 3.5 and Appendix A, the available
pharmacokinetic data provide greater support for the updated model structure. Under
"concentration equivalence dosimetry," this dose metric also accounts for the possible role of
TCE itself in pulmonary carcinogenicity (consistent with the assumption that the same average
concentration of TCE in blood will lead to a similar lifetime cancer risk across species).
5.2.1.2.1.4 Other sites
For all other sites listed in Table 5.2.3, there is insufficient information for site-specific
determinations of appropriate dose metrics. While TCE metabolites and/or metabolizing
enzymes have been reported in some of these tissues (e.g., male reproductive tract), their roles in
carcinogenicity for these specific sites have not been established. Although "primary" and
"alternative" dose metrics are defined, they do not differ appreciably in their degrees of
plausibility.
Given that the majority of the toxic and carcinogenic responses to TCE appear to be
associated with metabolism, total metabolism of TCE scaled by the 3/4 power of body weight was
selected as the primary dose metric (TotMetabBW34 [mg/kgy7wk]). This dose metric uses the
total flux of TCE metabolism as the toxicologically-relevant dose, and thus incorporates the
possible involvement of any TCE metabolite in carcinogenicity. Under the "empirical
dosimetry" approach, the underlying assumption for the TotMetabBW34 dose metric is that
equalizing the (whole body) rate of production of all metabolites (i.e., systemic production of
active moiety(ies)), scaled by the 3/4 power of body weight, yields equivalent lifetime cancer risk
across species. Under "concentration equivalence dosimetry," the TotMetabBW34 dose metric
6/22/2009
1124
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
is consistent with the assumptions that (i) active metabolites are delivered to the target tissue via
systemic circulation; (ii) the relative proportions and blood:tissue partitioning of the active
metabolites is similar across species; (iii) the same average concentrations of the active
metabolites in the target tissue leads to a similar lifetime cancer risk across species; and (iv) the
rates of clearance of the active metabolites scale by the 3/4 power of body weight (e.g., as is
assumed for enzyme activity or blood flow).
An alternative dose metric considered here is the AUC of TCE in blood. Under either the
"empirical dosimetry" or "concentration equivalence" approach, this dose metric would account
for the possibility that the determinant of carcinogenicity is local metabolism, governed
primarily by TCE delivered in blood via systemic circulation to the target tissue (the flow rate of
which scales as body weight to the 3/4 power). Under "concentration equivalence dosimetry,"
this dose metric also accounts for the possible role of TCE itself in carcinogenicity (consistent
with the assumption that the same average concentration of TCE in blood will lead to a similar
lifetime cancer risk across species).
5.2.1.2.2 Methods for dose-response analyses using internal dose metrics
As shown in Figure 5.2.1, the general approach taken for the use of internal dose metrics
in dose-response modeling was to first apply the rodent PBPK model to obtain rodent values for
the dose metrics corresponding to the applied doses in a bioassay. Then, dose-response
modeling for a tumor response was performed using the internal dose metrics and the multistage
model or the survival-adjusted modeling approaches described above to obtain a BMD and
BMDL in terms of the dose metric. On an internal dose basis, humans and rodents are presumed
to have similar lifetime cancer risks, and the relationship between human internal and external
doses is essentially linear at low doses up to 0.1 mg/kg/d or 0.1 ppm, and nearly linear up to 10
mg/kg/d or 10 ppm. Therefore, the BMD and BMDL were then converted human equivalent
doses (or exposures) using conversion ratios estimated from the human PBPK model at 0.001
mg/kg/d or 0.001 ppm (Table 5.2.4). Because the male and female conversions differed by less
than 11%, the human BMDLs were derived using the mean of the sex-specific conversion factors
(except for testicular tumors, for which only male conversion factors were used). Finally, a unit
risk estimate for that tumor response was derived from the human "BMDLs" as described above
(i.e., BMR/BMDL). Note that the converted "BMDs" and "BMDLs" are not actually human
equivalent BMDs and BMDLs corresponding to the BMR because the conversion was not made
in the dose range of the BMD; the converted BMDs and BMDLs are merely intermediaries to
obtain a converted unit risk estimate. In addition, it should be noted that median values of dose
metrics were used for rodents, whereas mean values were used for humans. Because the rodent
population model characterizes study-to-study variation, animals of the same sex/species/strain
combination within a study were assumed to be identical. Therefore, use of median dose metric
6/22/2009
1125
-------
1
2
3
4
5
6
7
8
9
10
11
12
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
values for rodents can be interpreted as assuming that the animals in the bioassay were all
"typical" animals and the dose-response model is estimating a "risk to the typical rodent." In
practice, the use of median or mean internal doses for rodents did not make much difference
except when the uncertainty in the dose metric was high (e.g., AMetLungBW34 dose metric in
mice). A quantitative analysis of the impact of the uncertainty in the rodent PBPK dose metrics
is included in Section 5.2.1.4.2. On the other hand, the human population model characterizes
individual-to-individual variation. Because the quantity of interest is the human population
mean risk, the expected value (averaging over the uncertainty) of the population mean (averaging
over the variability) dose metric was used for the conversion to human unit risks. Therefore, the
extrapolated unit risk estimates can be interpreted as the expected "average risk" across the
population based on rodent bioassays.
6/22/2009
1126
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Bioassay
experimental
paradigm
istribution
Rodent
model
parameters
PBPK
model
Rodent
internal
dose
edian
esponse
Model
.distribution (combined
ncertainty and variability)
0.001 ppm
in air or
0.001 mg/kg-d
continuous
exposure
fixed'.
Human
model
parameters
PBPK
model
istribution
.distribution (separate
ncertainty and variability)
BMD, BMDL
(internal
dose unit)
Site-specific
cancer unit
risk = BMR/
BMDL
(per internal
dose unit)
Human
internal
dose
population
Expected
average
human
internal dose
per ppm or
per mg/kg-d
1
2
3
4
5
6
Human site-
specific cancer
unit risk
(per ppm or
per mg/kg-d)
Figure 5.2.1.
Flow-chart for dose-response analyses of rodent bioassays using PBPK model-based dose
metrics. Square nodes indicate point values, circular nodes indicate distributions, and the
inverted triangles indicate a (deterministic) functional relationship.
6/22/2009
1127
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1
2 Table 5.2.3. Specific dose-response analyses performed and dose metrics used
Inhalation Bioassay Strain Endpoint Applied PBPK-based - PBPK-based - Time-
dose primary dose alternative dose to-
metric metric(s) tumor
Female mice
Fukuda et al. (1983)
Cij:CD-l
Lung adenomas and carcinomas
V
AMetLngBW 3 4
T otOxMetabBW 3 4
(ICR)
AUCCBld
Henschler et al. (1980)
Han:NMRI
Lymphoma
V
TotMetabBW34
AUCCBld
Maltoni et al. (1986)
B6C3F1
Liver hepatomas
V
AMetLivlBW34
T otOxMetabBW 3 4
Lung adenomas and carcinomas
V
AMetLngBW 3 4
T otOxMetabBW 3 4
AUCCBld
Combined risk
V
Male mice
Maltoni et al. (1986)
Swiss
Liver hepatomas
V
AMetLivlBW34
T otOxMetabBW 3 4
Female rats
None selected
Male rats
Maltoni et al. (1986)
Sprague-
Kidney adenomas and carcinomas
V
ABioactDCVCBW34
AMetGSHBW34
Dawley
TotMetabBW34
Leydig cell tumors
V
TotMetabBW34
AUCCBld
Leukemias
V
TotMetabBW34
AUCCBld
Combined risk
V
3
6/22/2009
1128
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Oral Bioassay
Strain
Endpoint
Applied
PBPK-based -
PBPK-based -
Time-
dose
primary dose
alternative dose
to-
metric
metric(s)
tumor
Female mice
NCI (1976)
B6C3F1
Liver carcinomas
V
AMetLivlBW34
T otOxMetabBW 3 4
Lung adenomas and carcinomas
V
AMetLngB W 3 4
T otOxMetabB W 3 4
AUCCBld
Multiple sarcomas/lymphomas
V
TotMetabBW34
AUCCBld
Combined risk
V
Male mice
NCI (1976)
B6C3F1
Liver carcinomas
V
AMetLivlBW34
T otOxMetabBW 3 4
Female rats
NTP (1988)
August
Leukemia
V
TotMetabBW34
AUCCBld
Male rats
NTP (1988)
August
Subcutaneous tissue sarcomas
V
TotMetabBW34
AUCCBld
NTP (1988)
Marshall
Testicular interstitial cell tumors
V
TotMetabBW34
AUCCBld
V
NTP (1988)
Osborne-
Kidney adenomas and carcinomas
V
ABioactDCVCBW34
AMetGSHBW34
V
Mendel
TotMetabBW34
NTP(1990)
F344/N
Kidney adenomas and carcinomas
V
ABioactDCVCBW34
AMetGSHBW34
V
TotMetabBW34
6/22/2009
1129
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 PBPK-based dose metric abbreviations:
2 ABioactDCVCBW34 = Amount of DCVC bioactivated in the kidney per unit body weight74 (mg DCVC/kgyYwk)
3 AMetGSHBW34 = Amount of TCE conjugated with GSH per unit body weight74 (mg TCE/kgyYwk)
4 AMetLivlBW34 = Amount of TCE oxidized per unit body weight74 (mg TCE/kgyYwk)
5 AMetLngBW34 = Amount of TCE oxidized in the respiratory tract per unit body weight74 (mg TCE/kgyYwk)
6 AUCCBld = Area under the curve of the venous blood concentration of TCE (mg hr/L/wk)
7 TotMetabBW34 = Total amount of TCE metabolized per unit body weight74 (mg TCE/kgyYwk)
8 TotOxMetabBW34 = Total amount of TCE oxidized per unit body weight 4 (mg TCE/kgyYwk)
9
6/22/2009
1130
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.2.4. Mean PBPK model predictions for weekly internal dose in humans exposed
continuously to low levels of TCE via inhalation (ppm) or orally (mg/kg/d).
Dose Metric
0.001
ppm
0.001 mg/kg/d
Female
Male
Female
Male
ABioactDCVCBW34
0.00324
0.00324
0.00493
0.00515
AMetGSHB W 3 4
0.00200
0.00200
0.00304
0.00318
AMetLivlBW34
0.00703
0.00683
0.0157
0.0164
AMetLngB W3 4
0.00281
0.00287
6.60xl0"5
6.08xl0"5
AUCCBld
0.00288
0.00298
0.000411
0.000372
TotMetabBW34
0.0118
0.0117
0.0188
0.0196
TotOxMetabBW34
0.00984
0.00970
0.0157
0.0164
See note to Table 5.2.3 for dose metric abbreviations. Values represent the mean of the
(uncertainty) distribution of population means for each sex and exposure scenario, generated
from Monte Carlo simulation of 500 populations of 500 individuals each.
5.2.1.3 Rodent dose-response analyses: Results
A summary of the "points of departure" (PODs) and unit risk estimates for each
sex/species/bioassay/tumor type is presented in Tables 5.2.5 (inhalation studies) and 5.2.6 (oral
studies). The PODs for individual tumor types were extracted from the modeling results in the
figures in Appendix G. For the applied dose (default dosimetry) analyses, the POD is the BMDL
from the male human ("M") BMDL entry at the top of the figure for the selected model; male
results were extracted because the default weight for males in the PBPK modeling is 70 kg,
which is the overall human weight in EPA's default dosimetry methods (for inhalation, male and
female results are identical). As described in Section 5.2.1.2 above, for internal dose metrics,
male and female results were averaged, and the converted human "BMDLs" are not true BMDLs
because they were converted outside the linear range of the PBPK models. It can be seen in
Appendix G that the male and female results were similar for all the dose metrics.
For two datasets, the highest dose (exposure) group was dropped to get a better fit when
using applied doses. This technique can improve the fit when the response tends to plateau with
increasing dose. Plateauing typically occurs when metabolic saturation alters the pattern of
metabolite formation or when survival is impacted at higher doses, and it is assumed that these
high-dose responses are less relevant to low-dose risk. The highest dose group was not dropped
to improve the fit for any of the internal dose metrics because it was felt that if the dose metric
was an appropriate reflection of internal dose of the reactive metabolite(s), then use of the dose
metric should have ameliorated the plateauing in the dose-response relationship (note that
survival-impacted datasets were addressed using survival adjustment techniques). For a 3rd
6/22/2009
1131
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
dataset (Henschler lymphomas), it might have helped to drop the highest exposure group, but
there were only two exposure groups, so this was not done. As a result, the selected model,
although it had an adequate fit overall, didn't fit the control group very well (the model estimated
a higher background response than was observed); thus, the BMD and BMDL were likely
overestimated and the risk underestimated. The estimates from the NCI 1976 oral male mouse
liver cancer dataset are also somewhat more uncertain because the response rate was
extrapolated down from a response rate of about 50% extra risk to the BMR of 10% extra risk.
Some general patterns can be observed in Tables 5.2.5 and 5.2.6. For inhalation, the unit
risk estimates for different dose metrics were generally similar (within about 2.5-fold) for most
tumor types. The exception was for kidney cancer, where the estimates varied by over 2 orders
of magnitude, with the AMetGSHBW34 and ABioactDCVCBW34 metrics yielding the highest
estimates. This occurs because pharmacokinetic data indicate, and the PBPK model predicts,
substantially more GSH conjugation (as a fraction of intake), and hence subsequent
bioactivation, in humans relative to rats. The range of the risk estimates for individual tumor
types overall (across tumor types and dose metrics) was encompassed by the range of estimates
across the dose metrics for kidney cancer in the male rat, which was from 4.4 x 10"4 per ppm
(applied dose) to 8.3 x 10"2 per ppm (ABioactDCVCBW34).
For oral exposure, the unit risk (slope factor) estimates are more variable across dose
metrics because of first-pass effects in the liver (median estimates for the fraction of TCE
metabolized in one pass through the liver in mice, rats, and humans are > 0.8). Here, the
exception is for the risk estimates for cancer of the liver itself, which are also within about a 2.5-
fold range, because the liver gets the full dose of all the metrics during that "first pass". For the
other tumor types, the range of estimates across dose metrics varies from about 30-fold to over 2
orders of magnitude, with the estimates based on AUCCBld and AMetLngBW34 being at the
low end and those based on AMetGSHBW34 and ABioactDCVCBW34 again being at the high
end. For AUCCBld, the PBPK model predicted the blood concentrations to scale more closely
to body weight rather than the 3/4 power of body weight, so the extrapolated human unit risks
using this dose metric are smaller than those obtained by applied dose or other dose metrics that
included 3/4 power body weight scaling. For AMetLngBW34, pharmacokinetic data indicate, and
the PBPK model predicts, that the human respiratory tract metabolizes a lower fraction of total
TCE intake than the mouse respiratory tract, so the extrapolated risk to humans based on this
metric is lower than that obtained using applied dose or other dose metrics. Overall, the oral unit
risk estimates for individual tumor types ranged from 1.6 x 10"5 per mg/kg/day (female rat
leukemia, AUCCBld) to 2.5 x 10"1 per mg/kg/day (male Osborne-Mendel rat kidney,
ABioactDCVCBW34), a range of over 4 orders of magnitude. It must be recognized, however,
that not all dose metrics are equally credible, and, as will be presented below, the unit risk
6/22/2009
1132
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
estimates for total cancer risk for the most sensitive bioassay response for each sex/species
combination using the primary (preferred) dose metrics fall within a very narrow range.
6/22/2009
1133
-------
1
2
3
4
5
6
7
8
9
10
11
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.2.5. Summary of POPs and unit risk estimates for each sex/species/bioassay/tumor type (inhalation)
Study
Tumor Type
BMR
PODs (ppm, in human equivalent exposures)3
applied
AUC
TotMetab
TotOxMetab
AMetLng
AMetLivl
AMetGSH
ABioact
dose
CBld
BW34
BW34
BW34
BW34
BW34
DCVCBW34
FEMALE MOUSE
Fukuda
lung ad + carc
0.1
26.3
55.5
31.3
38.8
Henschler
lymphoma
0.1
11.0b
__b
9.84
Maltoni
lung ad + carc
0.1
44.6
96.6
51.4
55.7
liver
0.05
37.1
45.8
41.9
combined
0.05
15.7
20.7
MALE MOUSE
Maltoni
liver
0.1
34.3
51
37.9
MALE RAT
Maltoni
leukemia
0.05
28.2°
b
28.3
kidney ad + carc
0.01
22.7
13.7
0.197
0.121
leydig cell
0.1
18.6°
d
18.1
combined
0.01
1.44
1.37
Study
Tumor Type
Unit Risk Estimate (ppm 1)e
applied dose
AUC
TotMetab
TotOxMetab
AMetLng
AMetLivl
AMetGSH
ABioact
CBld
BW34
BW34
BW34
BW34
BW34
DCVCBW34
FEMALE MOUSE
Fukuda
lung ad + carc
3.8
X 10"3
1.8 x 10"3
3.2 x 10"3
2.6 x 10 3
Henschler
lymphoma
9.1
X 10"3
1.0 x 10 2
Maltoni
lung ad + carc
2.2
X 10-3
1.0 x 10"3
1.9 x 10"3
1.8 x 10 3
liver
1.3
X 10-3
1.1 x 10"3
1.2 x 10 3
combined
3.2
X 10-3
2.4 x 10 3
MALE MOUSE
Maltoni
liver
2.9
X 10-3
2.0 x 10"3
2.6 x 10 3
MALE RAT
Maltoni
leukemia
1.8
X 10-3
1.8 x 10 3
kidney ad + carc
4.4
X 10"4
7.3 x 10"4
5.1 x 10"2
8.3 x 10 2
leydig cell
5.4
X 10"3
5.5 x 10 3
combined
7.0
X 10"3
7.3 x 10"3
a. for the applied doses, the PODs are BMDLs. However, for the internal dose metrics, the PODs are not actually human equivalent BMDLs corresponding to the BMR because the
interspecies conversion does not apply to the dose range of the BMDL; the converted BMDLs are merely intermediaries to obtain a converted unit risk estimate. The calculation that
was done is equivalent to using linear extrapolation from the BMDLs in terms of the internal dose metric to get a unit risk estimate for low-dose risk in terms of the internal dose metric
and then converting that estimate to a unit risk estimate in terms of human equivalent exposures. The PODs reported here are what one would get if one then used the unit risk estimate
to calculate the human exposure level corresponding to a 10% extra risk, but the unit risk estimate is not intended to be extrapolated upward out of the low-dose range, e.g., above 10"4
risk. In addition, for the internal dose metrics, the PODs are the average of the male and female human "BMDL" results presented in Appendix G.
b. inadequate fit to control group, but the primary metric, TotMetabBW34, fits adequately.
c. dropped highest dose group to improve model fit
d. inadequate overall fit
e. unit risk estimate = BMR/POD. Results for the primary dose metric are in bold.
6/22/2009
1134
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Study
Tumor Type
BMR
PODs (mg/kg/day, in human equivalent doses)3
applied
AUC
TotMetab
TotOxMetab
AMetLng
AMetLivl
AMetGSH
ABioact
dose
CBld
BW34
BW34
BW34
BW34
BW34
DCVCBW34
FEMALE MOUSE
NCI
liver carc
0.1
26.5
17.6
14.1
lung ad + carc
0.1
41.1
682
24.7
757
leukemias +
0.1
43.1
733
20.6
sarcomas
combined
0.05
7.43
5.38
MALE MOUSE
NCI
liver carc
0.1
8.23
4.34
3.45
FEMALE RAT
NTP 1988
leukemia
0.05
72.3
3220
21.7
MALE RAT
NTP 1990°
kidney ad + carc
0.1
32
11.5
0.471
0.292
NTP 1988
Marshall
testicular
0.1
3.95
167
1.41
August
subcut sarcoma
0.05
60.2
2560
21.5
Osborne-
kidney ad + carc
0.1
41.5
14.3
0.648
0.402
Mendef
Study
Tumor Type
Unit Risk Estimate ((mg/kg/day) 1)b
applied dose
AUC
TotMetab
TotOxMetab
AMetLng
AMetLivl
AMetGSH
ABioact
CBld
BW34
BW34
BW34
BW34
BW34
DCVCBW34
FEMALE MOUSE
NCI
liver carc
3.i
X 10"3
5.7 x 10"3
7.1 x 10 3
lung ad + carc
2.4 x 10"3
1.5 x 10"4
4.0 x 10"3
1.3 x 10 4
leukemias +
2.3 x 10"3
1.4 x 10"4
4.9 x 10 3
sarcomas
combined
6.7 x 10"3
9.3 x 10 3
MALE MOUSE
NCI
liver carc
1.2 x 10"2
2.3 x lO"2
2.9 x 10 2
FEMALE RAT
NTP 1988
leukemia
6.9 x 10"4
1.6 x 10"5
2.3 x 10 3
MALE RAT
NTP 1990°
kidney ad + carc
1.6 x 10"3
4.3 x 10"3
1.1 x 10"1
1.7 x 10 1
NTP 1988
Marshall
testicular
2.5 x 10"2
6.0 x 10"4
7.1 x 10 2
August
subcut sarcoma
8.3 x 10"4
2.0 x 10"5
2.3 x 10 3
Osborne-
kidney ad + carc
2.4 x 10"3
7.0 x 10"3
1.5 x 10"1
2.5 x 10 1
Mendef
6/22/2009
1135
-------
1
2
3
4
5
6
7
8
9
10
11
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
a. for the applied doses, the PODs are BMDLs. However, for the internal dose metrics, the PODs are not actually human equivalent BMDLs corresponding to the BMR because the
interspecies conversion does not apply to the dose range of the BMDL; the converted BMDLs are merely intermediaries to obtain a converted unit risk estimate. The calculation that
was done is equivalent to using linear extrapolation from the BMDLs in terms of the internal dose metric to get a unit risk estimate for low-dose risk in terms of the internal dose metric
and then converting that estimate to a unit risk (slope factor) estimate in terms of human equivalent doses. The PODs reported here are what one would get if one then used the unit risk
estimate to calculate the human dose level corresponding to a 10% extra risk, but the unit risk estimate is not intended to be extrapolated upward out of the low-dose range, e.g., above
10"4 risk. In addition, for the internal dose metrics, the PODs are the average of the male and female human "BMDL" results presented in Appendix G.
b. unit risk estimate = BMR/POD. Results for the primary dose metric are in bold.
c. using MSW adjusted incidences (see text and Table 5.2.7).
d. using poly-3 adjusted incidences (see text and Table 5.2.7).
6/22/2009
1136
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Results for survival-adjusted analyses are summarized in Table 5.2.7. For the time-
independent (BMDS) multistage model, the risk estimates using poly-3 adjustment are higher
than those without poly-3 adjustment. This is to be expected because the poly-3 adjustment
decreases denominators when accounting for early mortality, and, for these datasets, the higher
dose groups had greater early mortality. The difference was fairly modest for the kidney cancer
datasets (about 30% higher) but somewhat larger for the testicular cancer dataset (about 150%
higher).
In addition, the MSW time-to-tumor model generated higher risk estimates than the poly-
3 adjustment technique. The MSW results were about 40% higher for the NTP F344 rat kidney
cancer datasets and about 60% higher for the NTP Osborne-Mendel rat kidney cancer datasets.
For the NTP Marshall rat testicular cancer dataset, the discrepancies were greater; the results
ranged from about 100% to 180% higher for the different dose metrics. As discussed in Section
5.2.1.1, these two approaches differ in the way they take early mortality into account. The poly-
3 technique merely adjusts the tumor incidence denominators, using a constant power 3 of time,
to reflect the fact that animals are at greater risk of cancer at older ages. The MSW model
estimates risk as a function of time (and dose), and it estimates the power (of time) parameter for
each dataset.50 For the NTP F344 rat kidney cancer and NTP Marshall rat testicular cancer
datasets, the estimated power parameter was close to 3 in each case, ranging from 3.0 to 3.7; for
the NTP Osborne-Mendel rat kidney cancer datasets, however, the estimated power parameter
was about 10 for each of the dose metrics, presumably reflecting the fact that these were late-
occurring tumors (the earliest occurred at 92 weeks). Using a higher power parameter than 3 in
the poly-3 adjustment would give even less weight to non-tumor-bearing animals that die early
and would, thus, increase the adjusted incidence even more in the highest dose groups where the
early mortality is most pronounced, increasing the unit risk estimate. Nonetheless, as noted
above, the MSW results were only about 60% higher for the NTP Osborne-Mendel rat kidney
cancer datasets for which MSW estimated a power parameter of about 10.
In general, the risk estimates from the MSW model would be preferred because, as
discussed above, this model incorporates more information (e.g. tumor context) and estimates the
power parameter rather than using a constant value of 3. From Table 5.2.7, it can be seen that
the results from MSW yielded higher BMD:BMDL ratios than the results from the poly-3
technique. These ratios were only slightly higher and not unusually large for MSW model
analyses of the NTP (1988, 1990) kidney tumor estimates, and this, along with the adequate fit
50 Conceptually, the approaches differ most when different tumor contexts (incidental or fatal) are considered,
because the poly-3 technique only accounts for time of death, while the MSW model can account for the tumor
context and attempt to estimate an induction time (to), although this was not done for any of the datasets in this
assessment.
6/22/2009
1137
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 (assessed visually) of the MSW model, supports using the unit risk estimates from the MSW
2 modeling of rat kidney tumor incidence. On the other hand, the BMD:BMDL ratio was
3 relatively large for the applied dose analysis and, in particular, for the preferred dose metric
4 analysis (9.4- fold) of the NTP Marshall rat testicular tumor dataset. Therefore, for this
5 endpoint, the poly-3-adjusted results were used, although they may underestimate risk somewhat
6 as compared to the MSW model.
6/22/2009
1138
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1
2 Table 5.2.7. Comparison of survival-adjusted results for 3 oral male rat datasets3
dose metric
adjustment
BMR
POD
BMD:BMDL
unit risk estimate
method
(mg/kg/day)
(per m
5/kg/day)
NTP 1990 F344 RAT KIDNEY AD + CARC
applied dose
unadj BMDS
0.05
56.9
1.9
8.8
X 10-4
poly-3 BMDS
0.1
89.2
1.9
1.1
X 10"3
MSW
0.05
32.0
2.6
1.6
X 10"3
TotMetabBW34
unadj BMDS
0.05
20.2
2.1
2.5
X 10"3
poly-3 BMDS
0.1
31.8
1.7
3.1
X 10"3
MSW
0.05
11.5
3.1
4.3
X 10-3
AMetGSHBW34
unadj BMDS
0.05
0.841
1.9
5.9
X 10-2
poly-3 BMDS
0.1
1.32
1.9
7.6
X 10-2
MSW
0.05
0.471
2.4
1.1
X 10-1
ABioactDCVCBW34
unadj BMDS
0.05
0.522
1.9
9.6
X 10-2
poly-3 BMDS
0.1
0.817
1.9
1.2
X 10-1
MSW
0.05
0.292
2.4
1.7 x 10 1
NTP 1988 OSBORNE-MENDEL RAT KIDNEY AD + CARC
applied dose
unadj BMDS
0.1
86.6
1.7
1.2
X 10-3
poly-3 BMDS
0.1
65.9
1.7
1.5
X 10-3
MSW
0.1
41.5
2.0
2.4
X 10-3
TotMetabBW34
unadj BMDS
0.1
30.4
1.7
3.3
X 10"3
poly-3 BMDS
0.1
23.1
1.7
4.3
X 10"3
MSW
0.1
14.3
2.0
7.0
X 10"3
AMetGSHBW34
unadj BMDS
0.1
1.35
1.7
7.4
X 10"2
poly-3 BMDS
0.1
1.03
1.7
9.7
X 10"2
MSW
0.1
0.648
2.0
1.5
X 10-1
ABioactDCVCBW34
unadj BMDS
0.1
0.835
1.7
1.2
X 10-1
poly-3 BMDS
0.1
0.636
1.7
1.6
X 10-1
MSW
0.1
0.402
2.0
2.5 x 10 1
NTP 1988 MARSHALL RAT TESTICULAR TUMORS
applied dose
unadj BMDS
0.1
9.94
1.4
1.0
X 10-2
poly-3 BMDS
0.1
3.95
1.5
2.5
X 10-2
MSW
0.1
1.64
5.2
6.1
X 10-2
AUCCBld
unadj BMDS
0.1
427
1.4
2.3
X 10-4
poly-3 BMDS
0.1
167
1.6
6.0
X 10-4
MSW
0.1
60.4
2.6
1.7
X 10-3
TotMetabBW34
unadj BMDS
0.1
3.53
4.3
2.8
X 10"2
poly-3 BMDS
0.1
1.41
1.5
7.1 x 10 2
MSW
0.1
0.73
9.4
1.4
X 10"1
3 a. for the applied doses, the PODs are BMDLs. However, for the internal dose metrics, the PODs are not actually human
4 equivalent BMDLs corresponding to the BMR because the interspecies conversion does not apply to the dose range of the
5 BMDL; the converted BMDLs are merely intermediaries to obtain a converted unit risk estimate. Results for the primary dose
6 metric are in bold.
7
8 In addition to the results from dose-response modeling of individual tumor types, the
9 results of the combined tumor risk analyses for the three bioassays in which the rodents exhibited
10 increased risks at multiple sites are also presented in Tables 5.2.5 and 5.2.6, in the rows labeled
11 "combined" under the column heading "Tumor Type". These results were extracted from the
12 detailed results in Appendix G. Note that, because of the computational complexity of the
6/22/2009
1139
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
combined tumor analyses, dose-response modeling was only done using applied dose and a
common upstream internal dose metric, rather than using the different preferred dose metrics for
each tumor type within a combined tumor analysis.
For the Maltoni female mouse inhalation bioassay, the combined tumor risk estimates are
bounded by the highest individual tumor risk estimates and the sums of the individual tumor
risks estimates (the risk estimates are upper bounds, so the combined risk estimate, i.e., the upper
bound on the sum of the individual central tendency estimates, should be less than the sum of the
individual upper bound estimates), as one would expect. The common upstream internal dose
metric used for the combined analysis was TotOxMetabBW34, which is not the primary metric
for either of the individual tumor types. For the liver tumors, the primary metric was
AMetLivlBW34, but as can be seen in Table 5.2.5, it yields results similar to those for
TotOxMetabBW34. Likewise, for the lung tumors, the primary metric was AMetLngBW34,
which yields a unit risk estimate slightly smaller that for TotOxMetabBW34 Thus, the results of
the combined analysis using TotOxMetabBW34 as a common metric is not likely to substantially
over- or underestimate the combined risk based on preferred metrics for each of the tumor types.
For the Maltoni male rat inhalation bioassay, the combined risk estimates are also
reasonably bounded, as expected. The common upstream internal dose metric used for the
combined analysis was TotMetabBW34, which is the primary metric for two of the three
individual tumor types. However, as can be seen in Table 5.2.5, the risk estimate for the
preferred dose metric for the third tumor type, ABioactDCVCBW34 for the kidney tumors, is
substantially higher than the risk estimates for the primary dose metrics for the other two tumor
types and would dominate a combined tumor risk estimate across primary dose metrics; thus, the
ABioactDCVCBW34-based kidney tumor risk estimate alone can reasonably be used to
represent the total cancer risk for the bioassay using preferred internal dose metrics, although it
would underestimate the combined risk to some extent (e.g., the kidney-based estimate is 8.3 x
10"2 per ppm; the combined estimate would be about 9 x 10"2 per ppm, rounded to one significant
figure).
For the third bioassay (NCI female mouse oral bioassay), the combined tumor risk
estimates are once again reasonably bounded. The common upstream internal dose metric used
for the combined analysis was TotOxMetabBW34, which is not the primary metric for any of the
3 individual tumor types but was considered to be the most suitable metric to apply as a basis for
combining risk across these different tumor types. The unit risk estimate for the lung based on
the primary dose metric for that site becomes negligible compared to the estimates for the other 2
tumor types (see Table 5.2.6). However, the unit risk estimates for the remaining 2 tumor types
are both somewhat underestimated using the TotOxMetabBW34 metric rather than the primary
metrics for those tumors (the TotOxMetabBW34-based estimate for leukemias + sarcomas,
6/22/2009
1140
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
which is not presented in Table 5.2.6 because, in the absence of better mechanistic information,
more upstream metrics were used for that individual tumor type, is 4.1 x 10"3 per mg/kg/day).
Thus, overall, the combined estimate based on TotOxMetabBW34 is probably a reasonable
estimate for the total tumor risk in this bioassay, although it might overestimate risk slightly.
The most sensitive sex/species results are extracted from Tables 5.2.5 and 5.2.6 and
presented in Tables 5.2.8 (inhalation) and 5.2.9 (oral) below. The BMD:BMDL ratios for all the
results corresponding to the unit risk estimates based on the preferred dose metrics ranged from
1.3-2.1. For inhalation, the most sensitive bioassay responses based on the preferred dose
metrics ranged from 2.6 x 1CT3 per ppm to 8.3 x 1CT2 per ppm across the sex/species
combinations (with the exception of the female rat, which exhibited no apparent TCE-associated
response in the 3 available bioassays). For oral exposure, the most sensitive bioassay responses
based on the preferred dose metrics ranged from 2.3 x 10 3 per mg/kg/day to 2.5 x 10 1 per
mg/kg/day across the sex/species combinations. For both routes of exposure, the most sensitive
sex/species response was (or was dominated by, in the case of the combined tumors in the male
rat by inhalation) male rat kidney cancer based on the preferred dose metric of
ABioactDCVCBW34.
6/22/2009
1141
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1
2 Table 5.2.8. Inhalation: Most sensitive bioassay for each sex/species combination3
Sex/Species
Endpoint
Unit risk per ppm
(Study)
Preferred
Default
Alternative dose
dose metric
methodology
metrics, studies, or
endpoints
Female mouse
Lymphoma
(Henschler et al. 1980)
1.0 x 10~2
9.1 x 10~3
1 x 10~3 ~ 4 x 10~3
Male mouse
Liver hepatoma
(Maltoni et al. 1986)
2.6 x 10~3
2.9 x 10~3
2 x 10~3
Female rat
-
-
-
-
Male rat
Leukemia +
Kidney aden & carc +
Leydig cell tumors
(Maltoni et al. 1986)
8.3 x 10~2
7.0 x 10~3
4 x 10^-5 x 10~2
[individual site
results]
3 a. results extracted from Table 5.2.5
4
5 Table 5.2.9. Oral: Most sensitive bioassay for each sex/species combination3
Sex/Species Endpoint Unit risk per mg/kg/day
(Study)
Preferred
Default
Alternative dose
dose metric
methodology
metrics, studies, or
endpoints
Female mouse
Liver carcinomas +
lung aden & carc +
sarcomas + leukemias
(NCI 1976)
9.3 x 10~3
6.7 x 10~3
1 x 10^-7 x 10~3
[individual site
results]
Male mouse
Liver carcinomas
(NCI 1976)
2.9 x 10~2
1.2 x 10~2
2 x 10~2
Female rat
Leukemia
(NTP 1988)
2.3 x 10~3
6.9 x 10^
2 x 10~5
Male rat
Kidney adenomas +
carcinomas
(NTP 1988, Osborne-
Mendel)
2.5 x 10"1
2.4 x 10~3b
2 x 10~5 ~ 2 x 10"1
6 a. results extracted from Table 5.2.6
7 b. most sensitive male rat result using default methodology is 2.5 x 10 2 per mg/kg/day for NTP
8 (1988) Marshall rat testicular tumors
6/22/2009
1142
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.2.1.4 Uncertainties in dose-response analyses of rodent bioassays
5.2.1.4.1 Qualitative discussion of uncertainties
All risk assessments involve uncertainty, as study data are extrapolated to make
inferences about potential effects in humans from environmental exposure. The largest sources
of uncertainty in the TCE rodent-based cancer risk estimates are interspecies extrapolation and
low-dose extrapolation. Some limited human (occupational) data from which to estimate human
cancer risk are available, and cancer risk estimates based on these data are developed in Section
5.2.2 below. In addition, some quantitative uncertainty analyses of the interspecies differences
in pharmacokinetics were conducted and are presented in Section 5.2.1.4.2.
The rodent bioassay data offer conclusive evidence of carcinogenicity in both rats and
mice, and the available epidemiologic and mechanistic data support the relevance to humans of
the TCE-induced carcinogenicity observed in rodents. The epidemiologic data provide sufficient
evidence that TCE is carcinogenic to humans (see Section 4.10). There is even some evidence of
site concordance with the rodent findings, although site concordance is not essential to human
relevance and, in fact, is not observed across TCE-exposed rats and mice. The strongest
evidence in humans is for TCE-induced kidney tumors, with fairly strong evidence for
lymphomas and some lesser support for liver tumors; each of these tumor types has also been
observed in TCE rodent bioassays. Furthermore, the mechanistic data are supportive of human
relevance because, while the exact reactive species associated with TCE-induced tumors aren't
known, the metabolic pathways for TCE are qualitatively similar for rats, mice, and humans
(Section 3.3). The impact of uncertainties with respect to quantitative differences in TCE
metabolism is discussed in Section 5.2.1.4.2.
Typically, we estimate the cancer risk for the total cancer burden from all sites that
demonstrate an increased tumor incidence for the most sensitive experimental species and sex. It
is expected that this approach is protective of the human population, which is more diverse but is
exposed to lower exposure levels.
For the inhalation unit risk estimates, the preferred estimate from the most sensitive
species and sex was the estimate of 8.3 x 10 2 per ppm for the male rat, which was based on
multiple tumors observed in this sex/species but was dominated by the kidney tumor risk
estimated with the dose metric for bioactivated DCVC. This estimate was the high end of the
range of estimates (see Table 5.2.8) but was within an order of magnitude of other estimates,
such as the preferred estimate for the female mouse and the male rat kidney estimate based on
the GSH conjugation dose metric, which provide additional support for an estimate of this
magnitude. The preferred estimate for the male mouse was about an order of magnitude and a
6/22/2009
1143
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
half lower. The female rat showed no apparent TCE-associated tumor response in the 3 available
inhalation bioassays; however, this apparent absence of response is inconsistent with the
observations of increased cancer risk in occupationally exposed humans and in female rats in
oral bioassays. In Section 5.2.2.2, an inhalation unit risk estimate based on the human data is
derived and can be compared to the rodent-based estimate.
For the oral unit risk (slope factor) estimate, the preferred estimate from the most
sensitive species and sex was the estimate of 2.5 x 10 1 per mg/kg/day, again for the male rat,
based on the kidney tumor risk estimated with the dose metric for bioactivated DCVC. This
estimate was at the high end of the range of estimates (see Table 5.2.9) but was within an order
of magnitude of other estimates, such as the preferred male mouse estimate and the male rat
kidney estimate based on the GSH conjugation dose metric, which provide additional support for
an estimate of this magnitude. The preferred estimates for the female mouse and the female rat
were about another order of magnitude lower. Some of the oral unit risk estimates based on the
alternative dose metric of AUC for TCE in the blood were as much as 3 orders of magnitude
lower, but these estimates were considered less credible than those based on the preferred dose
metrics. In Section 5.2.2.3, an oral unit risk estimate based on the human (inhalation) data is
derived using the PBPK model for route-to-route extrapolation; this estimate can be compared to
the rodent-based estimate.
Furthermore, the male rat kidney tumor estimates from the inhalation (Maltoni et al.
1986) and oral (NTP 1988) studies were consistent on the basis of internal dose using the dose
metric for bioactivated DCVC. In particular, the linearly extrapolated slope (i.e., the
BMR/BMDL) per unit of internal dose derived from Maltoni et al. (1986) male rat kidney tumor
data was 2.4 x 10 1 per weekly mg DCVC bioactivated per unit body weight74, while the
analogous slope derived from NTP (1988) male rat kidney tumor data was 9.3 x 10 2 per weekly
mg DCVC bioactivated per unit body weight74 (MSW-modeled results), a difference of less than
3-fold.51 These results also suggest that differences between routes of administration are
adequately accounted for by the PBPK model using this dose metric.
51 For the Maltoni et al. (1986) male rat kidney tumors, the unit risk estimate of 8.3 x 10"2 per ppm using the
ABioactDCVCBW34 dose metric, from Table 5.2.5, is divided by the average male and female internal doses at
0.001 ppm, (0.0034/0.001), from Table 5.2.4, to yield a unit risk in internal dose units of 2.4 x 10"2. For the NTP
(1988) male rat kidney tumors, the unit risk estimate of 2.5 x 10"1 per mg/kg/d using the ABioactDCVCBW34 dose
metric, from Table 5.2.6, is divided by the average male and female internal doses at 0.001 mg/kg/d, (0.0027/0.001),
from Table 5.2.4, to yield a unit risk in internal dose units of 9.3 x 10"2. Note that the original BMDLs and unit risks
from BMD modeling were in internal dose units that were then converted to applied dose units using the values in
Table 5.2.4, so this calculation reverses that conversion.
6/22/2009
1144
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Regarding low-dose extrapolation, a key consideration in determining what extrapolation
approach to use is the MOA(s). However, MOA data are lacking or limited for each of the
cancer responses associated with TCE exposure, with the exception of the kidney tumors (see
Section 4.10). For the kidney tumors, the weight of the available evidence supports the
conclusion that a mutagenic MOA is operative (Section 4.3); this MOA supports linear low-dose
extrapolation. For the other TCE-induced tumors, the MOA(s) is unknown. When the MOA(s)
cannot be clearly defined, EPA generally uses a linear approach to estimate low-dose risk (U.S.
EPA, 2005a), based on the following general principles:
• A chemical's carcinogenic effects may act additively to ongoing biological processes,
given that diverse human populations are already exposed to other agents and have
substantial background incidences of various cancers.
• A broadening of the dose-response curve (i.e., less rapid fall-off of response with
decreasing dose) in diverse human populations and, accordingly, a greater potential for
risks from low-dose exposures (Ziese et al., 1987; Lutz et al., 2005) is expected for two
reasons: First, even if there is a "threshold" concentration for effects at the cellular level,
that threshold is expected to differ across individuals. Second, greater variability in
response to exposures would be anticipated in heterogeneous populations than in inbred
laboratory species under controlled conditions (due to, e.g., genetic variability, disease
status, age, nutrition, and smoking status).
• The general use of linear extrapolation provides reasonable upper-bound estimates that
are believed to be health-protective (U.S. EPA, 2005a) and also provides consistency
across assessments.
Additional uncertainties arise from the specific dosimetry assumptions, the model
structures and parameter estimates in the PBPK models, the dose-response modeling of data in
the observable range, and the application of the results to potentially sensitive human
populations. As discussed in Section 5.2.1.2.1, one uncertainty in the tissue-specific dose
metrics used here is whether to scale the rate of metabolism by tissue mass or body weight to the
3/4 in the absence of specific data on clearance; however, in the cases where this is an issue (the
lung, liver, and kidney), the impact of this choice is relatively modest (less than 2-fold to about
4-fold). An additional dosimetry assumption inherent in this analysis is that equal concentrations
of the active moiety over a lifetime yield equivalent lifetime risk of cancer across species, and
6/22/2009
1145
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
the extent to which this is true for TCE is unknown. Furthermore, it should be noted that use of
tissue-specific dosimetry inherently presumes site concordance of tumors across species.
With respect to uncertainties in the estimates of internal dose themselves, a quantitative
analysis of the uncertainty and variability in the PBPK model-predicted dose metric estimates
and their impacts on cancer risk estimates is presented in Section 5.2.1.4.2. Additional
uncertainties in the PBPK model were discussed Section 3.5. Furthermore, this assessment
examined a variety of dose metrics for the different tumor types using PBPK models for rats,
mice, and humans, so the impact of dose metric selection can be assessed. As discussed in
Section 5.2.1.2.1, there is strong support for the primary dose metrics selected for kidney, liver,
and, to a lesser extent, lung. For the other tumor sites, there is more uncertainty about dose
metric selection. The cancer unit risk estimates obtained using the preferred dose metrics were
generally similar (within about 3-fold) to those derived using default dosimetry assumptions
(e.g., equal risks result from equal cumulative equivalent exposures or doses), with the exception
of the bioactivated DCVC dose metric for rat kidney tumors and the metric for the amount of
TCE oxidized in the respiratory tract for mouse lung tumors occurring from oral exposure (see
Tables 5.2.8 and 5.2.9). The higher risk estimates for kidney tumors based on the bioactivated
DCVC dose metric are to be expected because pharmacokinetic data indicate, and the PBPK
model predicts, substantially more GSH conjugation (as a fraction of intake), and hence
subsequent bioactivation, in humans relative to rats. The lower risk estimates for lung tumors
from oral TCE exposure based on the metric for the amount of TCE oxidized in the respiratory
tract are because there is a greater first-pass effect in human liver relative to mouse liver
following oral exposure and because the gavage dosing used in rodent studies leads to a large
bolus dose that potentially overwhelms liver metabolism to a greater extent than a more graded
oral exposure. Both of these effects result in relatively more TCE being available for
metabolism in the lung for mice than for humans. In addition, mice have greater respiratory
metabolism relative to humans. However, because oxidative metabolites produced in the liver
may contribute to respiratory tract effects, using respiratory tract metabolism alone as a dose
metric may underestimate lung tumor risk. The unit risk estimates obtained using the alternative
dose metrics were also generally similar to those derived using default dosimetry assumptions,
with the exception of the metric for the amount of TCE conjugated with GSH for rat kidney
tumors, again because humans have greater GSH conjugation, and the AUC of TCE in blood for
all the tumor types resulting from oral exposure, again because of first-pass effects.
With respect to uncertainties in the dose-response modeling, the two-step approach of
modeling only in the observable range, as put forth in EPA's cancer assessment guidelines (U.S.
EPA, 2005a), is designed in part to minimize model dependence. The ratios of the BMDs to the
BMDLs give some indication of the uncertainties in the dose-response modeling. These ratios
6/22/2009
1146
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
did not exceed a value of 2.5 for all the primary analyses used in this assessment. Thus, overall,
modeling uncertainties in the observable range are considered to be negligible. Some additional
uncertainty is conveyed by uncertainties in the survival adjustments made to some of the
bioassay data; however, their impact is also believed to be minimal relative to the uncertainties
already discussed (i.e., interspecies and low-dose extrapolations).
Regarding the cancer risks to potentially sensitive human populations or life stages,
pharmacokinetic data on 42 individuals were used in the Bayesian population analysis of the
PBPK model discussed in Section 3.5. The impacts of these data on the predicted population
mean are incorporated in the quantitative uncertainty analyses presented in Section 5.2.1.4.2.
These data do not, however, reflect the full range of metabolic variability in the human
population (they are all from healthy, mostly male, human volunteers) and do not address
specific potentially sensitive subgroups (see Section 4.9). Moreover, there is inadequate
information about disease status, co-exposures, and other factors that make humans vary in their
responses to TCE. It will be a challenge for future research to quantify the differential risk
indicated by different risk factors or exposure scenarios.
5.2.1.4.2 Quantitative uncertainty analysis of PBPK model-based dose metrics
The Bayesian analysis of the PBPK model for TCE generates distributions of uncertainty
and variability in the internal dose metrics than can be readily fed into dose-response analysis.
As shown in Figure 5.2.2, the overall approach taken for the uncertainty analysis is similar to
that used for the point estimates except that distributions are carried through the analysis rather
than median or expected values. In particular, the PBPK model-based rodent internal doses are
carried through to a distribution of BMDs (which also includes sampling variance from the
number of responding and at risk animals in the bioassay). This distribution of BMDs generates
a distribution of cancer slope factors based on internal dose, which then is combined with the
(uncertainty) distribution of the human population mean conversion to applied dose or exposure.
The resulting distribution for the human population mean risk per unit dose or exposure accounts
for uncertainty in the PBPK model parameters (rodent and human) and the binomial sampling
error in the bioassays. These distributions can then be compared with the point estimates, based
on median rodent dose metrics and mean human population dose metrics, reported in Tables
5.2.5 and 5.2.6. Details of the implementation of this uncertainty analysis, which used the
WinBugs software in conjugation with the R statistical package, are reported in Appendix G.
6/22/2009
1147
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Bioassay
experimental
paradigm
Rodent
model
parameters
\
fixed 1
PBPK
model
[distribution (combined
"uncertainty and variability)
Rodent
internal
dose
Bioassay
responses
\
fixed \
^distribution
Dose- /
'esponse/
\Mode!/
'distribution
0.001 ppm
in air or
0.001 mg/kg-d
continuous
exposure
Human
model
parameters
\
fixedl
PBPK
model
[distribution (separate
"uncertainty and variability)
Human
internal
dose
'Uncertainty
(distribution of
[population mean
Cancer slope
factor =
BMR/BMD
(per internal
dose unit))
Population
mean human
internal dose
per ppm or
per mg/kg-d
Human cancer
slope factor
(per ppm or
per mg/kg-d)
distribution
2 Figure 5.2.2.
3 Flow-chart for uncertainty analysis of dose-response analyses of rodent bioassays using PBPK
4 model-based dose metrics. Square nodes indicate point values, circular nodes indicate
5 distributions, and the inverted triangles indicate a (deterministic) functional relationship.
6/22/2009
1148
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Overall, as shown in Tables 5.2.10 and 5.2.11, the 95% confidence upper bound of the
distributions for the linearly extrapolated risk per unit dose or exposure ranged from 1- to 8-fold
higher than the point unit risks derived using the BMDLs reported in Tables 5.2.5 and 5.2.6. The
largest differences, up to 4-fold for rat kidney tumors and 8-fold for mouse lung tumors,
primarily reflect the substantial uncertainty in the internal dose metrics for rat kidney DCVC and
GSH conjugation and for mouse lung oxidation (see Section 3.5). Additionally, despite the
differences in the degree of uncertainty due to the PBPK model across endpoints and dose
metrics, the only case where the choice of the most sensitive bioassay for each sex/species
combination would change based on the 95% confidence upper bounds reported in Tables 5.2.10
and 5.2.11 would be for female mouse inhalation bioassays. Even in this case, the difference
between unit risk estimate for the most sensitive and next most sensitive study/endpoint was only
2-fold.
6/22/2009
1149
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1
2
Table 5.2.10 Summary of PBPK model-based uncertainty analysis of unit risk estimates for each sex/species/bioassay/tumor
Study
Tumor Type
BMR
Dose Metric
Unit risk estimates ((mg/kg-d)')
From
Table
5.2.5
Summary statistics of unit risk distribution
Mean
5% lower
bound
Median
95% upper
bound
FEMALE
MOUSE
Fukuda
lung ad + carc
a
0.1
AMetLngBW34
2.6 x 10 3
5.65 x 10"3
2.34 x 10"4
1.49 x 10"3
2.18 x 10"2
T otOxMetabB W3 4
3.2 x 10"3
1.88 x 10"3
3.27 x 10"4
1.52 x 10"3
4.59 x 10"3
AUCCBld
1.8 x 10"3
1.01 x 10"3
1.54 x 10"4
8.36 x 10"4
2.44 x 10"3
Henschler
Lymphoma
b
0.1
TotMetabBW34
1.0 x 10 2
4.38 x 10"3
6.06 x 10"4
3.49 x 10"3
1.11 x 10"2
Maltoni
lung ad + carc
a
0.1
AMetLngBW34
1.8 x 10 3
3.88 x 10"3
1.48 x 10"4
1.04 x 10"3
1.52 x 10"2
TotOxMetabBW34
1.9 x 10"3
1.10 x 10"3
3.73 x 10"4
9.52 x 10"4
2.32 x 10"3
AUCCBld
1.0 x 10"3
5.25 x icr4
1.63 x lO"4
4.64 x icr4
l.io x icr3
liver
0.05
AMetLivlBW34
1.2 x 10 3
6.27 x lO"4
2.18 x lO"4
5.39 x 10"4
1.32 x icr3
TotOxMetabBW34
1.1 x 10"3
5.98 x lO"4
1.81 x 10"4
5.07 x icr4
1.31 x icr3
MALE
MOUSE
Maltoni
liver
0.1
AMetLivlBW34
2.6 x 10 3
1.35 x 10"3
4.28 x 10"4
1.16 x icr3
2.93 x icr3
TotOxMetabBW34
2.0 x 10"3
1.23 x 10"3
4.24 x 10"4
1.06 x 10"3
2.60 x icr3
MALE
RAT
Maltoni
Leukemia
b
0.05
TotMetabBW34
1.8 x 10 3
9.38 x 10"4
1.26 x 10"4
7.86 x lO"4
2.25 x icr3
kidney ad + carc
0.01
ABioactDCVCBW34
8.3 x 10 2
9.07 x 10"2
3.66 x 10"3
3.64 x 10"2
3.21 x icr1
AMetGSHBW34
5.1 x 10"2
3.90 x 10"2
2.71 x 10"3
2.20 x lO"2
1.30 x icr1
TotMetabBW34
7.3 x 10"4
3.94 x 10"4
8.74 x 10"5
3.42 x lO"4
8.74 x icr4
leydig cell
b
0.1
TotMetabBW34
5.5 x 10 3
4.34 x 10"3
1.99 x 10"3
3.98 x 10"3
7.87 x icr3
3
4
5
6
7
a WinBUGS dose-response analyses did not adequately converge for the AMetLngBW34 dose metric using
model (used for results in Table 5.2.5), but did converge when the 2nd-order model was used. Summary
2nd-order model calculations,
b Poor dose-response fits in point estimates for AUCCBld, so not included in uncertainty analysis.
the 3r -order multistage
statistics reflect results of
6/22/2009
1150
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.2.11 Summary of PBPK model-based uncertainty analysis of unit risk estimates for each sex/species/bioassay/tumor
type (oral) i i
Unit risk estimates ((mg/kg-d)')
From
Summary statistics of distribution
Study
Tumor Type
BMR
Dose Metric
Table
Mean
5% lower
Median
95% upper
5.2.6 or
bound
bound
5.2.7
FEMALE
MOUSE
NCI
liver carc
0.1
AMetLivlBW34
7.1 x 10 3
3.26 x 10"3
9.35 x 10"4
2.44 x 10"3
8.35 x 10"3
T otOxMetabB W3 4
5.7 x 10"3
2.63 x 10"3
8.76 x 10"4
2.01 x 10"3
6.60 x 10"3
lung ad + carc
a
0.1
AMetLngBW34
1.3 x 10 4
1.28 x 10"4
6.73 x 10"6
4.12 x 10"5
4.62 x 10"4
T otOxMetabB W3 4
4.0 x 10"3
1.84 x 10"3
5.29 x 10"4
1.39 x 10"3
4.73 x 10"3
AUCCBld
1.5 x 10"4
7.16 x 10"5
4.40 x 10"6
3.39 x 10"5
2.18 x 10"4
leukemias +
0.1
TotMetabBW34
4.9 x 10 3
1.60 x 10"3
1.42 x 10"4
1.13 x 10"3
4.65 x 10"3
sarcomas
AUCCBld
1.4 x 10"4
6.36 x 10"5
3.10 x 10"6
2.90 x 10"5
1.94 x 10"4
MALE
MOUSE
NCI
liver carc
0.1
AMetLivlBW34
2.9 x 10 2
1.65 x 10"2
4.70 x 10"3
1.25 x 10"2
4.25 x 10"2
TotOxMetabBW34
2.3 x 10"2
1.32 x 10"2
4.41 x 10"3
1.01 x 10"2
3.29 x 10"2
FEMALE
RAT
NTP 1988
leukemia
0.05
TotMetabBW34
2.3 x 10 3
1.89 x 10"3
5.09 x 10"4
1.43 x 10"3
4.69 x 10"3
AUCCBld
1.6 x 10"5
1.56 x 10"5
3.39 x 10"6
1.07 x 10"5
3.98 x 10"5
MALE
RAT
NTP 1990
kidney ad + carc
b
0.1
ABioactDCVCBW34
1.2 x 10 1
1.40 x 10"1
5.69 x 10"3
5.24 x 10"2
5.18 x 10"1
AMetGSHBW34
7.6 x 10"2
6.18 x 10"2
4.00 x 10"3
3.27 x 10"2
2.11 x 10"1
TotMetabBW34
3.1 x 10"3
2.49 x 10"3
7.14 x 10"4
1.96 x 10"3
5.96 x 10"3
NTP 1988
Marshall
testicular
b
0.1
TotMetabBW34
7.1 x 10 2
6.18 x 10"2
1.92 x 10"2
4.89 x 10"2
1.45 x 10"1
AUCCBld
6.0 x 10"4
5.45 x 10"4
1.18 x 10"4
3.70 x 10"4
1.44 x 10"3
August
subcut sarcoma
0.05
TotMetabBW34
2.3 x 10 3
1.65 x 10"3
4.58 x 10"4
1.27 x 10"3
4.04 x 10"3
AUCCBld
2.0 x 10"5
1.35 x 10"5
1.53 x 10"6
8.34 x 10"6
3.73 x 10"5
Osborne-
kidney ad + carc
b
0.1
ABioactDCVCBW34
1.6 x 10 1
1.61 x 10"1
5.45 x 10"3
6.35 x 10"2
6.02 x 10"1
Mendel
AMetGSHBW34
9.7 x 10"2
7.47 x 10"2
3.90 x 10"3
3.85 x 10"2
2.54 x 10"1
TotMetabBW34
4.3 x 10"3
2.73 x 10"3
5.40 x 10"4
2.10 x 10"3
6.89 x 10"3
a WinBUGS dose-response analyses did not adequately converge for AMetLngBW34 dose metric using the 3r -order multistage
model (used for results in Table 5.2.6), but did converge when the 2nd-order model was used. Summary statistics reflect results of 2nd-
order model calculations.
6/22/2009
1151
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1 b using poly-3 adjusted incidences from Table 5.2.7 (software for WinBUGS-based analyses using the MSW model was not
2 developed).
3
6/22/2009
1152
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.2.2 Dose-Response Analyses: Human Epidemiologic Data
Of the epidemiological studies of TCE and cancer, only one had sufficient exposure-
response information for dose-response analysis. This was the Charbotel et al. (2006) case-
control study of TCE and kidney cancer incidence, which was used to derive an inhalation unit
risk estimate for that endpoint (Section 5.2.2.1). Other epidemiological studies were used in
Section 5.2.2.2 below to provide information for a comparison of RR estimates across cancer
types. These epidemiologic data were used to derive an adjusted inhalation unit risk estimate for
the combined risk of developing kidney cancer, NHL, or liver cancer. The human PBPK model
was then used to perform route-to-route extrapolation to derive an oral unit risk estimate for the
combined risk of kidney cancer, NHL, or liver cancer (Section 5.2.2.3).
5.2.2.1 Inhalation Unit Risk Estimate for Renal Cell Carcinoma Derivedfrom Charbotel et
al. (2006) Data
The Charbotel et al. (2006) case-control study of 86 incident renal cell carcinoma (RCC)
cases and 316 age- and sex-matched controls, with individual cumulative exposure estimates for
TCE for each subject, provides a sufficient human dataset for deriving quantitative cancer risk
estimates for RCC in humans. The study is a high-quality study which used a detailed exposure
assessment (Fevotte et al., 2006) and took numerous potential confounding factors, including
exposure to other chemicals, into account (see Section 4.3). A significant dose-response
relationship was reported for cumulative TCE exposure and RCC (Charbotel et al., 2006).
The derivation of an inhalation unit risk estimate, defined as the plausible upper bound
lifetime risk of cancer from chronic inhalation of TCE per unit of air concentration, for RCC
incidence in the U.S. population, based on results of the Charbotel et al. case-control study, is
presented in the following subsections.
5.2.2.1.1 RCC results from the Charbotel et al. study
Charbotel et al. analyzed their data using conditional logistic regression, matching on sex
and age, and reported results (ORs) for cumulative TCE exposure categories, adjusted for
tobacco smoking and body mass index (Charbotel et al., 2006, Table 6). The exposure
categories were constructed as tertiles based on the cumulative exposure levels in the exposed
control subjects. The results are summarized in Table 5.2.12, with mean exposure levels kindly
provided by Dr. Charbotel (personal communication from Barbara Charbotel, University of
Lyon, to Cheryl Scott, U.S. EPA, 11 April 2008).
For additional details and discussion of the Charbotel et al. (2006) study, see Section 4.3
and Appendix B.
6/22/2009
1153
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.2.12. Results from Charbotel et al. on relationship between TCE exposure and
RCC
cumulative exposure category
mean cumulative exposure
adjusted OR
(ppm *y ears)
(95% CI)
non-exposed
1
low
62.4
1.62 (0.75, 3.47)
medium
253.2
1.15 (0.47, 2.77)
high
925.0
2.16(1.02,4.60)
5.2.2.1.2 Prediction of lifetime extra risk of RCC incidence from TCE exposure
The categorical results summarized in Table 5.2.12 were used for predicting the extra risk
of RCC incidence from continuous environmental exposure to TCE. Extra risk is defined as
Extra risk = (Rx-Ro)/(l-Ro),
where Rx is the lifetime risk in the exposed population and Ro is the lifetime risk in an
unexposed population (i.e., the background risk). Because kidney cancer is a rare event, the ORs
in Table 5.2.12 can be used as estimates of the relative risk ratio, RR = Rx/Ro (Rothman and
Greenland, 1998). A weighted linear regression model was used to model the dose-response data
in Table 5.2.12 to obtain a slope estimate (regression coefficient) for RR of RCC versus
cumulative exposure. This linear dose-response function was then used to calculate lifetime
extra risks in an actuarial program (lifetable analysis) that accounts for age-specific rates of death
and background disease, under the assumption that the RR is independent of age.52
For the weighted linear regression, the weights used for the RR estimates were the
inverses of the variances, which were calculated from the confidence intervals. Using this
approach,53 a linear regression coefficient of 0.001205 per ppm x year (SE = 0.0008195 per ppm
x year) was obtained from the categorical results.
For the lifetable analysis, U.S. age-specific all-cause mortality rates for 2004 for both
sexes and all race groups combined (NCHS, 2007) were used to specify the all-cause background
mortality rates in the actuarial program. Because we wish to estimate the unit risk for extra risk
52
This program is an adaptation of the approach previously used by the Committee on the Biological Effects of
Ionizing Radiation (BEIR, 1988). The same methodology was also used in EPA's 1,3-butadiene health risk
assessment (U.S. EPA, 2002). A spreadsheet illustrating the extra risk calculation for the derivation of the LEC0i
for RCC incidence is presented in Appendix H.
53
Equations for this weighted linear regression approach are presented in Rothman (1986) and summarized in
Appendix H.
6/22/2009
1154
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
of cancer incidence, not mortality, and because the Charbotel et al. data are incidence data, RCC
incidence rates were used for the cause-specific background "mortality" rates in the lifetable
analysis.54 Surveillance, Epidemiology, and End Results (SEER) 2001-2005 cause-specific
background incidence rates for RCC were obtained from the SEER public-use database.55 SEER
collects good-quality cancer incidence data from a variety of geographical areas in the United
States. The incidence data used here are from SEER 17, a registry of 17 states, cities, or regions
covering about 26% of the U.S. population (http://seer.cancer.gov). The risks were computed up
to age 85 years for continuous exposures to TCE56. Conversions between occupational TCE
exposures and continuous environmental exposures were made to account for differences in the
number of days exposed per year (240 vs. 365 days) and in the amount of air inhaled per day (10
vs. 20 m3; U.S. EPA, 1994). The standard error for the regression coefficient from the weighted
linear regression calculation described above was used to compute the 95% upper confidence
limit (UCL) for the slope estimate, and this value was used to derive 95% UCLs for risk
estimates (or 95% LCLs for corresponding exposure estimates), based on a normal
approximation.
Point estimates and one-sided 95% UCLs for the extra risk of RCC incidence associated
with varying levels of environmental exposure to TCE based on linear regression of the
Charbotel et al. (2006) categorical results were determined by the actuarial program; the results
are presented in 5.2.13. The models based on cumulative exposure yield extra risk estimates that
are fairly linear for exposures up to 1 ppm or so.
54 No adjustment was made for using RCC incidence rates rather than mortality rates to represent cause-specific
mortality in the actuarial program because the RCC incidence rates are negligible in comparison to the all-cause
mortality rates. Otherwise, all-cause mortality rates for each age interval would have been adjusted to reflect people
dying of a cause other than RCC or being diagnosed with RCC.
55 In accordance with the "SEER Program Coding and Staging Manual 2007"
(http://seer.cancer.gov/manuals/2007/SPCSM 2007 AppendixC p6.pdf). pages C-831 to C-833, RRC was
specified as ICD-0-3 histological types coded 8312, 8260, 8310, 8316-8320, 8510, 8959, and 8255 (mixed types).
56 Rates above age 85 years are not included because cause-specific disease rates are less stable for those ages. Note
that 85 years is not employed here as an average lifespan but, rather, as a cut-off point for the lifetable analysis,
which uses actual age-specific mortality rates.
6/22/2009
1155
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Table 5.2.13. Extra risk estimates for RCC incidence from various levels of lifetime
exposure to TCE, using linear cumulative exposure model
Exposure concentration
MLE of extra risk
95% UCL on extra risk
(ppm)
0.001
2.603 x 10"6
5.514 x 10"6
0.01
2.603 x 10"5
5.514 x 10"5
0.1
2.602 x 10"4
5.512 x 10"4
1.0
2.598 x 10"3
5.496 x 10"3
10.0
2.562 x 10"2
5.333 x 10"2
Consistent with EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
the same data and methodology were also used to estimate the exposure level (ECX: "effective
concentration") and the associated 95% lower confidence limit (LECX) corresponding to an extra
risk of 1% (x = 0.01). A 1% extra risk level is commonly used for the determination of the point
of departure (POD) for epidemiological data. Use of a 1% extra risk level for these data is
supported by the fact that, based on the actuarial program, the risk ratio (i.e., Rx/Ro) for an extra
risk of 1% for RCC incidence is 1.9, which is in the range of the ORs reported by Charbotel et al.
(Table 5.2.12). Thus, 1% extra risk was selected for determination of the POD, and, consistent
with the Guidelines for Carcinogen Risk Assessment, the LEC value corresponding to that risk
level was used as the actual POD. For the linear model that was selected, the unit risk is
independent of the benchmark risk level used to determine the POD (at low exposures/risk
levels; see Table 5.2.13); however, selection of a benchmark risk level is generally useful for
comparisons across models.
As discussed in Section 4.3, there is sufficient evidence to conclude that a mutagenic
MO A is operative for TCE-induced kidney tumors, which supports the use of linear low-dose
extrapolation from the POD. The ECoi, LECoi, and inhalation unit risk estimates for RCC
incidence using the linear cumulative exposure model are presented in Table 5.2.14. Converting
the units, 5.49 x 10"3 per ppm corresponds to a unit risk of 1.02 x 10"6 per (J,g/m3 for RCC
incidence.
Table 5.2.14. ECoi, LECoi, and unit risk estimates for RCC incidence, using linear
cumulative exposure model
ECoi (ppm)
LECoi (ppm)
unit risk (per ppm)a
3.87
1.82
5.49 x 10"3
a. unit risk = 0.01/LEC0i
6/22/2009
1156
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.2.2.1.3 Uncertainties in the RCC unit risk estimate
The two major sources of uncertainty in quantitative cancer risk estimates are generally
interspecies extrapolation and high-dose to low-dose extrapolation. The unit risk estimate for
RCC incidence derived from the Charbotel et al. (2006) results is not subject to interspecies
uncertainty because it is based on human data. A major uncertainty remains in the extrapolation
from occupational exposures to lower environmental exposures. There was some evidence of a
contribution to increased RCC risk from peak exposures; however, there remained an apparent
dose-response relationship for RCC risk with increasing cumulative exposure without peaks, and
the OR for exposure with peaks compared to exposure without peaks was not significantly
elevated (Charbotel et al., 2006). Although the actual exposure-response relationship at low
exposure levels is unknown, the conclusion that a mutagenic MOA is operative for TCE-induced
kidney tumors supports the linear low-dose extrapolation that was used (U.S. EPA, 2005a).
Another notable source of uncertainty in the cancer unit risk estimate is the dose-response
model used to model the study data to estimate the POD. A weighted linear regression across the
categorical ORs was used to obtain a slope estimate; use of a linear model in the observable
range of the data is often a good general approach for human data because epidemiological data
are frequently too limited (i.e., imprecise) to clearly identify an alternate model (U.S. EPA,
2005a). The Charbotel et al. study is a relatively small case-control study, with only 86 RCC
cases, 37 of which had TCE exposure; thus, the dose-response data upon which to specify a
model are indeed limited.
In accordance with U.S. EPA's Guidelines for Carcinogen Risk Assessment, the lower
bound on the ECoi is used as the POD; this acknowledges some of the uncertainty in estimating
the POD from the available dose-response data. In this case, the statistical uncertainty associated
with the ECoi is relatively small, as the ratio between the ECoi and the LECoi is about 2-fold.
The inhalation unit risk estimate of 5.49 x 10"3 per ppm presented above, which is calculated
based on a linear extrapolation from the POD (LECoi), is expected to provide an upper bound on
the risk of cancer incidence. [However, for certain applications, such as benefit-cost analyses,
estimates of "central tendency" for the risk below the POD are desired. Because a linear dose-
response model was used in the observable range of the human data and the POD was within the
low-dose linear range for extra risk as a function of exposure, linear extrapolation below the
LECoi has virtually the same slope as the 95% UCL on the actual (linear) dose-response model
in the low-dose range (i.e., below the POD). This is illustrated in Table 5.2.13, where the 95%
UCL on extra risk for RCC incidence predicted by the dose-response model is about 5.51 x 10"3
per ppm for exposures at or below about 0.1 ppm, which is virtually equivalent to the unit risk
estimate of 5.49 x 10"3 per ppm derived from the LECoi ( Table 5.2.14). The same holds for the
central tendency (weighted least squares) estimates of the extra risk from the (linear) dose-
6/22/2009
1157
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
response model (i.e., the dose-response model prediction of 2.60 x 10"3 per ppm from Table
5.2.13 is virtually identical to the value of 2.58 x 10"3 per ppm obtained from linear extrapolation
below the ECoi, i.e., by dividing 0.01 extra risk by the ECoi of 3.87 from Table 5.2.14). In other
words, because the dose-response model that was used to model the data in the observable range
is already low-dose linear near the POD, if one assumes that the same linear model is valid for
the low-dose range, one can use the central tendency (weighted least squares) estimates from the
model to derive a statistical "best estimate" of the slope rather than relying on an extrapolated
risk estimates (0.01/ECoi). The extrapolated risk estimates are not generally central tendency
estimates in any statistical sense because once risk is extrapolated below the ECoi using the
formulation 0.01/ECoi, it is no longer a function of the original model which generated the ECoiS
and the LECois.]
An important source of uncertainty in the underlying Charbotel et al. (2006) study is the
retrospective estimation of TCE exposures in the study subjects. This case-control study was
conducted in the Arve Valley in France, a region with a high concentration of workshops
devoted to screw cutting, which involves the use of TCE and other degreasing agents. Since the
1960s, occupational physicians of the region have collected a large quantity of well-documented
measurements, including TCE air concentrations and urinary metabolite levels (Fevotte et al.,
2006). The study investigators conducted a comprehensive exposure assessment to estimate
cumulative TCE exposures for the individual study subjects, using a detailed occupational
questionnaire with a customized task-exposure matrix for the screw-cutting workers and a more
general occupational questionnaire for workers exposed to TCE in other industries (Fevotte et
al., 2006). The exposure assessment even attempted to take dermal exposure from hand-dipping
practices into account by equating it with an equivalent airborne concentration based on
biological monitoring data. Despite the appreciable effort of the investigators, considerable
uncertainty associated with any retrospective exposure assessment is inevitable, and some
exposure misclassification is unavoidable. Such exposure misclassification was most likely for
the 19 deceased cases and their matched controls, for which proxy respondents were used, and
for exposures outside the screw-cutting industry (295 of 1486 identified job periods involved
TCE exposure; 120 of these were not in the screw-cutting industry).
Another noteworthy source of uncertainty in the Charbotel et al. (2006) study is the
possible influence of potential confounding or modifying factors. This study population, with a
high prevalence of metal-working, also had relatively high prevalences of exposure to petroleum
oils, cadmium, petroleum solvents, welding fumes, and asbestos (Fevotte et al., 2006). Other
exposures assessed included other solvents (including other chlorinated solvents), lead, and
ionizing radiation. None of these exposures was found to be significantly associated with RCC
at a p=0.05 significance level. Cutting fluids and other petroleum oils were associated with RCC
6/22/2009
1158
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
at a p=0.1 significance level; however, further modeling suggested no association with RCC
when other significant factors were taken into account (Charbotel et al., 2006). The medical
questionnaire included familial kidney disease and medical history, such as kidney stones,
infection, chronic dialysis, hypertension, and use of anti-hypertensive drugs, diuretics, and
analgesics. Body mass index (BMI) was also calculated, and lifestyle information such as
smoking habits and coffee consumption was collected. Univariate analyses found high levels of
smoking and BMI to be associated with increased odds of RCC, and these two variables were
included in the conditional logistic regressions. Thus, although impacts of other factors are
possible, this study took great pains to attempt to account for potential confounding or modifying
factors.
Some other sources of uncertainty associated with the epidemiological data are the dose
metric and lag period. As discussed above, there was some evidence of a contribution to
increased RCC risk from peak TCE exposures; however, there appeared to be an independent
effect of cumulative exposure without peaks. Cumulative exposure is considered a good
measure of total exposure because it integrates exposure (levels) over time. If there is a
contributing effect of peak exposures, not already taken into account in the cumulative exposure
metric, the linear slope may be overestimated to some extent. Sometimes cancer data are
modeled with the inclusion of a lag period to discount more recent exposures not likely to have
contributed to the onset of cancer. In an unpublished report (Charbotel et al., 2005), Charbotel et
al. also present the results of a conditional logistic regression with a 10-year lag period, and these
results are very similar to the untagged results reported in their published paper, suggesting that
the lag period might not be an important factor in this study.
Some additional sources of uncertainty are not so much inherent in the exposure-response
modeling or in the epidemiologic data themselves but, rather, arise in the process of obtaining
more general Agency risk estimates from the epidemiologic results. EPA cancer risk estimates
are typically derived to represent an upper bound on increased risk of cancer incidence for all
sites affected by an agent for the general population. From experimental animal studies, this is
accomplished by using tumor incidence data and summing across all the tumor sites that
demonstrate significantly increased incidences, customarily for the most sensitive sex and
species, to attempt to be protective of the general human population. However, in estimating
comparable risks from the Charbotel et al. (2006) epidemiologic data, certain limitations are
encountered. For one thing, these epidemiology data represent a geographically limited (Arve
Valley, France) and likely not very diverse population of working adults. Thus, there is
uncertainty about the applicability of the results to a more diverse general population.
Additionally, the Charbotel et al. (2006) study was a study of RCC only, and so the risk estimate
6/22/2009
1159
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
derived from it does not represent all the tumor sites that may be affected by TCE. We address
the issue of cancer risk at other sites in the next section (Section 5.2.2.2).
5.2.2.1.4 Conclusions regarding the RCC unit risk estimate
An ECoi of 3.9 ppm was calculated using a life-table analysis and linear modeling of the
categorical conditional logistic regression results for RCC incidence reported in a high-quality
case-control study. Linear low-dose extrapolation from the LECoi yielded a lifetime extra RCC
incidence unit risk estimate of 5.5 x 10"3 per ppm (1.0 x 10"6 per |ig/m3) of continuous TCE
exposure. The assumption of low-dose linearity is supported by the conclusion that a mutagenic
MOA is operative for TCE-induced kidney tumors. The inhalation unit risk estimate is expected
to provide an upper bound on the risk of RCC incidence; however, this is just the risk estimate
for RCC. A risk estimate for total cancer risk to humans would need to include the risk for other
potential TCE-associated cancers.
5.2.2.2 Adjustment of the Inhalation Unit Risk Estimate for Multiple Sites
Human data on TCE exposure and cancer risk sufficient for dose-response modeling are
only available for RCC, yet human and rodent data suggest that TCE exposure increases the risk
of cancer at other sites as well. In particular, there is evidence from human (and rodent) studies
for increased risks of lymphoma and liver cancer (see Section 4.10). Therefore, the inhalation
unit risk estimate derived from human data for RCC incidence was adjusted to account for
potential increased risk of those tumor types. To make this adjustment, a factor accounting for
the relative contributions to the extra risk for cancer incidence from TCE exposure for these
three tumor types combined versus the extra risk for RCC alone was estimated, and this factor
was applied to the unit risk estimate for RCC to obtain a unit risk estimate for the three tumor
types combined (i.e., lifetime extra risk for developing any of the 3 types of tumor). This
estimate is considered a better estimate of total cancer risk from TCE exposure than the estimate
for RCC alone.
Although only the Charbotel et al. (2006) study was found adequate for direct estimation
of inhalation unit risks, the available epidemiologic data provide sufficient information for
estimating the relative potency of TCE across tumor sites. In particular, the relative
contributions to extra risk (for cancer incidence) were calculated from two different datasets to
derive the adjustment factor for adjusting the unit risk estimate for RCC to a unit risk estimate
for the 3 types of cancers (RCC, lymphoma, and liver) combined. The first calculation is based
on the results of the meta-analyses of human epidemiologic data for the 3 tumor types (see
Appendix C); the second calculation is based on the results of the Raaschou-Nielsen et al. (2003)
study, the largest single human epidemiologic study by far with RR estimates for all 3 tumor
types. The approach for each calculation was to use the RR estimates and estimates of the
6/22/2009
1160
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
lifetime background risk in an unexposed population, Ro, to calculate the lifetime risk in the
exposed population, Rx, where Rx = RR * Ro, for each tumor type. Then, the extra risk from
TCE exposure for each tumor type could be calculated using the equation in Section 5.2.2.1.2.
Finally, the extra risks were summed across the 3 tumor types and the ratio of the sum of the
extra risks to the extra risk for RCC was derived. For the first calculation, the pooled relative
risk estimates (RRp's) from the meta-analyses for lymphoma, kidney cancer, and liver (&
biliary) cancer were used as the RR estimates. For the second calculation, the SIR estimates
from the Raaschou-Nielsen et al. (2003) study were used. For both calculations, Ro for RCC
was taken from the lifetable analysis described in Section 5.2.2.1.2 and presented in Appendix H,
which estimated a lifetime risk for RCC incidence up to age 85 years. For Ro values for the
other 2 sites, SEER statistics for the lifetime risk of developing cancer were used
(http://seer.cancer.gov/statfacts/html/nhl.html and
http://seer.cancer.uov/statfacts/html/livibd.html).
In both cases, an underlying assumption in deriving the relative potencies is that the
relative values of the age-specific background incidence risks for the person-years from the
epidemiologic studies for each tumor type approximate the relative values of the lifetime
background incidence risks for those tumor types. In other words, at least on a proportional
basis, the lifetime background incidence risks (for the U.S. population) for each site approximate
the age-specific background incidence risks for the study populations. A further assumption is
that the lifetime risk of RCC up to 85 years is an adequate approximation to the full lifetime risk,
which is what was used for the other 2 tumor types. The first calculation, based on the results of
the meta-analyses for the 3 tumor types, has the advantage of being based on a large dataset,
incorporating data from many different studies. However, this calculation relies on a number of
additional assumptions. First, it is assumed that the RRp's from the meta-analyses, which are
based on different groups of studies, reflect similar overall TCE exposures, i.e., that the overall
TCE exposures are similar across the different groups of studies that went into the different
meta-analyses for the 3 tumor types. Second, it is assumed that the RRp's, which incorporate
RR estimates for both mortality and incidence, represent good estimates for cancer incidence risk
from TCE exposure. In addition, it is assumed that the RRp for kidney cancer, for which RCC
estimates from individual studies were used when available, is a good estimate for the overall RR
for RCC and that the RRp estimate for lymphoma, for which different studies used different
classification schemes, is a good estimate for the overall RR for NHL. The second calculation,
based on the results of the Raaschou-Nielsen et al. (2003) study, the largest single study with RR
estimates for all 3 tumor types, has the advantage of having RR estimates that are directly
comparable. In addition, the Raaschou-Nielsen et al. study provided data for the precise tumor
types of interest for the calculation, i.e., RCC, NHL, and liver (& biliary) cancer.
6/22/2009
1161
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
The input data and results of the calculations are presented in Table 5.2.15. The value for
the ratio of the sum of the extra risks to the extra risk for RCC alone was 3.83 in calculation #1
and 4.36 in calculation #2, which together suggest that 4 is a reasonable factor to use to adjust
the inhalation unit risk estimate based on RCC for multiple sites to obtain a total cancer unit risk
estimate. Using this factor to adjust the unit risk estimate based on RCCs entails the further
fundamental assumption that the dose-response relationships for the other two tumor types (NHL
and liver cancer) are similarly linear, i.e., that the relative potencies are roughly maintained at
lower exposure levels. This assumption is consistent with U.S. EPA's Guidelines for
Carcinogen RiskAssessmen (U.S. EPA, 2005a), which recommends low-dose linear
extrapolation in the absence of sufficient evidence to support a non-linear MO A.
Applying the factor of 4 to the lifetime extra RCC incidence unit risk estimate of 5.49 x
10"3 per ppm (1.0 x 10"6 per |ig/m3) of continuous TCE exposure yields a cancer unit risk
estimate of 2.2 x 10"2 per ppm (4.1 x 10"6 per |ig/m3). Table 5.2.15 also presents calculations for
just kidney and lymphoma extra risks combined, because the strongest human evidence is for
those 2 tumor types. For those 2 tumor types, the calculations support a factor of 3. Applying
this factor to the RCC unit risk estimate yields an estimate of 1.6 x 10"2 per ppm, which results in
the same estimate as for the 3 tumor types combined when finally rounded to one significant
figure, i.e., 2 x 10"2 per ppm (or 3 x 10"6 per |ig/m3, which is still similar to the 3-tumor-type
estimate in those units).
Table 5.2.15. Relative contributions to extra risk for cancer incidence from TCE exposure
for multiple tumor types
RR Ro
Rx
extra risk
ratio to
kidney value
Calculation #1: using RR estimates from the meta-analyses
kidney (RCC)
lymphoma (NHL)
liver (& biliary) cancer
1.26 0.0107
1.27 0.0202
1.36 0.0066
0.01348
0.02565
0.008976
0.002812
0.005566
0.002392
1
1.98
0.85
sum
0.01077
3.83
kidney + NHL only
sum
0.008379
2.98
Calculation #2: using RR estimates from Rasschou-Nielsen etal. (2003)
kidney (RCC)
lymphoma (NHL)
liver (& biliary) cancer
1.20 0.0107
1.24 0.0202
1.35 0.0066
0.01284
0.02505
0.008910
0.002163
0.004948
0.002325
1
2.29
1.07
6/22/2009
1162
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
sum
0.009436
4.36
kidney + NHL only
sum
0.007111
3.29
In addition to the uncertainties in the underlying RCC estimate, there are uncertainties
related to the assumptions inherent in these calculations for adjusting to multiple sites, as
detailed above. Nonetheless, the fact that the calculations based on 2 different datasets yielded
comparable values for the adjustment factor provides more robust support for the use of the
factor of 4. Additional uncertainties pertain to the weight of evidence supporting the association
of TCE exposure with increased risk of cancer for the 3 tumor types. As discussed in Section
4.10.2, we found that the weight of evidence for kidney cancer was sufficient to classify TCE as
"carcinogenic to humans". We also concluded that there was strong evidence that TCE causes
NHL as well, although the evidence for liver cancer was more limited. In addition, the rodent
studies demonstrate clear evidence of multi-site carcinogenicity, with tumor types including
those for which associations with TCE exposure are observed in human studies, i.e., liver and
kidney cancers and lymphomas. Overall, we find the evidence sufficiently persuasive to support
the use of the adjustment factor of 4 based on these 3 tumor types, resulting in a cancer
inhalation unit risk estimate of 2.2 x 10"2 per ppm (4.1 x 10"6 per |ig/m3). Alternatively, if one
were to use the factor based only on the 2 tumor types with the strongest evidence, the cancer
inhalation unit risk estimate would be only slightly reduced (25%).
5.2.2.3 Route-to-route extrapolation using PBPK model
Route-to-route extrapolation of the inhalation unit risk estimate was performed using the
PBPK model described in Section 3.5. The (partial) unit risk estimates for lymphoma and liver
cancer were derived as for the total cancer inhalation unit risk estimate in Section 5.2.2.2 above,
except that the ratios of extra risk for the individual tumor types relative to kidney cancer were
used as adjustment factors rather than the ratio of the sum. As presented in Table 5.2.15, for
lymphoma, the ratios from the 2 different calculations were 1.98 and 2.29, so a factor of 2 was
used; for liver cancer, the ratios were 0.85 and 1.07, so a factor of 1 was used. With the ratio of
1 for kidney cancer itself, the combined adjustment factor is 4, consistent with the factor of 4
used to estimate the total cancer unit risk from the multiple sites in Section 5.2.2.2.
Because different internal dose metrics are preferred for each target tissue site, a separate
route-to-route extrapolation was performed for each site-specific unit risk estimate calculated in
sections 5.2.2.1 and 5.2.2.2. As shown in Figure 5.2.3, the approach taken to apply the human
PBPK model in the low dose range where external and internal doses are linearly related to
6/22/2009
1163
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
derive a conversion that is the ratio of internal dose per mg/kg/d to internal dose per ppm. The
expected value of the population mean for this conversion factor (in ppm per mg/kg/d) was used
to extrapolate each inhalation unit risk in units of risk per ppm to an oral slope factor in units of
risk per mg/kg/d. Note that this conversion is the mean of the ratio of internal dose predictions,
and is not the same as taking the ratio of the mean of internal dose predictions in Table 5.2.4.57
Table 5.2.16 shows the results of this route-to-route extrapolation for the "primary" and
"alternative" dose metrics. For reference, route-to-route extrapolation based on total intake (i.e.,
ventilation rate * air concentration = oral dose * BW) using the parameters in the PBPK model
would yield an expected population average conversion of 0.95 ppm per mg/kg/d. For
TotMetabBW34, TotOxMetabBW34, and AMetLivlBW34, the conversion is 2.0-2.8 ppm per
mg/kg/d, greater than that based on intake. This is because of the greater metabolic first pass in
the liver, which leads to a higher percentage of intake being metabolized via oral exposure
relative to inhalation exposure for the same intake. Conversely, for the AUC in blood, the
conversion is 0.14 ppm per mg/kg/d, less than that based on intake - the greater first pass in the
liver means lower blood levels of parent compound via oral exposure relative to inhalation for
the same intake. The conversion for the primary dose metric for the kidney,
ABioactDCVCBW34, is 1.7 ppm per mg/kg/d, less than that for total, oxidative, or liver
oxidative metabolism. This is because the majority of metabolism in first pass through the liver
is via oxidation, whereas with inhalation exposure, more parent compound reaches the kidney, in
which metabolism is via GSH conjugation.
When one sums the oral slope factor estimates based on the primary (preferred) dose
metrics for the 3 individual tumor types shown in Table 5.2.16, the resulting total cancer oral
unit risk (slope factor) estimate is 4.63 x 10"2 per mg/kg/day. In the case of the oral route-
extrapolated results, the ratio of the risk estimate for the 3 tumor types combined to the risk
estimate for kidney cancer alone is 5.0. This value differs from the factor of 4 used for the total
cancer inhalation unit risk estimate because of the different dose metrics used for the different
tumor types when the route-to-route extrapolation is performed. If only the kidney cancer and
NHL results, for which the evidence is strongest, were combined, the resulting total cancer oral
unit risk estimate would be 3.08 x 10"2 per mg/kg/day, and the ratio of this risk estimate to that
for kidney cancer alone would be 3.3.
If one were to use some of the risk estimates based on alternative dose metrics in Table
5.2.16, the total cancer risk estimate would vary depending on for which tumor type(s) an
alternative metric was used. The most extreme difference would occur when the alternative
57 For route-to-route extrapolation based on dose-response analysis performed on internal dose, as is the case for
rodent bioassays, it would be appropriate to use the values in Table 5.2.4 to first "unconvert" the unit risk based on
one route, and then recovert to a unit risk based on the other route.
6/22/2009
1164
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
metric is used for NHL and liver tumors; in that case, the resulting total cancer oral unit risk
estimate would be 2.20 x 10"2 per mg/kg/day, and the ratio of this risk estimate to that for kidney
cancer alone (based on the primary dose metric of ABioactDCVCBW34) would be 2.4.
The uncertainties in these conversions are relatively modest. As discussed in the note to
Table 5.2.16, the 95% confidence range for the route-to-route conversions at its greatest spans
3.4-fold. The greatest uncertainty is in the selection of the dose metric for NHL, since the use of
the alternative dose metric of AUCCBld yields a converted oral slope factor that is 14-fold lower
than that using the primary dose metric of TotMetabBW34. However, for the other two tumor
sites, the range of conversions is tighter, and lies within 3-fold of the conversion based solely on
intake.
0.001 ppm
in air or
0.001 mg/kg-d
continuous
exposure
PBPK
model
[distribution (separate
^uncertainty and variability)
/ [internal dosl
Site-specific
human
cancer
unit risk
per ppm
Human
model
parameters
per mg/kg/d]/i
[internal dose/
vper ppm]
"[population
ynean
i
^|ne
Expected
site-specific
human
cancer unit risk
per mg/kg-d
Figure 5.2.3.
Flow-chart for route-to-route extrapolation of human site-specific cancer inhalation unit risks to
oral slope factors. Square nodes indicate point values, circle nodes indicate distributions, and the
inverted triangle indicates a (deterministic) functional relationship.
6/22/2009
1165
-------
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
1
2 Table 5.2.16. Route-to-route extrapolation of site-specific inhalation unit risks to oral slope
3 factors.
Kidney
NHL
Liver
Inhalation unit risk
5.49 x 10"3
1.09 x 10"2
5.49 x 10"3
(risk per ppm)
Primary dose metric
ABioactDCVCBW34
a
TotMetabBW34
AMetLivlBW34
ppm per mg/kg/db
1.70
1.97
2.82
Oral slope factor
9.33 x 10"3
2.15 x 10"2
1.55 x 10"2
(risk per mg/kg/d)
Alternative dose metric
TotMetabBW34
AUCCBld
T otOxMetabBW 3 4
ppm per mg/kg-db
1.97
0.137
2.04
Oral slope factor
1.08 x 10"2
1.49 x 10"3
1.12 x 10"2
(risk per mg/kg-d)
4 a The AMetGSHBW34 dose metric gives the same route-to-route conversion because there is no route
5 dependence in the pathway between GSH conjugation and DCVC bioactivation.
6 b Average of expected population mean of males and females. Male and female estimates differed by
7 <1% for ABioactDCVCBW34; TotMetabBW34, AMetLivlBW34, and TotOxMetabBW34, and <15%
8 for AUCCBld. Uncertainty on the population mean route-to-route conversion, expressed as the ratio
9 between the 97.5% quantile the 2.5% quantile, is about 2.6-fold for ABioactDCVCBW34, 1.5-fold for
10 TotMetabBW34, AMetLivlBW34, and TotOxMetabBW34, and about 3.4-fold for AUCCBld.
11 5.2.3 Summary of unit risk estimates
12 5.2.3.1 Inhalation unit risk estimate
13 The inhalation unit risk for TCE is defined as a plausible upper bound lifetime extra risk
14 of cancer from chronic inhalation of TCE per unit of air concentration. The preferred estimate of
15 the inhalation unit risk for TCE is 2.20 x 10~2 per ppm (2 x 10~2 per ppm [4 x 10~6 per jig/m3]
16 rounded to 1 significant figure), based on human kidney cancer risks reported by Charbotel et al.
17 (2006) and adjusted for potential risk for tumors at multiple sites. This estimate is based on
18 good-quality human data, thus avoiding the uncertainties inherent in interspecies extrapolation.
19 This value is supported by inhalation unit risk estimates from multiple rodent bioassays,
20 the most sensitive of which range from 1 x 10~2 to 2 x 10_1 per ppm [2 x 10~6 to 3 x 10~5 per
21 jng/m3| From the inhalation bioassays selected for analysis in section 5.2.1.1, and using the
22 preferred PBPK model-based dose metrics, the inhalation unit risk estimate for the most sensitive
23 sex/species is 8 x 10~2 per ppm [2 x 10~5 per |ig/m3], based on kidney adenomas and carcinomas
24 reported by Maltoni et al. (1986) for male Sprague-Dawley rats. Leukemias and Ley dig cell
25 tumors were also increased in these rats, and, although a combined analysis for these tumor types
6/22/2009
1166
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
which incorporated the different site-specific preferred dose metrics was not performed, the
result of such an analysis is expected to be similar, about 9 x 10 2 per ppm [2 x 10 5 per |ig/m3].
The next most sensitive sex/species from the inhalation bioassays is the female mouse, for which
lymphomas were reported by Henschler et al. (1980); these data yield a unit risk estimate of
1.0 x 10~2 per ppm [2 x 10-6 per |ig/m3]. In addition, the 90% confidence intervals reported in
Table 5.2.10 for male rat kidney tumors from Maltoni et al. (1986) and female mouse
lymphomas from Henschler et al. (1980), derived from the quantitative analysis of PBPK model
uncertainty, both included the estimate based on human data of 2 x 10 2 per ppm. Furthermore,
PBPK model-based route-to-route extrapolation of the results for the most sensitive sex/species
from the oral bioassays, kidney tumors in male Osborne-Mendel rats and testicular tumors in
Marshall rats (NTP 1988), leads to inhalation unit risk estimates of 2 x 10 1 per ppm [3 x 10 5
per |ig/m3] and 4 x 10 2 per ppm [8 x 10 6 per |ig/m3], respectively, with the preferred estimate
based on human data falling within the route-to-route extrapolation of the 90% confidence
intervals reported in Table 5.2.11.58 Finally, for all these estimates, the ratios of BMDs to the
BMDLs did not exceed a value of 3, indicating that the uncertainties in the dose-response
modeling for determining the POD in the observable range are small.
Although there are uncertainties in these various estimates, as discussed in Sections
5.2.1.4, 5.2.2.1.3, and 5.2.2.2, confidence in the proposed inhalation unit risk estimate of 2 x 10~2
per ppm [4 x 10 6 per |ig/m3], based on human kidney cancer risks reported by Charbotel et al.
(2006) and adjusted for potential risk for tumors at multiple sites (as discussed in Section
5.2.2.2), is further increased by the similarity of this estimate to estimates based on multiple
rodent datasets.
5.2.3.2 Oral unit risk estimate
The oral unit risk (or slope factor) for TCE is defined as a plausible upper bound lifetime
extra risk of cancer from chronic ingestion of TCE per mg/kg/d oral dose. The preferred
estimate of the oral unit risk is 4.63 x 10 2 per mg/kg/d (5 x 10~2 per mg/kg/d rounded to 1
significant figure), resulting from PBPK model-based route-to-route extrapolation of the
inhalation unit risk estimate based on the human kidney cancer risks reported in Charbotel et al.
(2006) and adjusted for potential risk for tumors at multiple sites. This estimate is based on
58 For oral-to-inhalation extrapolation of NTP (1988) male rat kidney tumors, the unit risk estimate of 2.5 x 10"1 per
mg/kg/d using the ABioactDCVCBW34 dose metric, from Table 5.2.6, is divided by the average male and female
internal doses at 0.001 mg/kg/d, (0.00504/0.001), and then multiplied by the average male and female internal doses
at 0.001 ppm, (0.00324/0.001), both from Table 5.2.4, to yield a unit risk of 1.6 x 10"1 [3.0 x 10 s per |ig/m31. For
oral-to-inhalation extrapolation of NTP (1988) male rat testicular tumors, the unit risk estimate of 7.1 x 10"2 per
mg/kg/d using the TotMetabBW34 dose metric, from Table 5.2.6, is divided by the male internal dose at 0.001
mg/kg/d, (0.0192/0.001), and then multiplied by the male internal doses at 0.001 ppm, (0.0118/0.001), both from
Table 5.2.4, to yield a unit risk of 4.4 x 10"2 [8.1 x 10 6 per |ig/m31.
6/22/2009
1167
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
good-quality human data, thus avoiding uncertainties inherent in interspecies extrapolation. In
addition, uncertainty in the PBPK model-based route-to-route extrapolation is relatively low
(Chiu and White 2006, Chiu 2006). In this particular case, extrapolation using different dose
metrics yielded expected population mean risks within about a 2-fold range, and, for any
particular dose metric, the 95% confidence interval for the extrapolated population mean risks
for each site spanned a range of no more than about 3-fold.
This value is supported by oral unit risk estimates from multiple rodent bioassays, the
most sensitive of which range from 3 x 10~2 to 3 x 10-1 per mg/kg/d. From the oral bioassays
selected for analysis in section 5.2.1.1, and using the preferred PBPK model-based dose metrics,
the oral unit risk estimate for the most sensitive sex/species is 3 x 10 1 per mg/kg/d, based on
kidney tumors in male Osborne-Mendel rats (NTP 1988). The oral unit risk estimate for
testicular tumors in male Marshall rats (NTP 1988) is somewhat lower at 7 x 10 2 per mg/kg/d.
The next most sensitive sex/species result from the oral studies is for male mouse liver tumors
(NCI 1976), with an oral unit risk estimate of 3 x 10 2 per mg/kg/d. In addition, the 90%
confidence intervals reported in Table 5.2.11 for male Osborne-Mendel rat kidney tumors (NTP
1988), male F344 rat kidney tumors (NTP 1990), and male Marshall rat testicular tumors (NTP
1988), derived from the quantitative analysis of PBPK model uncertainty, all included the
estimate based on human data of 5 x 10 2 per mg/kg/d, while the upper 95% confidence bound
for male mouse liver tumors from NCI (1976) was slightly below this value at 4 x 10 2 per
mg/kg/d. Furthermore, PBPK model-based route-to-route extrapolation of the most sensitive
endpoint from the inhalation bioassays, male rat kidney tumors from Maltoni et al. (1986), leads
to an oral unit risk estimate of 1 x 10 1 per mg/kg/d, with the preferred estimate based on human
data falling within the route-to-route extrapolation of the 90% confidence interval reported in
Table 5.2.10.59 Finally, for all these estimates, the ratios of BMDs to the BMDLs did not exceed
a value of 3, indicating that the uncertainties in the dose-response modeling for determining the
POD in the observable range are small.
Although there are uncertainties in these various estimates, as discussed in Sections
5.2.1.4, 5.2.2.1.3, 5.2.2.2, and 5.2.2.3, confidence in the proposed oral unit risk estimate of
5 x 10~2 per mg/kg/d, resulting from PBPK model-based route-to-route extrapolation of the
inhalation unit risk estimate based on the human kidney cancer risks reported in Charbotel et al.
(2006) and adjusted for potential risk for tumors at multiple sites (as discussed in Section
59 For the Maltoni et al. (1986) male rat kidney tumors, the unit risk estimate of 8.3 x 10"2 per ppm using the
ABioactDCVCBW34 dose metric, from Table 5.2.5, is divided by the average male and female internal doses at
0.001 ppm, (0.00324/0.001) and then multiplied by the average male and female internal doses at 0.001 mg/kg/d,
(0.00504/0.001), both from Table 5.2.4, to yield a unit risk of 1.3 x 10"1 per mg/kg/d.
6/22/2009
1168
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.2.2.2), is further increased by the similarity of this estimate to estimates based on multiple
rodent datasets.
5.2.3.3 Application of age-dependent adjustment factors
When there is sufficient weight of evidence to conclude that a carcinogen operates
through a mutagenic MO A, and in the absence of chemical-specific data on age-specific
susceptibility, U.S. EPA's Supplemental Guidance for Assessing Susceptibility from Early-Life
Exposure to Carcinogens (U.S. EPA, 2005b) recommends the application of default age-
dependent adjustment factors (ADAFs) to adjust for potential increased susceptibility from early-
life exposure. See the Supplemental Guidance for detailed information on the general
application of these adjustment factors. In brief, the Supplemental Guidance establishes ADAFs
for three specific age groups. The current ADAFs and their age groupings are 10 for <2 years, 3
for 2 to <16 years, and 1 for 16 years and above (U.S. EPA, 2005b). For risk assessments based
on specific exposure assessments, the 10-fold and 3-fold adjustments to the unit risk estimates
are to be combined with age-specific exposure estimates when estimating cancer risks from
early-life (<16 years age) exposure. The ADAFs and their age groups may be revised over time.
The most current information on the application of ADAFs for cancer risk assessment can be
found at www.epa.gov/cancerguidelines.
In the case of TCE, the inhalation and oral unit risk estimates reflect lifetime risk for
cancer at multiple sites, and a mutagenic MOA has been established for one of these sites, the
kidney. The following subsections illustrate how one might apply the ADAFs to the inhalation
and oral unit risk estimates for TCE. These are sample calculations, and individual risk
assessors should use exposure-related parameters (e.g., age-specific water ingestion rates) that
are appropriate for their particular risk assessment applications.
In addition to the uncertainties discussed above for the inhalation and oral total cancer
unit risk estimates, there are uncertainties in the application of ADAFs to adjust for potential
increased early-life susceptibility. For one thing, the adjustment is made only for the kidney
cancer component of total cancer risk because that is the tumor type for which the weight of
evidence was sufficient to conclude that TCE-induced carcinogenesis operates through a
mutagenic MOA. However, it may be that TCE operates through a mutagenic MOA for other
tumor types as well or that it operates through other MO As that might also convey increased
early-life susceptibility. Additionally, the ADAFs are general default factors, and it is uncertain
to what extent they reflect increased early-life susceptibility for exposure to TCE, if increased
early-life susceptibility occurs.
6/22/2009
1169
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
5.2.3.3.1 Example application of ADAFs for inhalation exposures
For inhalation exposures, assuming ppm equivalence across age groups, i.e., equivalent
risk from equivalent exposure levels, independent of body size, the calculation is fairly
straightforward. The ADAF-adjusted lifetime cancer unit risk estimate for kidney cancer alone
is calculated as follows:
kidney cancer risk from exposure to constant TCE exposure level of 1 |ig/m3 from ages 0-70:
unit risk
exposure
duration
partial
Aee eroup
ADAF
(per ng/m3)
cone (us/m3)
adjustment
risk
0 - < 2 years
10
1.0 x 10"6
1
2 years/70 years
2.9 x 10"7
2 - < 16 years
3
1.0 x 10"6
1
14 years/70 years
6.0 x 10"7
>16 years
1
1.0 x 10"6
1
54 years/70 years
7.7 x 10"7
total risk =
1.7 x 10"6
Note that the partial risk for each age group is the product of the values in columns 2-5 [e.g., 10 x
(1.0 x 10"6) x 1 x 2/70 = 2.9 x 10"7], and the total risk is the sum of the partial risks. This 70-year
risk estimate for a constant exposure of 1 (.ig/nr3 is equivalent to a lifetime unit risk of 1.7 x 10"6
per |ig/m3. adjusted for early-life susceptibility, assuming a 70-year lifetime and constant
exposure across age groups.
In other words, the lifetime unit risk estimate for kidney cancer alone, adjusted for
potential increased early-life susceptibility is 1.7-times the unadjusted unit risk estimate. Adding
a 3-fold factor to the unadjusted unit risk estimate to account for potential risk at multiple sites
("1-fold" of the factor of 4 for multiple sites is already included in the 1.7-times adjustment for
early-life susceptibility) yields a total adjustment factor of 4.7. Applying a factor of 4.7 to the
unit risk estimate based on kidney cancer alone results in a total cancer unit risk estimate of 2.6 x
10"2 per ppm (4.8 x 10"6 per pg/m3) of constant lifetime TCE exposure, adjusted for potential
early-life susceptibility.
Note that the above calculation for adjusting the ADAF-adjusted lifetime unit risk
estimate for multiple sites is equivalent to adjusting each ADAF by adding a factor of 3 and
applying those factors as age-specific adjustment factors for both early-life susceptibility and
multiple sites to the unadjusted kidney cancer unit risk estimate (i.e., 13, 6, and 4 for <2 years, 2
to <16 years, and > 16 years, respectively. The total cancer risk estimate of 4.7 x 10"6 per pg/m3,
adjusted for potential increased early-life susceptibility, derived below for a constant exposure of
1 pg/m3 differs from the unit risk estimate of 4.8 x 10"6 per pg/m3 presented above only because
of round-off error.
6/22/2009
1170
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
total cancer risk from exposure to constant TCE exposure level of 1 (.ig/nr3 from ages 0-70:
combined
adjustment
unit risk
exposure
duration
partial
Aee eroup
factor
(per ng/m3)
cone (ue/m3)
adjustment
risk
0 - < 2 years
13
1.0 x 10"6
1
2 years/70 years
3.7 x 10"7
2 - < 16 years
6
1.0 x 10"6
1
14 years/70 years
1.2 x 10"6
>16 years
4
1.0 x 10"6
1
54 years/70 years
3.1 x 10"6
total risk =
4.7 x 10"6
Note that the partial risk for each age group is the product of the values in columns 2-5 [e.g., 13 x
(1.0 x 10"6) x 1 x 2/70 = 3.7 x 10"7], and the total risk is the sum of the partial risks. This 70-year
risk estimate for a constant exposure of 1 (.ig/nr3 is equivalent to a lifetime unit risk of 4.7 x 10"6
per |ig/m3. adjusted for early-life susceptibility, assuming a 70-year lifetime and constant
exposure across age groups.
This total cancer unit risk estimate of 2.6 x 10"2 per ppm (4.8 x 10"6 per (j,g/m3), adjusted
for potential increased early-life susceptibility, is only minimally (17.5%) increased over the
unadjusted total cancer unit risk estimate because the kidney cancer risk estimate that gets
adjusted for potential increased early-life susceptibility is only part of the total cancer risk
estimate. Thus, foregoing the ADAF adjustment in the case of full lifetime calculations will not
seriously impact the resulting risk estimate. For less-than-lifetime exposure calculations, the
impact of applying the ADAFs will increase as the proportion of time at older ages decreases.
The maximum impact will be when exposure is for only the first 2 years of life, in which case the
partial lifetime total cancer risk estimate for exposure to 1 (J,g/m3 adjusted for potential increased
early-life susceptibility is 13 x (1 (j,g/m3) x (1.0 x 10~6 per (J,g/m3) x (2/70), or 3.7 x 10"7, which
is over 3 times greater than the unadjusted partial lifetime total cancer risk estimate for exposure
to 1 (J,g/m3 of 4 x (1 (j,g/m3) x (1.0 x 10~6 per (J,g/m3) x (2/70), or 1.1 x 10"7.
5.2.3.3.2 Example application of ADAFs for oral exposures
For oral exposures, the calculation of risk estimates adjusted for potential increased early-
life susceptibility is complicated by the fact that for a constant exposure level, e.g., a constant
concentration of TCE in drinking water, doses will vary by age because of different age-specific
uptake rates, e.g., drinking water consumption rates. Different U.S. EPA Program or Regional
Offices may have different default age-specific uptake rates that they use for risk assessments for
specific exposure scenarios, and the calculations presented below are merely to illustrate the
general approach to applying ADAFs for oral TCE exposures, using lifetime exposure to 1 p,g/L
of TCE in drinking water as an example.
6/22/2009
1171
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Age-specific water ingestion rates in L/kg/day were taken from U.S. EPA's Child-
Specific Exposure Factors Handbook (U.S. EPA, 2008). Values for the 90th percentile were
taken from Table 3-19 (consumers only estimates of combined direct and indirect water
ingestion from community water). The 90th percentile was based on the policy in the U.S. EPA
Office of Water for determining risk through direct and indirect consumption of drinking water.
Community water was used in the illustration because U.S. EPA only regulates community water
sources, and not private wells and cisterns or bottled water. Data for "consumers only" (i.e.,
excluding individuals who did not ingest community water) were used because formula-fed
infants (as opposed to breast-fed infants, who consume very little community water), children,
and young adolescents are often the population of concern with respect to water consumption.
For the 16+ age group, the standard default rate for adults was used (i.e., 2 L/day 70 kg, or
0.029 L/kg/day) (U.S. EPA, 1997, page 3-1), which is identical to the 90th percentile for the 18 to
<21 age group. For the purposes of this illustration, the different age-specific rates were
collapsed into the same age groupings as the ADAFs using a time-weighted averaging. These
age-specific water ingestion rates are presented in Table 5.2.17.
Table 5.2.17. Estimates of age-specific water ingestion rates (90th percentile)3
age
ingestion rate (L/kg/day)
birth to <1 month
0.238
1 to <3 months
0.228
3 to <6 months
0.148
6 to <12 months
0.112
1 to <2 years
0.056
0 to <2 years
0.103
2 to <3 years
0.052
3 to <6 years
0.049
6 to <11 years
0.035
11 to <16 years
0.026
2 to <16 years
0.036
> 16 yearsb
0.029
a. values in bold are time-weighted averages corresponding to the ADAF age groupings
b. for this age grouping, the standard adult default rate is presented
For simplicity, the adjustments for potential cancer risk at multiple sites and for potential
increased early-life susceptibility are made simultaneously using age-specific combined
adjustment factors, as was done in the second (equivalent) lifetime risk calculation for inhalation
6/22/2009
1172
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
exposures in Section 5.2.3.3.1. In the case of oral cancer risk, however, the ratio for total risk
relative to kidney cancer risk was about 5 (see Section 5.2.2.3); thus, a factor of 4 is added to
each of the ADAFs to account for risk of tumor types other than kidney cancer. The
calculations for the combined adjustment are shown in Table 5.2.18.
Table 5.2.18. Sample calculation for total lifetime cancer risk based on the kidney unit risk
estimate, adjusting for potential risk at multiple sites and for potential increased early-life
susceptibility: constant lifetime exposure to 1 iig/mL of TCE in drinking water
age group
combined
unit risk8
exposure
water
duration
partial riskc
(years)
adjustment
(per
concb
ingestion
adjustment
factor
mg/kg/day)
(mg/L)
rate
(fraction of
(L/kg/day)
years)
0 to <2 years
14
9.33 x 10"3
0.001
0.103
2/70
3.8 x 10"7
2 to <16 years
7
9.33 x 10"3
0.001
0.036
14/70
4.7 x 10"7
>16 years
5
9.33 x 10"3
0.001
0.029
54/70
1.04 x 10"6
total lifetime riskd
1.9 x 10 6
a. unit risk estimate for kidney cancer based on primary dose metric, from Table 5.2.16.
b. from Table 5.2.17
c. the partial risk for each tumor type is the product of the values in columns 2-6
d. the total lifetime risk estimate is the sum of the partial risks
Because the TCE intake is not constant across age groups, we do not calculate a lifetime
unit risk estimate in terms of risk per mg/kg/day adjusted for potential increased early-life
susceptibility. One could calculate a unit risk estimate for TCE in drinking water in terms of
[j,g/L from the result in Table 5.2.18, but this is not something that is commonly reported, and it
is dependent on the water ingestion rates used.
As with the adjusted inhalation risk estimate in Section 5.2.3.3.1, the lifetime total cancer
risk estimate of 1.9 x 10"6 calculated for lifetime exposure to 1 |ig/L of TCE in drinking water
adjusted for potential increased early-life susceptibility is only minimally (25%) increased over
the unadjusted total cancer unit risk estimate. [This calculation is not shown, but if one uses just
the factor of 5 for potential cancer risk at multiple sites for each of the age groups in Table
5.2.18, the resulting total lifetime risk estimate is 1.5 x 10"6.] Unlike with inhalation exposure
under the assumption of ppm equivalence, the oral intake rates are higher in the potentially more
susceptible younger age groups. This would tend to yield a larger relative impact of adjusting
for potential increased early-life susceptibility for oral risk estimates compared to inhalation risk
estimates. In the case of TCE, however, this impact is partially offset by the lesser proportion of
the total oral cancer risk that is accounted for by the kidney cancer risk, which is the component
of total risk that is being adjusted for potential increased early-life susceptibility, based on the
6/22/2009
1173
-------
1
2
3
4
5
6
7
8
9
10
11
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
primary dose metrics (1/5 versus 1/4 for inhalation). Thus, as with lifetime inhalation risk,
foregoing the ADAF adjustment in the case of full lifetime calculations will not seriously impact
the resulting risk estimate. For less-than-lifetime exposure calculations, the impact of applying
the ADAFs will increase as the proportion of time at older ages decreases. The maximum
impact will be when exposure is for only the first 2 years of life, in which case the partial
lifetime total cancer risk estimate for exposure to 1 [j,g/L adjusted for potential increased early-
life susceptibility is 3.8 x 10"7 (from Table 5.2.18), which is almost 3 times greater than the
unadjusted partial lifetime total cancer risk estimate for exposure to 1 [j,g/L of 5 x (0.001 mg/L)
x (0.103 L/kg/day) x (9.33 x 10~3 per mg/kg/d) x (2/70), or 1.4 x 10"7.
6/22/2009
1174
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
REFERENCES
Albee, RR; Spencer, PJ; Johnson, KA; et al. (2006) Lack of trigeminal nerve toxicity in rats
exposed to trichloroethylene vapor for 13 weeks. Int J Toxicol 25(6):531-540.
Andersen ME. (1987). Tissue Dosimetxy in Risk Assessment, or What's the Problem Here
Anyway? in Drinking Water and Health, Pharmacokinetics in Risk Assessment, Vol. 8,
Washington DC: National Academy Press
Anna CH, Maronpot RR, Pereira MA, Foley JF, Malarkey DE, Anderson MW. 1994. ras proto-
oncogene activation in dichloroacetic-, trichloroethylene- and tetrachloroethylene-
induced liver tumors in B6C3F1 mice. Carcinogenesis 15:2255-2261.
Arito, H; Takahashi, M; Ishikawa, T. (1994) Effect of subchronic inhalation exposure to low-
level TCEchloroethylene on heart rate and wakefulness-sleep in freely moving rats. Jpn J
Ind Health (Sangyo Igaku) 36(1): 1-8.
Barret L, Faure J, Guilland B, Chomat D, Didier B, Debru JL. (1984) Trichloroethylene
occupational exposure: elements for better prevention. Int Arch Occup Environ Health
53(4):283-9.
Barton HA, Clewell HJ 3rd. (2000) Evaluating noncancer effects of trichloroethylene: dosimetry,
mode of action, and risk assessment. Environ Health Perspect. Suppl 10:323-334
BEIR ([Committee on the] Biological Effects of Ionizing Radiation). (1988) Health risks of
radon and other internally deposited alpha emitters. BEIR IV. Washington, DC: National
Academy Press.
Blain, L; Lachapelle, P; Molotchnikoff, S. (1994) Electroretinal responses are modified by
chronic exposure to trichloroethylene. Neurotoxicology 15(3):627—631.
Bogen, K.T. 1990. Uncertainty in Environmental Health Risk Assessment. London: Taylor &
Francis.
Buben, JA; O'Flaherty, EJ. (1985) Delineation of the role of metabolism in the hepatotoxicity of
trichloroethylene and perchloroethylene: a dose-effect study. Toxicol Appl Pharmacol
78:105-122.
6/22/2009
1175
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Bull, RJ; Orner, GA; Cheng, RS; et al. (2002) Contribution of dichloroacetate and
trichloroacetate to liver tumor induction in mice by trichloroethylene. Toxicol Appl
Pharmacol 182:55-65.
Cai H, Guengerich P. (1999). Mechanism of aqueous decomposition of trichloroethylene oxide. J
AmChem Soc. 121:11656-11663.
Cai P, Konig R, Boor PJ, Kondraganti S, Kaphalia BS, Khan MF, et al. 2008. Chronic exposure
to trichloroethene causes early onset of SLE-like disease in female MRL +/+ mice.
Toxicol Appl Pharmacol 228:68-75.
Carney, EW; Thorsrud, BA; Dugard, PH; Zablotny, CL. (2006) Developmental toxicity studies
in Crl:Cd (SD) rats following inhalation exposure to trichloroethylene and
perchloroethylene. Birth Defects Research (Part B) 77:405-412.
Charbotel B, Fevotte J, Hours M, Martin JL, Bergeret A. (2006) Case-control study on renal cell
cancer and occupational exposure to trichloroethylene. Part II: Epidemiological aspects.
Ann Occup Hyg 50(8):777-87.
Chia SE, Ong CN, Tsakok MF, Ho A. (1996) Semen parameters in workers exposed to
trichloroethylene. Reprod Toxicol 10(4):295-299.
Chiu WA, White P. (2006). Steady-state solutions to PBPK models and their applications to risk
assessment I: Route-to-route extrapolation of volatile chemicals. Risk Anal. 26:769-80.
Chiu WA. (2006). Statistical issues in physiologically-based pharmacokinetic modeling.
Toxicokinetics and Risk Assessment, ed. Lipscomb, JC, Ohanian, EV. New York:
Informa healthcare. 269-284.
Clewell HJ, Anderson ME. (2004) Applying mode-of-action and pharmacokinetic considerations
in contemporary cancer risk assessments: an example with trichloroethylene. Crit Rev
Toxicol. 34(5):385-445.
Crofton, KM; Zhao, X. (1997) The ototoxicity of trichloroethylene: extrapolation and relevance
of high-concentration, short-duration animal exposure data. Fundam Appl Toxicol
38(1): 101—106.
Dalbey W, Bingham E. 1978. Metabolism of trichloroethylene by the isolated perfused lung.
Toxicol Appl Pharmacol. 43:267-277.
6/22/2009
1176
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Dekant, W., K. Berthold, S. Vamvakas, D. Hencschler and M.W. Anders. 1988. Thioacylating
intermediates as metabolites of S-(l,2-dichlorovinyl)-L-cysteine and S-( 1,2,2-
trichlorovinyl)-L-cysteine formed by cysteine conjugate P-lyase. Chem. Res. Toxicol.
1:175-178.
DuTeaux, SB; Berger, T; Hess, RA; Sartin, BL ; Miller, MG. (2004b) Male reproductive toxicity
of trichloroethylene: sperm protein oxidation and decreased fertilizing ability. Biol
Reprod 70:1518-1526.
DuTeaux, SB; Hengel, MJ; DeGroot, DE; Jelks, KA; Miller, MG. (2003) Evidence for
trichloroethylene bioactivation and adduct formation in the rat epididymis and efferent
ducts. Biol Reprod 69:771-779.
Falk Filipsson A, Victorin K. (2003). Comparison of available benchmark dose softwares and
models using trichloroethylene as a model substance. Regul Toxicol Pharmacol.
37(3):343-55.
Fevotte J, Charbotel B, Muller-Beaute P, Martin JL, Hours M, Bergeret A. (2006). Case-control
study on renal cell cancer and occupational exposure to trichloroethylene. Part I:
Exposure assessment. Ann Occup Hyg. 50(8):765-75.
Fisher JW, Whittaker TA, Taylor DH, Clewell HJ 3rd, Andersen ME. (1989) Physiologically
based pharmacokinetic modeling of the pregnant rat: a multiroute exposure model for
trichloroethylene and its metabolite, trichloroacetic acid. Toxicol Appl Pharmacol
99(3):395-414.
Forkert, P-G; Lash, LH; Nadeau V; Tardif, R; Simmonds, A. (2002) Metabolism and toxicity of
trichloroethylene in epididymis and testis. Toxicol Appl Pharmacol 182:244-254.
Forkert, PG; Millen, B; Lash, LH; Putt, DA, Ghanayem, BI. (2006) Pulmonary bronchiolar
cytotoxicity and formation of dichloroacetyl lysine protein adducts in mice treated with
trichloroethylene. J Pharmacol Exp Ther 316(2):520-529.
Fredriksson, A; Danielsson, BRG; Eriksson, P. (1993) Altered behavior in adult mice orally
exposed to tri- and tetrachloroethylene as neonates. Toxicology Letters 66:13-19.
Fukuda, K; Takemoto, K; Tsuruta, H. (1983) Inhalation carcinogenicity of trichloroethylene in
mice and rats. Ind Health 21:243-254.
6/22/2009
1177
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Gash, DM; Rutland, K; Hudson, NL; et al. (2007) Trichloroethylene: Parkinsonism and complex
1 mitochondrial neurotoxicity. Ann Neurol .
George, JD; Reel, JR; Myers, CB; Lawton, AD; Lamb, JC. (1985) Trichloroethylene:
reproduction and fertility assessment in CD-I mice when administered in the feed. RTI
Project No. 310-2344, NTP-86-068. National Institute of Environmental Health Sciences,
National Toxicology Program, RTP, NC.
George, JD; Reel, JR; Myers, CB; Lawton, AD; Lamb, JC. (1986) Trichloroethylene:
reproduction and fertility assessment in F344 rats when administered in the feed. RTI
Project No. 310-2344, NTP-86-085. National Institute of Environmental Health Sciences,
National Toxicology Program, RTP, NC.
Ghantous, H; Danielsson, BRG; Dencker, L; Gorczak J; Vesterberg, O. (1985) Trichloroacetic
acid accumulates in murine amniotic fluid after tri- and tetrachloroethylene inhalation.
Acta Pharmacol Toxicol (Copenh) 58:105-114.
Griffin JM, Gilbert KM, Lamps LW, Pumford NR. 2000b. CD4(+) T-cell activation and
induction of autoimmune hepatitis following trichloroethylene treatment in MRL+/+
mice. Toxicol Sci 57:345-352.
Healy, TEJ; Poole, TR; Hopper, A. (1982) Rat fetal development and maternal exposure to
trichloroethylene 100 p.p.m. British Journal of Anaesthesiology 54:337-340.
Henschler D, Elsasser H, Romen W, Eder E. 1984. Carcinogenicity study of trichloroethylene,
with and without epoxide stabilizers, in mice. J Cancer Res Clin Oncol 107:149-156.
Henschler D, Romen W, Elsasser HM, Reichaert D, Eder E, Radwan Z. 1980. Carcinogenicity
study of trichloroethyleneby longterm inhalation in three animal species. Arch Toxicol
43: 237-248 (1980).
IPCS (International Programme on Chemical Safety). (2005). Chemical-specific adjustment
factors for interspecies differences and human variability : guidance document for use of
data in dose/concentration-response assessment. Geneva: World Health Organization.
Isaacson, LG; Spohler, SA; Taylor, DH. (1990) Trichloroethylene affects learning and decreases
myelin in the rat hippocampus. Neurotoxicol Teratol 12(4):375—381.
6/22/2009
1178
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Isaacson, LG; Taylor, DH. (1989) Maternal exposure to 1,1,2-trichloroethylene affects myelin in
the hippocampal formation of the developing rat. Brain Res 488(l-2):403-407.
Johnson, PD; Goldberg, SJ; Mays, MZ; Dawson, BV. (2003) Threshold of trichloroethylene
contamination in maternal drinking waters affecting fetal heart development in the rat.
Environ Health Perspect 111(3):289-292.
Kan, FW; Forkert, PG; Wade, MG. (2007) Trichloroethylene exposure elicits damage in
epididymal epithelium and spermatozoa in mice. Histol Histopathol 22(9):977-988.
Kaneko T, Wang P-Y, Sato A. 1994. Enzymes induced by ethanol differently affect the
pharmacokinetics of trichloroethylene and 1,1,1-trichloroethane. Occupational and
Environmental Medicine 51:113-119.
Keil, DE; Peden-Adams, MM; Wallace, S; Ruiz, P; Gilkeson, GS. (2009) Assessment of
trichloroethylene (TCE) exposure in murine strains genetically-prone and non-prone to
develop autoimmune disease. Journal of Environmental Science and Health, Part A 44:
443-453.
Kjellstrand P, Holmquist B, Aim P, Kanje M, Romare S, Jonsson I, Mansson L, Bjerkemo M.
(1983a) Trichloroethylene: further studies of the effects on body and organ weights and
plasma butyrylcholinesterase activity in mice. Acta Pharmacol Toxicol (Copenh)
53(5):375-84.
Kjellstrand P, Holmquist B, Mandahl N, Bjerkemo M. (1983b) Effects of continuous
trichloroethylene inhalation on different strains of mice. Acta Pharmacol Toxicol
(Copenh) 53(5):369-74.
Kjellstrand, P; Kanje, M; Bjerkemo, M. (1987) Regeneration of the sciatic nerve in mice and rats
exposed to trichloroethylene. Toxicol Lett 38(1-2): 187—191.
Kopylev, L; Chen, C; White, P. (2007) Towards quantitative uncertainty assessment for cancer
risks: central estimates and probability distributions of risk in dose-response modeling.
Regul Toxicol Pharmacol 49(3):203-207.
Kulig, BM. (1987) The effects of chronic trichloroethylene exposure on neurobehavioral
functioning in the rat. Neurotoxicol Teratol 9(2): 171-178.
6/22/2009
1179
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Kumar, P; Prasad, AK; Dutta, KK. (2000b) Steroidogenic alterations in testes and sera of rats
exposed to trichloroethylene (TCE) by inhalation. Hum Exp Toxicol 19:117-121.
Kumar, P; Prasad, AK; Mani, U; Maji, BK; Dutta, KK. (2001) Trichloroethylene induced
testicular toxicity in rats exposed by inhalation. Hum Exp Toxicol 20:585-589.
Kumar, P; Prasad, AK; Saxena, DK; Mani, U; Maji, BK; Dutta, KK. (2000a) Fertility and
general reproduction studies in trichloroethylene exposed rats. Indian J Occup Health
43(3): 117-125.
Land, PC; Owen, EL; Linde, HW. (1981) Morphologic changes in mouse spermatozoa after
exposure to inhalational anesthetics during early spermatogenesis. Anesthesiology 54:53-
56.
Lutz WK, Gaylor DW, Conolly RB, Lutz RW. (2005). Nonlinearity and thresholds in dose-
response relationships for carcinogenicity due to sampling variation, logarithmic dose
scaling, or small differences in individual susceptibility. Toxicol Appl Pharmacol. 207(2
Suppl):565-9.
Maltoni, C; Lefemine, G; Cotti, G. (1986) Experimental research on trichloroethylene
carcinogenesis. In: Maltoni, C; Mehlman MA., eds. Vol. 5. Archives of research on
industrial carcinogenesis. Princeton, NJ: Princeton Scientific Publishing;
Manson, JM; Murphy, M; Richdale, N; Smith, MK. (1984) Effects of oral exposure to
trichloroethylene on female reproductive function. Toxicology 32:229-242.
Mhiri, C; Choyakh, F; Ben, HM; et al. (2004) Trigeminal somatosensory evoked potentials in
trichloroethylene-exposed workers. Neurosciences 9(2): 102-107.
Moser, Cheek & MacPhail. A multidisciplinary approach to toxicological screening III.
Neurotbehavioral toxicity. J Toxicol. Environ. Hlth., 1995, 45, 173-210.
Narotsky, MG; Weller, EA; Chinchilli, VM; Kavlock, RJ. (1995) Nonadditive developmental
toxicity in mixtures of trichloroethylene, di(2-ethylhexyl) phthalate, and heptachlor in a 5
x 5 x 5 design. Fundam Appl Toxicol 27:203-216.
NCHS (National Center for Health Statistics). (2007) National Vital Statistics Reports, vol. 55,
no. 19; August 21, 2007, Table 3. National Center for Health Statistics, Hyattsville, MD.
6/22/2009
1180
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
NCI (National Cancer Institute). (1976) Carcinogenesis bioassay of trichloroethylene. Division
of Cancer Cause and Prevention, National Cancer Institute, U.S. Department of Health,
Education, and Welfare, DHEW Publication No. (NIH) 76-802, Technical Report Series
No. 2, 218 pages; NCI-CG-TR-2; NTIS PB254122.
http://ntp.niehs.nih.gov/ntp/htdocs/LT_rpts/tr002.pdf.
NRC (National Research Council) (1986). Drinking Water and Health, Vol. 6. Washington DC:
National Academy Press.
NRC (National Research Council) (1987). Drinking Water and Health, Pharmacokinerics in
Risk Assessment, Vol. 8. Washington DC: National Academy Press.
NRC (National Research Council). 1994. Science and Judgment in Risk Assessment.
Washington, DC: National Academy Press
NTP (National Toxicology Program). (1988) Toxicology and carcinogenesis studies of
trichloroethylene (CAS no. 79-01-6) in four strains of rats (ACI, August, Marshall,
Osborne-Mendel) (gavage studies). Public Health Service, U.S. Department of Health
and Human Services; NTP TR-273; NIH Publication No. 88-2529. Available from the
National Institute of Environmental Health Sciences, Research Triangle Park, NC, and
the National Technical Information Service, Springfield, VA; PB88-218896.
http://ntp.niehs.nih.gov/ntp/htdocs/LT_rpts/tr273.pdf.
NTP (National Toxicology Program). (1990) Carcinogenesis Studies of Trichloroethylene
(Without Epichlorhydrin) (CAS No. 79-01-6) in F344/N Rats and B6C3F1 Mice (Gavage
Study). NTP TR 243. Research Triangle Park, NC: U.S Department of Health and
Human Services.
Nunes J, Ehrich M, Robertson J. 2001. Toxicosis associated with dual oral exposure of rats to
lead and trichloroethylene. Toxicologic Pathology. 29(4): 451-457.
Peden-Adams MM, Eudaly JG, Heesemann LM, Smythe J, Miller J, Gilkeson GS, et al. (2006).
Developmental immunotoxicity of trichloroethylene (TCE): studies in B6C3F1 mice. J
Environ Sci Health A Tox Hazard Subst Environ Eng 41:249-271.
Peden-Adams MM, EuDaly JG, Lee AM, Miller DE, Keil DE, Gilkeson GS. (2008)
Developmental effects of trichloroethylene (TCE) on immunity: studies in MRL (+/+)
mice. J Environ Sci Health A Tox Hazard Subst Environ Eng A43(12): 1402-9.
6/22/2009
1181
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Portier CJ, Bailer AJ (1989) Testing for increased carcinogenicity using a survival-adjusted
quantal response test. Fund Appl Toxicol 12:731-737.
Raaschou-Nielsen O, Hansen J, McLaughlin JK, Kolstad H, Christensen JM, Tarone RE, Olsen
JH. (2003) Cancer risk among workers at Danish companies using trichloroethylene: a
cohort study. Am J Epidemiol 158:1182-1192.
Rasmussen, K; Sabroe, S. (1986) Neuropsychological symptoms among metal workers exposed
to halogenated hydrocarbons. Scand J Soc Med 14(3): 161-168.
Rebert CS, Day VL, Matteucci MJ, Pryor GT (1991). Sensory-evoked potentials in rats
chronically exposed to trichloroethylene: predominant auditory dysfunction.
Neurotoxicol Teratol. 13(l):83-90
Rhomberg LR. (2000). Dose-response analyses of the carcinogenic effects of trichloroethylene in
experimental animals. Environ Health Perspect 108(suppl 2):343-358.
Rice D, Barone S Jr. (2000). Critical periods of vulnerability for the developing nervous system:
evidence from humans and animal models. Environ Health Perspect. Suppl 108, 3:511-
33.
Rothman KJ (1986). Modern Epidemiology. Little, Brown and Company, Boston.
Rothman KJ, Greenland S (1998). Modern Epidemilogy, Second Edition. Lippincott Williams
& Wilkins, Philadelphia.
Ruijten, M; Verberk, MM; Salle, HJ. (1991) Nerve function in workers with long term exposure
to trichloroethene. Br J Ind Med 48(2):87-92.
Sanders VM, Tucker AN, White KL, Jr., Kauffmann BM, Hallett P, Carchman RA, et al. 1982.
Humoral and cell-mediated immune status in mice exposed to trichloroethylene in the
drinking water. Toxicol Appl Pharmacol 62:358-368.
Sausen, P.J. and A. A. Elfarra. 1991. Reactivity of cysteine s-conjugate sulfoxides: formation of
S-[l-chloro-2-(S-glutathionyl)vinyl]-L-cysteine sulfoxide by the reaction of S (1,2
dichlorovinyl)-L-cycsteine sulfoxide with glutathione. Chem. Res. Toxicol. 4:655-660.
6/22/2009
1182
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
Schwetz, BA; Leong, BKJ; Gehring, PJ. (1975) The effect of maternally inhaled
trichloroethylene, perchloroethylene, methyl chloroform, and methylene chloride on
embryonal and fetal development in mice and rats. Toxicol Appl Pharmacol 32:84-96.
Sheffels, P; Schroeder, JL; Altuntas, TG; Liggitt, HD; Kharasch, ED. (2004) Role of cytochrome
P4503 A in cysteine S-conjugates sulfoxidation and the nephrotoxicity of the sevoflurane
degradation product fluoromethyl-2,2-difluoro-l-(trifluoromethyl)vinyl ether (compound
A) in rats. Chem Res Toxicol 17(9): 1177-1189.
Spiegelhalter, D; Thomas, A; Best, N; et al. (2003) WinBUGS user manual. Version 1.4.
Available online atwww.mrc-bsu.cam.ac.uk/bugs/winbugs/manuall4.pdf.
Taylor, DH; Lagory, KE; Zaccaro, DJ; Pfohl, RJ; Laurie, RD. (1985) Effect of trichloroethylene
on the exploratory and locomotor activity of rats exposed during development. Sci Total
Environ 47:415-420.
Travis CC, White RK. (1988) Interspecific scaling of toxicity data. Risk Anal. 8:119-25.
Travis CC. (1990). Tissue dosimetry for reactive metabolites. Risk Anal. 10:317-21.
U.S. EPA (Environmental Protection Agency) (1992). Draft Report: A cross-species scaling
factor for carcinogen risk assessment based on equivalence of mg/kg3/4/Day; Notice.
Federal Register 57(109): 24152-2424173. Friday, June 5, 1992.
U.S. EPA (Environmental Protection Agency) (2008b). Benchmark Dose Technical Guidance
(Inter-Agency Review Draft).
U.S. EPA (Environmental Protection Agency). (1994) Methods for derivation of inhalation
reference concentrations and application of inhalation dosimetry. Environmental Criteria
and Assessment Office, Office of Health and Environmental Assessment, Washington,
Washington, DC; EPA/600/8-90/066F. Available from: National Technical Information
Service, Springfield, VA; PB2000-500023.
U.S. EPA (Environmental Protection Agency). (1997) Exposure Factors Handbook, Volume I -
General Factors. National Center for Environmental Assessment, Office of Research and
Development, Washington, DC, EPA/600/P-95/002Fa
U.S. EPA (Environmental Protection Agency). (2002a) Health assessment of 1,3-butadiene.
Office of Research and Development, Washington, DC. EPA/600/P-98/001F.
6/22/2009
1183
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
INTER-AGENCY REVIEW DRAFT- DO NOT QUOTE OR CITE
U.S. EPA. (2002b) 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. Risk Assessment Forum,
Washington, DC; EPA/630/P-03/001B. Available from:
.
U.S. EPA. (2005b) Supplemental guidance for assessing susceptibility from early-life exposure
to carcinogens. Risk Assessment Forum, Washington, DC; EPA/630/R-03/003F.
Available from: .
U.S. EPA. Child-Specific Exposure Factors Handbook (2008). U.S. Environmental Protection
Agency, Washington, D.C., EPA/600/R-06/096F, 2008.
Van Duuren BL, Goldschmidt BM, Loewengart G, Smith AC, Meichionne S, Seidman I, Roth
D. 1979. Carcinogenicity of halogenated olefinic and aliphatic hydrocarbons in mice.
JNCI 63(6): 1433-1439.
Waseem, M; Ali, M; Dogra, S; et al. (2001) Toxicity of trichloroethylene following inhalation
and drinking contaminated water. J Appl Toxicol 21(6):441-444.
Woolhiser, MR; Krieger, SM; Thomas, J; Hotchkiss, JA. (2006) Trichloroethylene (TCE):
Immunotoxicity potential in CD rats following a 4-week vapor inhalation exposure. Dow
Chemical Company, Toxicology & Environmental Research and Consulting, Midland,
MI, Study ID 031020, July 5, 2006, unpublished.
Xu, H; Tanphaichitr, N; Forkert, P-G; Anupriwan, A; Weerachatyanukul, W; Vincent, R;
Leader, A; Wade, MG. (2004) Exposure to trichloroethylene and its metabolites causes
impairment of sperm fertilizing ability in mice. Toxicol Sci 82:590-597.
Zeise L, Wilson R, Crouch EA. (1987). Dose-response relationships for carcinogens: a review.
Environ Health Perspect. 73:259-306.
Zenick H, Blackburn K, Hope E, Richdale N, Smith MK. (1984) Effects of trichloroethylene
exposure on male reproductive function in rats. Toxicology 31(3-4):237-50.
6/22/2009
1184
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
6 MAJOR CONCLUSIONS IN Till CHARACTERIZATION OF
HAZARD AND DOSE RESPONSE
6.1 Human Hazard Potential
This section summarizes the human hazard potential for TCE. For extensive discussions
and references, see Chapter 2 for Exposure, Chapter 3 for toxicokinetics and PBPK modeling,
and Sections 4.0-4.8 for the epidemiologic and experimental studies of TCE non-cancer and
cancer toxicity. Section 4.9 summarizes information on susceptibility, and Section 4.10 provides
a more detailed summary and references for non-cancer toxicity and carcinogenicity.
6.1.1 Exposure (Chapter 2)
TCE is a volatile compound with moderate water solubility. Most TCE produced today
is used for metal degreasing. The highest environmental releases are to the air. Ambient air
monitoring data suggests that mean levels have remained fairly constant since 1999 at about 0.3
[j,g/m3 (0.06 ppb). As discussed in Chapter 2, in 2006, ambient air monitors (n=258) had annual
means ranging from 0.03 to 7.73 (J,g/m3 with a median of 0.13 and an overall average of 0.23
[j,g/m3. Indoor levels are commonly 3 or more times higher than outdoor levels due to releases
from building materials and consumer products. TCE is among the most common groundwater
contaminants and the median level based on a large survey by USGS for 1985-2001 is 0.15 (J,g/L.
It has also been detected in a wide variety of foods in the 1-100 (J,g/kg range. None of the
environmental sampling has been done using statistically based national surveys. However, a
substantial amount of air and groundwater data have been collected allowing reasonably well
supported estimates of typical daily intakes by the general population: inhalation - 13 (j,g/day
and water ingestion - 0.2 (j,g/day. The limited food data suggests an intake of about 5 jig/day, but
this must be considered preliminary.
Much higher exposures have occurred to various occupational groups. For example, past
studies of aircraft workers have shown short term peak exposures in the hundreds of ppm
(>500,000 (J,g/m3) and long term exposures in the low tens of ppm (>50,000 (J,g/m3).
Occupational exposures have likely decreased in recent years due to better release controls and
improvements in worker protection.
Exposure to a variety of TCE related compounds, which include metabolites of TCE and
other parent compounds that produce similar metabolites, can alter or enhance TCE metabolism
and toxicity by generating higher internal metabolite concentrations than would result from TCE
6/22/2009
1185
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
exposure by itself. Available estimates suggest that exposures to most of these TCE related
compounds are comparable to or greater than TCE itself.
6.1.2 Toxicokinetics and PBPK modeling (Chapter 3 and Appendix A)
TCE is a lipophilic compound that readily crosses biological membranes. Exposures may
occur via the oral, dermal, and inhalation route, with evidence for systemic availability from
each route. TCE can also be transferred transplacental^ and through breast milk ingestion. TCE
is rapidly and nearly completely absorbed from the gut following oral administration, and animal
studies indicate that exposure vehicle may impact the time course of administration: oily
vehicles may delay absorption whereas aqueous vehicles result in a more rapid increase in blood
concentrations. See Section 3.1 for additional discussion of TCE absorption.
Following absorption to the systemic circulation, TCE distributes from blood to solid
tissues by each organ's solubility. This process is mainly determined by the blood:tissue
partition coefficients, which are largely determined by tissue lipid content. Adipose partitioning
is high, adipose tissue may serve as a reservoir for TCE, and accumulation into adipose tissue
may prolong internal exposures. TCE attains high concentrations relative to blood in the brain,
kidney, and liver—all of which are important target organs of toxicity. TCE is cleared via
metabolism mainly in three organs: the kidney, liver, and lungs. See Section 3.2 for additional
discussion of TCE distribution.
The metabolism of TCE is an important determinant of its toxicity. Metabolites are
thought to be responsible for toxicity at multiple sites, particularly in the liver and kidney.
Initially, TCE may be oxidized via cytochrome P450 (CYP) isoforms or conjugated with
glutathione by glutathione-S-transferase enzymes. While CYP2E1 is generally accepted to be
the CYP isoform most responsible for TCE oxidation, others forms may also contribute. There
are conflicting data as to which GST isoforms are responsible for TCE conjugation, with one rat
study indicating alpha-class GSTs and another rat study indicating mu and pi-class GST. The
balance between oxidative and conjugative metabolites generally favors the oxidative pathway,
especially at lower concentrations, and inhibition of CYP-dependent oxidation in vitro increases
GSH conjugation in renal preparations. However, in humans, direct comparison of in vitro rates
of oxidation and conjugation, as well as in vivo data on the amount of the TCE GSH conjugation
to dichlorovinyl glutathione in blood, support a flux through the GSH pathway that may be one
or more orders of magnitude greater than the <0.1% inferred from excretion of GSH conjugation
derived urinary mercapturates. See Section 3.3 for additional discussion of TCE metabolism.
Once absorbed, TCE is excreted primarily either in breath as unchanged TCE or CO2, or
in urine as metabolites. Minor pathways of elimination include excretion of metabolites in
6/22/2009
1186
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
saliva, sweat, and feces. Following oral administration or upon cessation of inhalation exposure,
exhalation of unmetabolized TCE is a major elimination pathway. Initially, elimination of TCE
upon cessation of inhalation exposure demonstrates a steep concentration-time profile: TCE is
rapidly eliminated in the minutes and hours postexposure, and then the rate of elimination via
exhalation decreases. Following oral or inhalation exposure, urinary elimination of parent TCE
is minimal, with urinary elimination of the metabolites trichloroacetic acid and trichloroethanol
accounting for the bulk of the absorbed dose of TCE. See Section 3.4 for additional discussion
of TCE excretion.
As part of this assessment, a comprehensive Bayesian PBPK model-based analysis of the
population toxicokinetics of TCE and its metabolites was developed in mice, rats, and humans.
This analysis considered a wider range of physiological, chemical, in vitro, and in vivo data than
any previously published analysis of TCE. The toxicokinetics of the "population average," its
population variability, and their uncertainties are characterized and estimates of experimental
variability and uncertainty were included in this analysis. The experimental database included
separate sets for model calibration and evaluation for rats and humans; fewer data were available
in mice, and were all used for model calibration. The total combination of these approaches and
PBPK analysis substantially supports the model predictions. In addition, the approach employed
yields an accurate characterization of the uncertainty in metabolic pathways for which available
data were sparse or relatively indirect, such as GSH conjugation and respiratory tract
metabolism. Key conclusions from the model predictions include: (1) as expected, TCE is
substantially metabolized, primarily by oxidation at doses below saturation; (2) GSH conjugation
and subsequent bioactivation in humans appears to be 10- to 100-fold greater than previously
estimated; and (3) mice had the greatest rate of respiratory tract oxidative metabolism compared
to rats and humans. The predictions of the PBPK model are subsequently used in non-cancer
and cancer dose-response analyses for inter- and intra-species extrapolation of toxicokinetics
(see below). See Section 3.5 and Appendix A for additional discussion of and details about
PBPK modeling of TCE and metabolites.
6.1.3 Non-cancer toxicity
This section summarizes the weight of evidence for TCE non-cancer toxicity. Based on
the available human epidemiologic data and experimental and mechanistic studies, it is
concluded that TCE poses a potential human health hazard for non-cancer toxicity to the central
nervous system, the kidney, the liver, the immune system, the male reproductive system, and the
developing fetus. The evidence is more limited for TCE toxicity to the respiratory tract and
6/22/2009
1187
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
female reproductive system. The conclusions pertaining to specific endpoints within these
tissues and systems are summarized below.
6.1.3.1 Neurological effects (Sections 4.2 and 4.10.1.1 and Appendix D)
Both human and animal studies have associated TCE exposure with effects on several
neurological domains. Multiple epidemiologic studies in different populations have reported
abnormalities in trigeminal nerve function in association with TCE exposure. Two small studies
did not report an association between TCE exposure and trigeminal nerve function. However,
statistical power was limited, exposure misclassification was possible, and in one case methods
for assessing trigeminal nerve function was not available. As a result, these studies do not
provide substantial evidence against a causal relationship between TCE exposure and trigeminal
nerve impairment. Laboratory animal studies have also demonstrated TCE-induced changes in
the morphology of the trigeminal nerve following short-term exposures in rats. However, one
study reported no significant changes in trigeminal somatosensory evoked potential in rats
exposed to TCE for 13 weeks. See Section 4.2.1 for additional discussion of studies of
alterations in nerve conduction and trigeminal nerve effects. Human chamber, occupational, and
geographic based/drinking water studies have consistently reported subjective symptoms such as
headaches, dizziness, and nausea which are suggestive of vestibular system impairments. One
study reported changes in nystagmus threshold (a measure of vestibular system function)
following an acute TCE exposure. There are only a few laboratory animal studies relevant to
this neurological domain, with reports of changes in nystagmus, balance, and handling reactivity.
See Section 4.2.3 for additional discussion of TCE effects on vestibular function. Fewer and
more limited epidemiologic studies are suggestive of TCE exposure being associated with
delayed motor function, and changes in auditory, visual, and cognitive function or performance
(see Sections 4.2.2, 4.2.4, 4.2.5, and 4.2.6). Acute and subchronic animal studies show
disruption of the auditory system, changes in visual evoked responses to patterns or flash
stimulus, and neurochemical and molecular changes. Animal studies suggest that while the
effects on the auditory system lead to permanent function impairments and histopathology,
effects on the visual system may be reversible with termination of exposure. Additional acute
studies reported structural or functional changes in hippocampus, such as decreased myelination
or decreased excitability of hippocampal CA1 neurons, although the relationship of these effects
to overall cognitive function is not established (see Section 4.2.9). An association between TCE
exposure and sleep changes has also been demonstrated in rats (see Section 4.2.7). Some
evidence exists for motor-related changes in rats/mice exposed acutely/subchronically to TCE,
but these effects have not been reported consistently across all studies (see Section 4.2.6).
Gestational exposure to TCE in humans has been reported to be associated with
6/22/2009
1188
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
neurodevelopmental abnormalities including neural tube defects, encephalopathy, impaired
cognition, aggressive behavior, and speech and hearing impairment. Developmental
neurotoxicological changes have also been observed in animals including aggressive behaviors
following an in utero exposure to TCE and suggestion of impaired cognition as noted by
decreased myelination in the CA1 hippocampal region of the brain. See Section 4.2.8 for
additional discussion of developmental neurological effects of TCE. Therefore, overall, the
strongest neurological evidence of human toxicological hazard is for changes in trigeminal nerve
function or morphology and impairment of vestibular function, based on both human and
experimental studies, while fewer and more limited evidence exists for delayed motor function,
changes in auditory, visual, and cognitive function or performance, and neurodevelopmental
outcomes.
6.1.3.2 Kidney effects (Sections 4.3.1, 4.3.4, 4.3.6, and 4.10.1.2)
Kidney toxicity has also been associated with TCE exposure in both human and animal
studies. There are few human data pertaining to TCE-related non-cancer kidney toxicity;
however, several available studies reported elevated excretion of urinary proteins, considered
nonspecific markers of nephrotoxicity, among TCE-exposed subjects compared to unexposed
controls. While some of these studies include subjects previously diagnosed with kidney cancer,
other studies report similar results in subjects that are disease free. Some additional support for
TCE nephrotoxicity in humans is provided by a study reporting a greater incidence of end-stage
renal disease in TCE-exposed workers as compared to unexposed controls, although some
subjects in this study were also exposed to hydrocarbons, JP-4 gasoline, and multiple solvents,
including TCE and 1,1,1-trichloroethane. See Section 4.3.1 for additional discussion of human
data on the non-cancer kidney effects of TCE. Laboratory animal and in vitro data provide
additional support for TCE nephrotoxicity. TCE causes renal toxicity in the form of cytomegaly
and karyomegaly of the renal tubules in male and female rats and mice following either oral or
inhalation exposure. In rats, the pathology of TCE-induced nephrotoxicity appears distinct from
age-related nephropathy. Increased kidney weights have also been reported in some rodent
studies. See Section 4.3.4 for additional discussion of laboratory animal data on the non-cancer
kidney effects of TCE. Further studies with TCE metabolites have demonstrated a potential role
for DCVC, TCOH, and TCA in TCE-induced nephrotoxicity. Of these, available data suggest
that DCVC induced renal effects are most similar to those of TCE and that DCVC is formed in
sufficient amounts following TCE exposure to account for these effects. TCE or DCVC have
also been shown to be cytotoxic to primary cultures of rat and human renal tubular cells. See
Section 4.3.6 for additional discussion on the role of metabolism in the non-cancer kidney effects
of TCE. Overall, multiple lines of evidence support the conclusion that TCE causes
6/22/2009
1189
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
nephrotoxicity in the form of tubular toxicity, mediated predominantly through the TCE GSH
conjugation product DCVC.
6.1.3.3 Liver effects (Sections 4.4.1, 4.4.3, 4.4.4, 4.4.6,and 4.10.1.3, and Appendix E)
Liver toxicity has also been associated with TCE exposure in both human and animal
studies. Although there are few human studies on liver toxicity and TCE exposure, several
available studies have reported TCE exposure to be associated with significant changes in serum
liver function tests, widely used in clinical settings in part to identify patients with liver disease,
or changes in plasma or serum bile acids. Additional, more limited human evidence for TCE
induced liver toxicity includes reports suggesting an association between TCE exposure and liver
disorders, and case reports of liver toxicity including hepatitis accompanying immune-related
generalized skin diseases, jaundice, hepatomegaly, hepatosplenomegaly, and liver failure in
TCE-exposed workers. Cohort studies examining cirrhosis mortality and either TCE exposure or
solvent exposure are generally null, but cannot rule out an association with TCE because of their
use of death certificates where there is a high degree (up to 50%) of underreporting. Overall,
while some evidence exists of liver toxicity as assessed from liver function tests, the data are
inadequate for making conclusions regarding causality. See Section 4.4.1 for additional
discussion of human data on the non-cancer liver effects of TCE. In rats and mice, TCE
exposure causes hepatomegaly without concurrent cytotoxicity. Like humans, laboratory
animals exposed to TCE have been observed to have increased serum bile acids, although the
toxicologic importance of this effect is unclear. Other effects in the rodent liver include small
transient increases in DNA synthesis, cytomegaly in the form of "swollen" or enlarged
hepatocytes, increased nuclear size probably reflecting polyploidization, and proliferation of
peroxisomes. Available data also suggest that TCE does not induce substantial cytotoxicity,
necrosis, or regenerative hyperplasia, as only isolated, focal necroses and mild to moderate
changes in serum and liver enzyme toxicity markers having been reported. These effects are
consistently observed across rodent species and strains, although the degree of response at a
given mg/kg/d dose appears to be highly variable across strains, with mice on average appearing
to be more sensitive. See Sections 4.4.3 and 4.4.4 for additional discussion of laboratory animal
data on the non-cancer liver effects of TCE. While it is likely that oxidative metabolism is
necessary for TCE-induced effects in the liver, the specific metabolite or metabolites responsible
is less clear. However, the available data are strongly inconsistent with TCA being the sole or
predominant active moiety for TCE-induced liver effects, particularly with respect to
hepatomegaly. See Section 4.4.6 for additional discussion on the role of metabolism in the non-
cancer liver effects of TCE. Overall, TCE, likely through its oxidative metabolites, clearly leads
to liver toxicity in laboratory animals, with mice appearing to be more sensitive than other
6/22/2009
1190
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
laboratory animal species, but there is only limited epidemiologic evidence of hepatotoxicity
being associated with TCE exposure.
6.1.3.4 Immunological effects (Sections 4.5.1.1, 4.5.2, and 4.10.1.4)
Effects related the immune system have also been associated with TCE exposure in both
human and animal studies. A relationship between systemic autoimmune diseases, such as
scleroderma, and occupational exposure to TCE has been reported in several recent studies, and a
meta-analysis of scleroderma studies resulted in a statistically significant combined odds ratio for
any exposure in men (OR = 2.5, 95% CI 1.1, 5.4), with a lower relative risk seen in women in
women (OR = 1.2, 95% CI 0.58, 2.6). The human data at this time do not allow a determination
of whether the difference in effect estimates between men and women reflects the relatively low
background risk of scleroderma in men, gender-related differences in exposure prevalence or in
the reliability of exposure assessment, a gender-related difference in susceptibility to the effects
of TCE, or chance. Additional human evidence for the immunological effects of TCE includes
studies reporting TCE-associated changes in levels of inflammatory cytokines in occupationally-
exposed workers and infants exposed via indoor air at air concentrations typical of such exposure
scenarios (see 6.1.1, above); a large number of case reports (mentioned above) of a severe
hypersensitivity skin disorder, distinct from contact dermatitis and often accompanied by
heptatitis; and a reported association between increased history of infections and exposure to
TCE contaminated drinking water. See Section 4.5.1.1 for additional discussion of human data
on the immunological effects of TCE. Immunotoxicity has also been reported in experimental
rodent studies of TCE. Numerous studies have demonstrated accelerated autoimmune responses
in autoimmune-prone mice, including changes in cytokine levels similar to those reported in
human studies, with more severe effects, including autoimmune hepatitis, inflammatory skin
lesions, and alopecia, manifesting at longer exposure periods. Immunotoxic effects have been
also reported in B6C3F1 mice, which do not have a known particular susceptibility to
autoimmune disease. Developmental immunotoxicity in the form of hypersensitivity responses
have been reported in TCE-treated guinea pigs and mice via drinking water pre- and post-natally.
Evidence of localized immunosuppression has also been reported in mice and rats. See Section
4.5.2 for additional discussion of laboratory animal data on the immunological effects of TCE.
Overall, the human and animal studies of TCE and immune-related effects provide strong
evidence for a role of TCE in autoimmune disease and in a specific type of generalized
hypersensitivity syndrome, while there are less data pertaining to immunosuppressive effects.
6.1.3.5 Respiratory tract effects (Sections 4.6.1.1, 4.6.2.1, 4.6.3, and4.10.1.5)
The very few human data on TCE and pulmonary toxicity are too limited for drawing
conclusions (see Section 4.6.1.1), but laboratory studies in mice and rats have shown toxicity in
6/22/2009
1191
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
the bronchial epithelium, primarily in Clara cells, following acute exposures to TCE (see Section
4.6.2.1). A few studies of longer duration have reported more generalized toxicity, such as
pulmonary fibrosis in mice and pulmonary vasculitis in rats. However, respiratory tract effects
were not reported in other longer-term studies. Acute pulmonary toxicity appears to be
dependent on oxidative metabolism, although the particular active moiety is not known. While
earlier studies implicated chloral produced in situ by P450 enzymes in respiratory tract tissue in
toxicity, the evidence is inconsistent and several other possibilities are viable. Although humans
appear to have lower overall capacity for enzymatic oxidation in the lung relative to mice, P450
enzymes do reside in human respiratory tract tissue, suggesting that, qualitatively, the respiratory
tract toxicity observed in rodents is biologically plausible in humans. See Section 4.6.3 for
additional discussion of the role of metabolism in the non-cancer respiratory tract toxicity of
TCE. Therefore, overall, data are suggestive of TCE causing respiratory tract toxicity, based
primarily on short-term studies in mice and rats, with available human data too few and limited
to add to the weight of evidence for pulmonary toxicity.
6.1.3.6 Reproductive effects (Sections 4.7.1 and 4.10.1.6)
A number of human and laboratory animal studies suggest that TCE exposure has the
potential for male reproductive toxicity, with a more limited number of studies examining female
reproductive toxicity. Human studies have reported TCE exposure to be associated (in all but
one case statistically-significantly) with increased sperm density and decreased sperm quality,
altered sexual drive or function, or altered serum endocrine levels. Measures of male fertility,
however, were either not reported or reported to be unchanged with TCE exposure, though the
statistical power of the available studies is quite limited. Epidemiologic studies have identified
possible associations of TCE exposure with effects on female fertility and with menstrual cycle
disturbances, but these data are fewer than those available for male reproductive toxicity. See
Section 4.7.1.1 for additional discussion of human data on the reproductive effects of TCE.
Evidence of similar effects, particularly for male reproductive toxicity, is provided by several
laboratory animal studies that reported effects on sperm, libido/copulatory behavior, and serum
hormone levels, although some studies that assessed sperm measures did not report treatment-
related alterations. Additional adverse effects on male reproduction have also been reported,
including histopathological lesions in the testes or epididymides and altered in vitro sperm-
oocyte binding or in vivo fertilization due to TCE or metabolites. While reduced fertility in
rodents was only observed in one study, this is not surprising given the redundancy and
efficiency of rodent reproductive capabilities. In addition, although the reduced fertility
observed in the rodent study was originally attributed to systemic toxicity, the database as a
whole suggests that TCE does induce reproductive toxicity independent of systemic effects.
6/22/2009
1192
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Fewer data are available in rodents on female reproductive toxicity. While in vitro oocyte
fertilizability has been reported to be reduced as a result of TCE exposure in rats, a number of
other laboratory animal studies did not report adverse effects on female reproductive function.
See Section 4.7.1.2 for additional discussion of laboratory animal data on the reproductive
effects of TCE. Very limited data are available to elucidate the MOA for these effects, though
some aspects of a putative MOA (e.g., perturbations in testosterone biosynthesis) appear to have
some commonalities between humans and animals (Section 4.7.1.3.2). Together, the human and
laboratory animal data support the conclusion that TCE exposure poses a potential hazard to the
male reproductive system, but are more limited with regard to the potential hazard to the female
reproductive system.
6.1.3.7 Developmental effects (Sections 4.7.3 and 4.10.1.7)
The relationship between TCE exposure (direct or parental) and developmental toxicity
has been investigated in a number of epidemiologic and laboratory animal studies. Postnatal
developmental outcomes examined include developmental neurotoxicity (addressed above with
neurotoxicity), developmental immunotoxicity (addressed above with immunotoxicity), and
childhood cancers. Prenatal effects examined include death (spontaneous abortion, perinatal
death, pre- or post-implantation loss, resorptions), decreased growth (low birth weight, small for
gestational age, intrauterine growth restriction, decreased postnatal growth), and congenital
malformations, in particular cardiac defects. Some epidemiological studies have reported
associations between parental exposure to TCE and spontaneous abortion or perinatal death, and
decreased birth weight or small for gestational age, although other studies reported mixed or null
findings. While comprising both occupational and environmental exposures, these studies are
overall not highly informative due to the small numbers of cases and limited exposure
characterization or to the fact that exposures were to a mixture of solvents. See Section 4.7.3.1
for additional discussion of human data on the developmental effects of TCE. However,
multiple well-conducted studies in rats and mice show analogous effects of TCE exposure,
including pre- or post-implantation losses, increased resorptions, perinatal death, and decreased
birth weight Interestingly, the rat studies reporting these effects used Fischer 344 or Wistar rats,
while several other studies, all of which used Sprague-Dawley rats, reported no increased risk in
these developmental measures, suggesting a strain difference in susceptibility. See Section
4.7.3.2 for additional discussion of laboratory animal data on the developmental effects of TCE.
Therefore, overall, based on weakly suggestive epidemiologic data and fairly consistent
laboratory animal data, it can be concluded that TCE exposure poses a potential hazard for
prenatal losses and decreased growth or birth weight of offspring.
6/22/2009
1193
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
With respect to congenital malformations, epidemiology and experimental animal studies
of TCE have reported increases in total birth defects, CNS defects, oral cleft defects, eye/ear
defects, kidney/urinary tract disorders, musculoskeletal birth anomalies, lung/respiratory tract
disorders, skeletal defects, and cardiac defects. Human occupational cohort studies, while not
consistently reporting positive results, are generally limited by the small number of observed or
expected cases of birth defects. While only one of the epidemiological studies specifically
reported observations of eye anomalies, studies in rats have identified increases in the incidence
of fetal eye defects following oral exposures during the period of organogenesis with TCE or its
oxidative metabolites DCA and TCA. The epidemiological studies, while individually limited,
as a whole show relatively consistent elevations, some of which were statistically significant, in
the incidence of cardiac defects in TCE-exposed populations compared to reference groups. In
laboratory animal models, avian studies were the first to identify adverse effects of TCE
exposure on cardiac development, and the initial findings have been confirmed multiple times.
Additionally, administration of TCE and its metabolites TCA and DCA in maternal drinking
water during gestation has been reported to induce cardiac malformations in rat fetuses. It is
notable that a number of other studies, several of which were well-conducted, did not report
induction of cardiac defects in rats, mice, or rabbits in which TCE was administered by
inhalation or gavage. However, many of these studies used a traditional free-hand section
technique on fixed fetal specimens, and a fresh dissection technique that can enhance detection
of anomalies was used in the positive studies by Dawson et al. (1993) and Johnson et al. (2003,
2005). Nonetheless, two studies that used the same or similar fresh dissection technique did not
report cardiac anomalies. Differences in other aspects of experimental design may have been
contributing factors to the differences in observed response. In addition, mechanistic studies,
such as the treatment-related alterations in endothelial cushion development observed in avian in
ovo and in vitro studies, provide a plausible mechanistic basis for defects in septal and valvular
morphogenesis observed in rodents, and consequently support the plausibility of cardiac defects
induced by TCE in humans. Therefore, while the studies by Dawson et al. (1993) and Johnson et
al. (2003, 2005) have significant limitations, including the lack of clear dose-response
relationship for the incidence of any specific cardiac anomaly and the pooling of data collected
over an extended period, there is insufficient reason to dismiss their findings. See Section
4.7.3.3.2 for additional discussion of the conclusions with respect to TCE-induced cardiac
malformations. Therefore, overall, based on weakly suggestive, but overall consistent,
epidemiologic data, in combination with evidence from experimental animal and mechanistic
studies, it can be concluded that TCE exposure poses a potential hazard for congenital
malformations, including cardiac defects, in offspring.
6/22/2009
1194
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
6.1.4 Carcinogenicity (Sections 4.0, 4.1, 4.3.2, 4.3.5, 4.3.7, 4.4.2, 4.4.5, 4.4.6, 4.4.7, 4.5.1.2,
4.5.2.4, 4.6.1.2, 4.6.2.2, 4.6.4, 4.7.2, 4.8, and 4.10.2, and Appendices B and C)
In 1995, International Agency for Research on Cancer (IARC) concluded that
trichloroethylene is "probably carcinogenic to humans" (IARC, 1995). In 2000, National
Toxicology Program (NTP) concluded that trichloroethylene is "reasonably anticipated to be a
human carcinogen." (NTP, 2000). In 2001, the draft U.S. EPA health risk assessment of TCE
concluded that TCE was "highly likely" to be carcinogenic in humans. In 2006, a committee of
the National Research Council stated that "findings of experimental, mechanistic, and
epidemiologic studies lead to the conclusion that trichloroethylene can be considered a potential
human carcinogen" (NRC, 2006).
Following U.S. EPA (2005a) Guidelines for Carcinogen Risk Assessment, based on the
available data as of 2009, TCE is characterized as carcinogenic in humans by all routes of
exposure. This conclusion is based on convincing evidence of a causal association between TCE
exposure in humans and kidney cancer. The strong consistency of the epidemiologic data on
TCE and kidney cancer argues against chance, bias, and confounding as explanations for the
elevated kidney cancer risks. In addition, statistically significant exposure-response trends are
observed in high-quality studies. These studies were designed to examine kidney cancer in
populations with high TCE exposure intensity. These studies addressed important potential
confounders and biases, further supporting the observed associations with kidney cancer as
causal. See Section 4.3.2 for additional discussion of the human epidemiologic data on TCE
exposure and kidney cancer. Meta-analyses of 14 high-quality studies show that estimated
relative risks or odds ratios in cohort and case-control studies are consistent, robust, and
insensitive to individual study inclusion, with no indication of publication bias or significant
heterogeneity. A statistically significant pooled relative risk estimate was observed for overall
TCE exposure [RRp=1.26 (95% CI: 1.11, 1.42)], and the pooled relative risk estimate was
greater for the highest TCE exposure groups [RRp=1.55 (95% CI: 1.24, 1.94)]. See Section
4.3.2.5 and Appendix C for additional discussion of the kidney cancer meta-analysis. Given the
modest relative risk estimates and the relative rarity of the cancers observed, and therefore the
limited statistical power of individual studies, the consistency of the database is compelling. It
would require a substantial amount of high-quality negative data in order to rule out this
observed association.
The human evidence of carcinogenicity from epidemiologic studies of TCE exposure is
compelling for Non-Hodgkins Lymphoma but less convincing than for kidney cancer. High
quality studies generally reported excess relative risk estimates, with statistically significant
increases in three studies, and a statistically significant trend with TCE exposure in one study
6/22/2009
1195
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
(see Section 4.5.1.2). The consistency of the association between TCE exposure and lymphoma
is further supported by the results of meta-analyses (see Section 4.5.1.2.2 and Appendix C). A
statistically significant pooled relative risk estimate was observed for overall TCE exposure
[RRp=l .27 (95% CI: 1.04, 1.53)], and, as with kidney cancer, the pooled relative risk estimate
was greater for the highest TCE exposure groups [RRp=1.50 (95% CI: 1.20, 1.88)] than for
overall TCE exposure. Sensitivity analyses indicated that this result and its statistical
significance were not overly influenced by most individual studies or choice of individual
(study-specific) risk estimates, although in a few cases the resulting pooled relative risk estimates
had a lower confidence bound of 0.99 or 1.00. Some heterogeneity was observed, particularly
between cohort and case-control studies, and, in addition, there was some evidence of potential
publication bias. Thus, while the evidence is strong for NHL, issues of study heterogeneity,
potential publication bias, and weaker exposure-response results contribute greater uncertainty.
The evidence is more limited for liver and biliary tract cancer mainly because only cohort
studies are available and most of these studies have small numbers of cases due the comparative
rarity of liver and biliary tract cancer. While most high quality studies reported excess relative
risk estimates, they were generally based on small numbers of cases or deaths, with the result of
wide confidence intervals on the estimates. The low number of liver cancer cases in the
available studies made assessing exposure-response relationships difficult. See Section 4.4.2 for
additional discussion of the human epidemiologic data on TCE exposure and liver cancer. A
consistency of the association between TCE exposure and liver cancer is supported by the results
of meta-analyses (see Section 4.4.2 and Appendix C). These meta-analyses found a statistically
significant increased pooled relative risk estimate for liver and biliary tract cancer of 1.34 (95%
CI: 1.09, 1.65) with overall TCE exposure; but the meta-analyses using only the highest
exposure groups yielded a lower, and non-statistically significant, pooled estimate for primary
liver cancer [1.25 (95% CI: 0.87, 1.79)]. Although there was no evidence of heterogeneity or
publication bias and the pooled estimates were fairly insensitive to the use of alternative relative
risk estimates, the statistical significance of the pooled estimates depends heavily on the one
large study by Raaschou-Nielsen et al. (2003). There were fewer adequate, high quality studies
available for meta-analysis of liver cancer (9 versus 15 for NHL and 14 for kidney), leading to
lower statistical power, even with pooling. Thus, while there is epidemiologic evidence of an
association between TCE exposure and liver cancer, the much more limited database, both in
terms of number of available studies and number of cases upon which the studies are based,
contributes to greater uncertainty as compared to the evidence for either kidney cancer or NHL.
There are several other lines of supporting evidence for TCE carcinogenicity in humans.
First, multiple chronic bioassays in rats and mice have reported increased incidences of tumors
with TCE treatment, including tumors in the kidney, liver, and lymphoid tissues - target tissues
6/22/2009
1196
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
of TCE carcinogenicity also seen in epidemiological studies. Of particular note is the site-
concordant finding of low, but biologically and sometimes statistically significant, increases in
the incidence of kidney tumors in multiple strains of rats treated with TCE by either inhalation or
corn oil gavage (see Section 4.3.5). The increased incidences were greater in male than female
rats, although, notably, pooled incidences in females from five rat strains tested by NTP (1988,
1990) resulted in a statistically significant trend. While individual studies provide only
suggestive evidence of renal carcinogenicity, the database as a whole supports the conclusion
that TCE is a kidney carcinogen in rats, with males being more sensitive than females. No other
tested laboratory species (i.e., mice and hamsters) have exhibited increased kidney tumors, with
no adequate explanation for these species differences (particularly with mice, which have been
extensively tested). With respect to the liver, TCE and its oxidative metabolites CH, TCA, and
DCA are clearly carcinogenic in mice, with strain and sex differences in potency that appear to
parallel, qualitatively, differences in background tumor incidence. Data in other laboratory
animal species are limited; thus, except for DCA which is carcinogenic in rats, inadequate
evidence exists to evaluate the hepatocarcinogenicity of these compounds in rats or hamsters.
However, to the extent that there is hepatocarcinogenic potential in rats, TCE is clearly less
potent in the strains tested in this species than in B6C3F1 and Swiss mice. See Section 4.4.5 for
additional discussion of laboratory animal data on TCE-induced liver tumors. Additionally,
there is more limited evidence for TCE-induced lymphatic cancers in rats and mice, lung tumors
in mice, and testicular tumors in rats. With respect to the lymphatic cancers, two studies in mice
reported increased incidences of lymphomas in females of two different strains, and two studies
in rats reported leukemias in males of one strain and females of another. However, these tumors
had relatively modest increases in incidence with treatment, and were not reported to be
increased in other studies. See Section 4.5.2.4 for additional discussion of laboratory animal data
on TCE-induced lymphatic tumors. With respect to lung tumors, rodent bioassays have
demonstrated a statistically significant increase in pulmonary tumors in mice following chronic
inhalation exposure to TCE, and non-statistically significant increases in mice exposed orally;
but pulmonary tumors were not reported in other species tested (i.e., rats and hamsters) (see
Section 4.6.2.2). Finally, increased testicular (interstitial or Leydig cell) tumors have been
observed in multiples studies of rats exposed by inhalation and gavage, although in some cases
high (>75%) control rates of testicular tumors in rats limited the ability to detect a treatment
effect. See Section 4.7.2.2 for additional discussion of laboratory animal data on TCE-induced
tumors of the reproductive system. Overall, TCE is clearly carcinogenic in rats and mice. The
apparent lack of site concordance across laboratory animal studies may be due to limitations in
design or conduct in a number of rat bioassays and/or genuine inter-species differences in
qualitative or quantitative sensitivity (i.e., potency). Nonetheless, these studies have shown
6/22/2009
1197
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
carcinogenic effects across different strains, sexes, and routes of exposure, and site-concordance
is not necessarily expected for carcinogens.
A second line of supporting evidence for TCE carcinogenicity in humans consists of
toxicokinetic data indicating that TCE absorption, distribution, metabolism, and excretion are
qualitatively similar in humans and rodents. As summarized above, there is no evidence of
major qualitative differences across species in TCE absorption, distribution, metabolism, and
excretion. Extensive in vivo and in vitro data show that mice, rats, and humans all metabolize
TCE via two primary pathways: oxidation by CYP450s and conjugation with glutathione via
GSTs. Several metabolites and excretion products from both pathways have been detected in
blood and urine from exposed humans was well as from at least one rodent species. In addition,
the subsequent distribution, metabolism, and excretion of TCE metabolites are qualitatively
similar among species. Therefore, humans possess the metabolic pathways that produce the TCE
metabolites thought to be involved in the induction of rat kidney and mouse liver tumors, and
internal target tissues of both humans and rodents experience a similar mix of TCE and
metabolites. See Sections 3.1-3.4 for additional discussion of TCE toxicokinetics. Quantitative
inter-species differences in toxicokinetics do exist, and are addressed through PBPK modeling
(see Section 3.5 and Appendix A). Importantly, these quantitative differences affect only inter-
species extrapolations of carcinogenic potency, and do not affect inferences as to the
carcinogenic hazard for TCE.
Finally, available mechanistic data do not suggest a lack of human carcinogenic hazard
from TCE exposure. In particular, these data do not suggest qualitative differences between
humans and test animals that would preclude any of the hypothesized key events in the
carcinogenic MO A in rodents from occurring in humans. For the kidney, the predominance of
positive genotoxicity data in the database of available studies of TCE metabolites derived from
GSH conjugation (in particular DCVC), together with toxicokinetic data consistent with their
systemic delivery to and in situ formation in the kidney, supports the conclusion that a mutagenic
MOA is operative in TCE-induced kidney tumors. While supporting the biological plausibility of
this hypothesized MOA, available data on the VHL gene in humans or transgenic animals do not
conclusively elucidate the role of VHL mutation in TCE-induced renal carcinogenesis.
Cytotoxicity and compensatory cell proliferation, similarly presumed to be mediated through
metabolites formed after GSH-conjugation of TCE, have also been suggested to play a role in the
MOA for renal carcinogenesis, as high incidences of nephrotoxicity have been observed in
animals at doses that induce kidney tumors. Human studies have reported markers for
nephrotoxicity at current occupational exposures, although data are lacking at lower exposures.
Nephrotoxicity alone appears to be insufficient, or at least not rate-limiting, for rodent renal
carcinogenesis, since, although very high incidences of toxicity are observed in both mice and
6/22/2009
1198
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
rats, kidney tumors are only observed at low incidences in rats. In addition, nephrotoxicity has
not been shown to be necessary for kidney tumor induction by TCE in rodents. In particular,
there is a lack of experimental support for causal links, such as compensatory cellular
proliferation or clonal expansion of initiated cells, between nephrotoxicity and kidney tumors
induced by TCE. Furthermore, it is not clear if nephrotoxicity is one of several key events in a
MO A, if it is a marker for an "upstream" key event (such as oxidative stress) that may contribute
independently to both nephrotoxicity and renal carcinogenesis, or if it is incidental to kidney
tumor induction. Moreover, while toxicokinetic differences in the GSH conjugation pathway
along with their uncertainty are addressed through PBPK modeling, no data suggest that any of
the proposed key events for TCE-induced kidney tumors in rats are precluded in humans. See
Section 4.3.7 for additional discussion of the MOA for TCE-induced kidney tumors. Therefore,
TCE-induced rat kidney tumors provide additional support for the convincing human evidence of
TCE-induced kidney cancer, with mechanistic data supportive of a mutagenic MOA.
With respect to other tumor sites, data are insufficient to conclude that any of the other
hypothesized MO As are operant. In the liver, a mutagenic MOA mediated by CH, which has
evidence for genotoxic effects, or some other oxidative metabolite of TCE cannot be ruled out,
but data are insufficient to conclude it is operant. A second MOA hypothesis for TCE-induced
liver tumors involves activation of the PPARa receptor. Clearly, in vivo administration of TCE
leads to activation of PPARa in rodents and likely does so in humans as well. However, the
evidence as a whole does not support the view that PPAR-a is the sole operant MOA mediating
TCE hepatocarcinogenesis. Rather, there is evidential support for multiple TCE metabolites and
multiple toxicity pathways contributing to TCE-induced liver tumors. Furthermore, recent
experiments have demonstrated that PPAR-a activation and the sequence of key events in the
hypothesized MOA are not sufficient to induce hepatocarcinogenesis (Yang et al., 2007).
Moreover, the demonstration that the PPAR-a agonist DEHP induces tumors in PPAR-a-null
mice supports the view that the events comprising the hypothesized PPAR-a activation MOA are
not necessary for liver tumor induction in mice by this PPARa agonist (Ito et al. 2007). See
Section 4.4.7 for additional discussion of the MOA for TCE-induced liver tumors. For mouse
lung tumors, as with the liver, a mutagenic MOA involving CH has also been hypothesized, but
there are insufficient data to conclude that it is operant. A second MOA hypothesis for mouse
lung tumors has been posited involving other effects of oxidative metabolites including
cytotoxicity and regenerative cell proliferation, but experimental support remains limited, with
no data on proposed key events in experiments of duration 2 weeks or longer. See Section 4.6.4
for additional discussion of the MOA for TCE-induced lung tumors. A MOA subsequent to in
situ oxidative metabolism, whether involving mutagenicity, cytotoxicity, or other key events,
may also be relevant to other tissues where TCE would undergo P450 metabolism. For instance,
6/22/2009
1199
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
CYP2E1, oxidative metabolites, and protein adducts have been reported in the testes of rats
exposed to TCE, and, in some rat bioassays, TCE exposure increased the incidence of rat
testicular tumors. However, inadequate data exist to adequately define a MOA hypothesis for
this tumor site (see Section 4.7.2.3 for additional discussion of the MOA for TCE-induced
testicular tumors).
6.1.5 Susceptibility (Sections 4.9 and 4.10.3)
There is some evidence that certain populations may be more susceptible to exposure to
TCE. Factors affecting susceptibility examined include lifestage, gender, genetic
polymorphisms, race/ethnicity, pre-existing health status, and lifestyle factors and nutrition
status. Factors that impact early lifestage susceptibilty include exposures such as transplacental
transfer and breast milk ingestion, early lifestage-specific toxicokinetics, and differential
outcomes in early lifestages such as developmental cardiac defects (see Section 4.9.1). Because
the weight of evidence supports a mutagenic MOA for TCE carcinogenicity in the kidney (see
Section 4.3.7), and there is an absence of chemical-specific data to evaluate differences in
carcinogenic susceptibility, early-life susceptibility should be assumed and the age-dependent
adjustment factors (ADAFs) should be applied, in accordance with the Supplemental Guidance
(see summary below in Section 6.2.2.5). Fewer data are available on later lifestages, although
there is suggestive evidence to indicate that older adults may experience increased adverse
effects than younger adults due to greater tissue distribution of TCE. In general, more studies
specifically designed to evaluate effects in early and later lifestages are needed in order to more
fully characterize potential life stage-related TCE toxicity. Gender-specific (Section 4.9.2.1)
differences also exist in toxicokinetics (e.g., cardiac outputs, percent body fat, expression of
metabolizing enzymes) and susceptibility to toxic endpoints (e.g., gender-specific effects on the
reproductive system, gender differences in baseline risks to endpoints such as scleroderma or
liver cancer). Genetic variation (Section 4.9.2.2) likely has an effect on the toxicokinetics of
TCE. Increased CYP2E1 activity and GST polymorphisms may influence susceptibility of TCE
due to effects on production of toxic metabolites or may play a role in variability in toxic
response. Differences in genetic polymorphisms related to the metabolism of TCE have been
observed among various race/ethnic groups (Section 4.9.2.3). Pre-existing diminished health
status (Section 4.9.2.4) may alter the response to TCE exposure. Individuals with increased body
mass may have an altered toxicokinetic response due to the increased uptake of TCE into fat.
Other conditions that may alter the response to TCE exposure include diabetes and hypertension,
and lifestyle and nutrition factors (Section 4.9.2.5) such alcohol consumption, tobacco smoking,
nutritional status, physical activity, and socioeconomic status. Alcohol intake has been
6/22/2009
1200
-------
1
2
3
4
5
6
7
8
9
10
11
12
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
associated with inhibition of TCE metabolism in both humans and experimental animals. In
addition, such conditions have been associated with increased baseline risks for health effects
also associated with TCE, such as kidney cancer and liver cancer. However, the interaction
between TCE and known risk factors for human diseases is not known, and further evaluation of
the effects due to these factors is needed.
In sum, there is some evidence that certain populations may be more susceptible to
exposure to TCE. Factors affecting susceptibility examined include lifestage, gender, genetic
polymorphisms, race/ethnicity, pre-existing health status, and lifestyle factors and nutrition
status. However, except in the case of toxicokinetic variability characterized using the PBPK
model described in Section 3.5, there are inadequate chemical-specific data to quantify the
degree of differential susceptibility due to such factors.
6/22/2009
1201
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
6.2 Dose-Response Assessment
This section summarizes the major conclusions of the dose-response analysis for TCE
non-cancer effects and carcinogenicity, with more detailed discussions in Chapter 5.
6.2.1 Non-cancer effects (Section 5.1)
6.2.1.1 Background and methods
As summarized above, based on the available human epidemiologic data and
experimental and mechanistic studies, it is concluded that TCE poses a potential human health
hazard for non-cancer toxicity to the central nervous system, the kidney, the liver, the immune
system, the male reproductive system, and the developing fetus. The evidence is more limited
for TCE toxicity to the respiratory tract and female reproductive system.
Dose-response analysis for a non-cancer endpoint generally involves two steps: (i) the
determination of a point of departure (POD) derived from a benchmark dose (BMD)60, a no
observed adverse effect level (NOAEL), or a lowest observed adverse effect level (LOAEL); and
(ii) adjustment of the POD by endpoint/study-specific "uncertainty factors" (UFs), accounting
for adjustments and uncertainties in the extrapolation from the study conditions to conditions of
human exposure.
Because of the large number of non-cancer health effects associated with TCE exposure
and the large number of studies reporting on these effects, in contrast to toxicological reviews for
chemicals with smaller databases of studies, a formal, quantitative screening process (see Section
5.1) was used to reduce the number of endpoints and studies to those that would best inform the
selection of the critical effects for the inhalation reference concentration (RfC) and oral reference
dose (RfD).61 As described in Section 5.1, for all studies described in Chapter 4 which report
adverse non-cancer health effects and provided quantitative dose-response data, PODs on the
basis of applied dose, adjusted by endpoint/study-specific UFs, were used to develop candidate
RfCs (cRfCs) and candidate RfDs (cRfDs) intended to be protective for each endpoint
individually. Candidate critical effects - those with the lowest cRfCs and cRfDs taking into
60 more precisely, it is the benchmark dose lower bound (BMDL), i.e., the (one-sided) 95% lower confidence bound
on the dose corresponding to the benchmark response (BMR) for the effect, that is used as the POD
61 In EPA non-cancer health assessments, the RfC [RfD] is an estimate (with uncertainty spanning perhaps an order
of magnitude) of a continuous inhalation [daily oral] exposure to the human population (including sensitive
subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime. It can be derived
from a NOAEL, LOAEL, or benchmark concentration [dose], with uncertainty factors generally applied to reflect
limitations of the data used.
6/22/2009
1202
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
account the confidence in each estimate - were selected within each of the following health
effect domains: (i) neurological, (ii) systemic/organ system; (iii) immunological; (iv)
reproductive; and (v) developmental. For each of these candidate critical effects, the PBPK
model developed in Section 3.5 was used for inter-species, intra-species, and route-to-route
extrapolation on the basis of internal dose to develop PBPK-model-based PODs. Plausible
internal dose metrics were selected based on what is understood about the role of different TCE
metabolites in toxicity and the MOA for toxicity. These PODs were then adjusted by
endpoint/study-specific UFs, taking into account the use of the PBPK model, to develop PBPK
model-based candidate RfCs (p-cRfCs) and candidate RfDs (p-cRfDs). The most sensitive
cRfCs, p-cRfCs, cRfDs, and p-cRfDs were then evaluated, taking into account the confidence in
each estimate, to arrive at overall candidate RfCs and RfDs for each health effect type. Then, the
RfC and RfD for TCE was selected so as to be protective of the most sensitive effects. In
contrast to the approach used in most assessments, in which the RfC and RfD are each based on
a single critical effect, the final RfC and RfD for TCE were based on multiple critical effects that
resulted in very similar candidate RfC and RfD values at the low end of the full range of values.
This approach was taken here because it provides robust estimates of the RfC and RfD and
because it highlights the multiple effects that are all yielding very similar candidate values.
6.2.1.2 Uncertainties and application of uncertainty factors (UFs) (Section 5.1.1 and 5.1.4)
An underlying assumption in deriving reference values for noncancer effects is that the
dose-response relationship for these effects has a threshold. Thus, a fundamental uncertainty is
the validity of that assumption. For some effects, in particular effects on very sensitive processes
(e.g., developmental processes) or effects for which there is a nontrivial background level and
even small exposures may contribute to background disease processes in more susceptible
people, a practical threshold (i.e., a threshold within the range of environmental exposure levels
of regulatory concern) may not exist.
Nonetheless, under the assumption of a threshold, the desired exposure level to have as a
reference value is the maximum level at which there is no appreciable risk for an adverse effect
in sensitive subgroups (of humans). However, because it is not possible to know what this level
is, "uncertainty factors" are used to attempt to address quantitatively various aspects, depending
on the dataset, of qualitative uncertainty.
First there is uncertainty about the "point of departure" for the application of UFs.
Conceptually, the POD should represent the maximum exposure level at which there is no
appreciable risk for an adverse effect in the study population under study conditions (i.e., the
threshold in the dose-response relationship). Then, the application of the relevant UFs is
intended to convey that exposure level to the corresponding exposure level for sensitive human
6/22/2009
1203
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
subgroups exposed continuously for a lifetime. In fact, it is again not possible to know that
exposure level even for a laboratory study because of experimental limitations (e.g. the power to
detect an effect, dose spacing, measurement errors, etc.), and crude approximations like the
NOAEL or a BMDL are used. If a LOAEL is used as the POD, the LOAEL-to-NOAEL UF is
applied as an adjustment factor to better approximate the desired exposure level (threshold),
although the necessary extent of adjustment is unknown. The standard value for the LOAEL-to-
NOAEL UF is 10, although sometimes a value of 3 is used if the effect is considered minimally
adverse at the response level observed at the LOAEL or even 1 if the effect is an early marker for
an adverse effect. For one POD in this assessment, a value of 30 was used for the LOAEL-to-
NOAEL UF because the incidence rate for the adverse effect was > 90% at the LOAEL.
If a BMDL is used as the POD, there are uncertainties regarding the appropriate dose-
response model to apply to the data, but these should be minimal if the modeling is in the
observable range of the data. There are also uncertainties about what BMR to use to best
approximate the desired exposure level (threshold, see above). For continuous endpoints, in
particular, it is often difficult to identify the level of change that constitutes the "cut-point" for an
adverse effect. Sometimes, to better approximate the desired exposure level, a BMR somewhat
below the observable range of the data is selected. In such cases, the model uncertainty is
increased, but this is a trade-off to reduce the uncertainty about the POD not being a good
approximation for the desired exposure level.
For each of these types of PODs, there are additional uncertainties pertaining to
adjustments to the administered exposures (doses). Typically, administered exposures (doses)
are converted to equivalent continuous exposures (daily doses) over the study exposure period
under the assumption that the effects are related to concentration x time, independent of the daily
(or weekly) exposure regimen (i.e., a daily exposure of 6 hours to 4 ppm is considered equivalent
to 24 hours of exposure to 1 ppm). However, the validity of this assumption is generally
unknown, and, if there are dose-rate effects, the assumption of C x t equivalence would tend to
bias the POD downwards. Where there is evidence that administered exposure better correlates
to the effect than equivalent continuous exposure averaged over the study exposure period (e.g.,
visual effects), administered exposure was not adjusted. For the PBPK analyses in this
assessment, the actual administered exposures are taken into account in the PBPK modeling, and
equivalent daily values (averaged over the study exposure period) for the dose metrics are
obtained (see above, 5.1.3.2). Additional uncertainties about the PBPK-based estimates include
uncertainties about the appropriate dose metric for each effect, although for some effects there
was better information about relevant dose metrics than for others (see Section 5.1.3.1).
There is also uncertainty about the other UFs. The human variability UF is to some
extent an adjustment factor because for more sensitive people, the dose-response relationship
6/22/2009
1204
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
shifts to lower exposures. But there is uncertainty about the extent of the adjustment required,
i.e., about the distribution of human susceptibility. Therefore, in the absence of data on a
susceptible population(s) or on the distribution of susceptibility in the general population, an UF
of 10 is generally used, which breaks down (approximately) to a factor of 3 for pharmacokinetic
variability and a factor of 3 for pharmacodynamic variability. This standard value was used for
all the PODs based on applied dose in this assessment with the exception of the PODs for a few
immunological effects that were based on data from a sensitive (autoimmune-prone) mouse
strain. For those PODs, an UF of 3 (reflecting pharmacokinetics only) was used for human
variability. The PBPK analyses in this assessment attempt to account for the pharmacokinetic
portion of human variability using human data on pharmacokinetic variability. For PBPK
model-based candidate reference values, the pharmacokinetic component of this UF was omitted.
A quantitative uncertainty analysis of the PBPK-derived dose metrics used in the assessment is
presented in Section 5.1.4.2 in Chapter 5. There is still uncertainty regarding the susceptible
subgroups for TCE exposure and the extent of pharmacodynamic variability.
If the data used to determine a particular POD are from laboratory animals, an
interspecies extrapolation UF is used. This UF is also to some extent an adjustment factor for the
expected scaling for toxicologically-equivalent doses across species (i.e., according to body
weight to the 3/4 power for oral exposures). However, there is also uncertainty about the true
extent of interspecies differences for specific noncancer effects from specific chemical
exposures. For oral exposures, the standard value for the interspecies UF is 10, which can be
viewed as breaking down (approximately) to a factor of 3 for the "adjustment" (nominally
pharmacokinetics) and a factor of 3 for the "uncertainty" (nominally pharmacodynamics). For
inhalation exposures for systemic toxicants such as TCE, no adjustment across species is
generally assumed for fixed air concentrations (ppm equivalence), and the standard value for the
interspecies UF is 3 reflects "uncertainty" (nominally pharmacodynamics only). The PBPK
analyses in this assessment attempt to account for the "adjustment" portion of interspecies
extrapolation using rodent pharmacokinetic data to estimate internal doses for various dose
metrics. Equal doses of these dose metrics, appropriately scaled, are then assumed to convey
equivalent risk across species. For PBPK model-based candidate reference values, the
"adjustment" component of this UF was omitted. With respect to the "uncertainty" component,
quantitative uncertainty analyses of the PBPK-derived dose metrics used in the assessment are
presented in Section 5.1.4.2 in Chapter 5. However, these only address the pharmacokinetic
uncertainties in a particular dose metric, and there is still uncertainty regarding the true dose
metrics. Nor do the PBPK analyses address the uncertainty in either cross-species
pharmacodynamic differences (i.e., about the assumption that equal doses of the appropriate dose
metric convey equivalent risk across species for a particular endpoint from a specific chemical
6/22/2009
1205
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
exposure) or in cross-species pharmacokinetic differences not accounted for by the PBPK model
dose metrics (e.g., departures from the assumed inter-species scaling of clearance of the active
moiety, in the cases where only its production is estimated). A value of 3 is typically used for
the "uncertainty" about cross-species differences, and this generally represents true uncertainty
because it is usually unknown, even after adjustments have been made to account for the
expected interspecies differences, whether humans have more or less susceptibility, and to what
degree, than the laboratory species in question.
RfCs and RfDs apply to lifetime exposure, but sometimes the best (or only) available
data come from less-than-lifetime studies. Lifetime exposure can induce effects that may not be
apparent or as large in magnitude in a shorter study; consequently, a dose that elicits a specific
level of response from a lifetime exposure may be less than the dose eliciting the same level of
response from a shorter exposure period. If the effect becomes more severe with increasing
exposure, then chronic exposure would shift the dose-response relationship to lower exposures,
although the true extent of the shift is unknown. PODs based on subchronic exposure data are
generally divided by a subchronic-to-chronic UF, which has a standard value of 10. If there is
evidence suggesting that exposure for longer time periods does not increase the magnitude of an
effect, a lower value of 3 or 1 might be used. For some reproductive and developmental effects,
chronic exposure is that which covers a specific window of exposure that is relevant for eliciting
the effect, and subchronic exposure would correspond to an exposure that is notably less than the
full window of exposure.
Sometimes a database UF is also applied to address limitations or uncertainties in the
database. The overall database for TCE is quite extensive, with studies for many different types
of effects, including 2-generation reproductive studies, as well as neurological and
immunological studies. In addition, there were sufficient data to develop a reliable PBPK model
to estimate route-to-route extrapolated doses for some candidate critical effects for which data
were only available for one route of exposure. Thus, there is a high degree of confidence that the
TCE database was sufficient to identify some sensitive endpoints, and no database UF was used
in this assessment.
6.2.1.3 Candidate critical effects and reference values (Sections 5.1.2 and 5.1.3)
A large number of endpoints and studies were considered within each health effect
domain. Chapter 5 contains a comprehensive discussion of all endpoints/studies which were
considered for developing candidate reference values (cRfCs, cRfDs, p-cRfCs, and p-cRfDs),
their PODs, and the UFs applied. The summary below reviews the selection of candidate critical
effects for each health effect domain, the confidence in the reference values, the selection of
6/22/2009
1206
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
PBPK model-based dose metrics, and the impact of PBPK modeling on the candidate reference
values.
6.2.1.3.1 Neurological effects
Candidate reference values were developed for several neurological domains for which
there was evidence of hazard (Tables 5.1.1 and 5.1.8). There is higher confidence in the
candidate reference values for trigeminal nerve, auditory, or psychomotor effects, but the
available data suggest that the more sensitive indicators of TCE neurotoxicity are changes in
wakefulness, regeneration of the sciatic nerve, demyelination in the hippocampus and
degeneration of dopaminergic neurons. Therefore, these more sensitive effects are considered
the candidate critical effects for neurotoxicity, albeit with more uncertainty in the corresponding
candidate reference values. Of these more sensitive effects, there is greater confidence in the
changes in wakefulness reported by Arito et al. (1994). In addition, trigeminal nerve effects are
considered a candidate critical effect because this is the only type of neurological effect for
which human data are available, and the POD for this effect is similar to that from the most
sensitive rodent study (Arito et al., 1994, for changes in wakefulness). Between the two human
studies of trigeminal nerve effects, Ruitjen et al. (1991) is preferred for deriving non-cancer
reference values because its exposure characterization is considered more reliable.
Because of the lack of specific data as to the metabolites involved and the MOA for the
candidate critical neurologic effects, PBPK model predictions of total metabolism (scaled by
body weight to the % power) were selected as the preferred dose metric based on the general
observation that TCE toxicity is associated with metabolism. The AUC of TCE in blood was
used as an alternative dose metric. With these dose metrics, the candidate reference values
derived using the PBPK model were only modestly (~3-fold or less) different than those derived
on the basis of applied dose.
6.2.1.3.2 Kidney effects
High confidence candidate reference values were developed for histopathological and
weight changes in the kidney (Tables 5.1.2 and 5.1.9), and these are considered to be candidate
critical effects for several reasons. First, they appear to be the most sensitive indicators of
toxicity that are available for the kidney. In addition, as discussed in Sections 3.3 and 3.5,
pharmacokinetic data indicate substantially more production of GSH-conjugates thought to
mediate TCE kidney effects in humans relative to rats and mice. Several studies are considered
reliable for developing candidate reference values for these endpoints. For histopathological
6/22/2009
1207
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
changes, these were the only available inhalation study (Maltoni et al. 1986), the NTP (1988)
study in rats, and the NCI (1976) study in mice. For kidney weight changes, both available
studies (Kjellstrand et al. 1983b; Woolhiser et al. 2006) were chosen as candidate critical studies.
Due to the substantial evidence supporting the role of GSH conjugation metabolites in
TCE-induced nephrotoxicity, the preferred PBPK model dose metrics for kidney effects were the
amount of DCVC bioactivated in the kidney for rat studies and the amount of GSH conjugation
(both scaled by body weight to the % power) for mouse studies (inadequate toxicokinetic data are
available in mice for predicting the amount of DCVC bioactivation). With these dose metrics,
the candidate reference values derived using the PBPK model were 300- to 400-fold lower than
those derived on the basis of applied dose. As discussed above and in Chapter 3, this is due to
the available data supporting not only substantially more GSH conjugation in humans than in
rodents, but also substantial inter-individual toxicokinetic variability.
6.2.1.3.3 Liver effects
Hepatomegaly appears to be the most sensitive indicator of toxicity that is available for
the liver and is therefore considered a candidate critical effect. Several studies are considered
reliable for developing high confidence candidate reference values for this endpoint. Since they
all indicated similar sensitivity but represented different species and/or routes of exposure, they
were all considered candidate critical studies (Tables 5.1.2 and 5.1.10).
Due to the substantial evidence supporting the role of oxidative metabolism in TCE-
induced hepatomegaly (and evidence against TC A being the sole mediator of TCE-induced
hepatomegaly [Evans et al., 2009]), the preferred PBPK model dose metric for liver effects was
the amount of hepatic oxidative metabolism (scaled by body weight to the % power). Total
(hepatic and extra-hepatic) oxidative metabolism (scaled by body weight to the % power) was
used as an alternative dose metric. With these dose metrics, the candidate reference values
derived using the PBPK model were only modestly (~3-fold or less) different than those derived
on the basis of applied dose.
6.2.1.3.4 Immunological effects
There is high qualitative confidence for TCE immunotoxicity and moderate confidence in
the candidate reference values that can be derived from the available studies (Tables 5.1.3 and
5.1.11). Decreased thymus weight reported at relatively low exposures in non-autoimmune-
prone mice is a clear indicator of immunotoxity (Keil et al. 2009), and is therefore considered a
candidate critical effect. A number of studies have also reported changes in markers of
immunotoxicity at relatively low exposures. Among markers for autoimmune effects, the more
sensitive measures of autoimmune changes in liver and spleen (Kaneko et al. 2000) and
increased anti-dsDNA and anti-ssDNA antibodies (early markers for systemic lupus
6/22/2009
1208
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
erythematosus) (Keil et al. 2009) are considered the candidate critical effects. For markers of
immunosuppression, the more sensitive measures of decreased PFC response (Woolhiser et al.
2006), decreased stem cell bone marrow recolonization, and decreased cell-mediated response to
sRBC (both from Sanders et al. 1982 ) are considered the candidate critical effects.
Developmental immunological effects are discussed below as part of the summary of
developmental effects (Section 6.2.1.3.6).
Because of the lack of specific data as to the metabolites involved and the MOA for the
candidate critical immunologic effects, PBPK model predictions of total metabolism (scaled by
body weight to the % power) was selected as the preferred dose metric based on the general
observation that TCE toxicity is associated with metabolism. The AUC of TCE in blood was
used as an alternative dose metric. With these dose metrics, the candidate reference values
derived using the PBPK model were, with one exception, only modestly (~3-fold or less)
different than those derived on the basis of applied dose. For the Woolhiser et al. (2006)
decreased PFC response, with the alternative dose metric of AUC of TCE in blood, BMD
modeling based on internal doses changed the candidate reference value by 17-fold higher than
the cRfC based on applied dose. However, the dose-response model fit for this effect using this
metric was substantially worse than the fit using the preferred metric of total oxidative
metabolism, with which the change in candidate reference value was only 1.3-fold.
6.2.1.3.5 Reproductive effects
While there is high qualitative confidence in the male reproductive hazard posed by TCE,
there is lower confidence in the reference values that can be derived from the available studies of
these effects (Tables 5.1.4 and 5.1.12). Relatively high PODs are derived from several studies
reporting less sensitive endpoints (George et al. 1985, 1986; Land et al. 1981), and
correspondingly higher cRfCs and cRfDs suggest that they are not likely to be critical effects.
The studies reporting more sensitive endpoints also tend to have greater uncertainty. For the
human study by Chia et al. (1996), there are uncertainties in the characterization of exposure and
the adversity of the effect measured in the study. For the Kumar et al. (2000a, 2000b, 2001),
Forkert et al. (2002) and Kan et al. (2007) studies, the severity of the sperm and testes effects
appears to be continuing to increase with duration even at the end of the study, so it is plausible
that a lower exposure for a longer duration may elicit similar effects. For the DuTeaux et al.
(2004b) study, there is also duration- and low-dose extrapolation uncertainty due to the short
duration of the study in comparison to the time period for sperm development as well as the lack
of a NOAEL at the tested doses. Overall, even though there are limitations in the quantitative
assessment, there remains sufficient evidence to consider these to be candidate critical effects.
There is moderate confidence both in the hazard and the candidate reference values for
reproductive effects other than male reproductive effects. While there are multiple studies
6/22/2009
1209
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
suggesting decreased maternal body weight with TCE exposure, this systemic change may not be
indicative of more sensitive reproductive effects. None of the estimates developed from other
reproductive effects is particularly uncertain or unreliable. Therefore, delayed parturition
(Narotsky et al. 1995) and decreased mating (George et al. 1986), which yielded the lowest
cRfDs, were considered candidate critical effects. These effects were also included so that
candidate critical reproductive effects from oral studies would not include only that reported by
DuTeaux et al. (2004b), from which deriving the cRfD entailed a higher degree of uncertainty.
Because of the general lack of specific data as to the metabolites involved and the MOA
for the candidate critical reproductive effects, PBPK model predictions of total metabolism
(scaled by body weight to the % power) was selected as the preferred dose metric based on the
general observation that TCE toxicity is associated with metabolism. The AUC of TCE in blood
was used as an alternative dose metric. The only exception to this was for the DuTeaux et al.
(2004) study, which suggested that local oxidative metabolism of TCE in the male reproductive
tract was involved in the effects reported. Therefore, in this case, AUC of TCE in blood was
considered the preferred dose metric, while total oxidative metabolism (scaled by body weight to
the 3/4 power) was considered the alternative metric. With these dose metrics, the candidate
reference values derived using the PBPK model were only modestly (~3.5-fold or less) different
than those derived on the basis of applied dose.
6.2.1.3.6 Developmental effects
There is moderate-to-high confidence both in the hazard and the candidate reference
values for developmental effects of TCE (Tables 5.1.5 and 5.1.13). It is also noteworthy that the
PODs for the more sensitive developmental effects were similar to or, in most cases, lower than
the PODs for the more sensitive reproductive effects, suggesting that developmental effects are
not a result of paternal or maternal toxicity. Among inhalation studies, candidate reference
values were only developed for effects in rats reported in Healy et al. (1982), of resorptions,
decreased fetal weight, and delayed skeletal ossification. These were all considered candidate
critical developmental effects. Because resorptions were also reported in oral studies, the most
sensitive (rat) oral study for this effect (and most reliable for dose-response analysis) of Narotsky
et al. (1995) was also selected as a candidate critical study. The confidence in the oral studies
and candidate reference values developed for more sensitive endpoints is more moderate, but still
sufficient for consideration as candidate critical effects. The most sensitive endpoints by far are
the increased fetal heart malformations in rats reported by Johnson et al. (2003) and the
developmental immunotoxicity in mice reported by Peden-Adams et al. (2006), and these are
both considered candidate critical effects. Neurodevelopmental effects are a distinct type among
developmental effects. Thus, the next most sensitive endpoints of decreased rearing post-
exposure in mice (Fredricksson et al. 1993), increased exploration post-exposure in rats (Taylor
6/22/2009
1210
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
et al. 1985) and decreased myelination in the hippocampus of rats (Isaacson and Taylor 1989) are
also considered candidate critical effects.
Because of the general lack of specific data as to the metabolites involved and the MOA
for the candidate critical reproductive effects, PBPK model predictions of total metabolism
(scaled by body weight to the 3/4 power) was selected as the preferred dose metric based on the
general observation that TCE toxicity is associated with metabolism. The AUC of TCE in blood
was used as an alternative dose metric. The only exception to this was for the Johnson et al.
(2003) study, which suggested that oxidative metabolites were involved in the effects reported
based on similar effects being reported from TCA and DCA exposure. Therefore, in this case,
total oxidative metabolism (scaled by body weight to the % power) was considered the preferred
dose metric, while AUC of TCE in blood was considered the alternative metric. With these dose
metrics, the candidate reference values derived using the PBPK model were, with one exception,
only modestly (~3-fold or less) different than those derived on the basis of applied dose. For
resorptions reported by Narotsky et al. (1995), BMD modeling based on internal doses changed
the candidate reference value by 7- to 8-fold larger than the corresponding cRfD based on
applied dose. However, there is substantial uncertainty in the low-dose curvature of the dose-
response curve for modeling both with applied and internal dose, so the BMD remains somewhat
uncertain for this endpoint/study. Finally, for two studies (Isaacson and Taylor, 1989; Peden-
Adams et al., 2006), PBPK modeling of internal doses was not performed due to the inability to
model the complicated exposure pattern (in utero, followed by lactational transfer, followed by
drinking water post-weaning).
6.2.1.3.7 Summary of most sensitive candidate reference values
As shown in Section 5.1.3 and 5.1.5, the most sensitive candidate reference values are for
developmental effects of heart malformations in rats (candidate RfC of 0.0004 ppm and
candidate RfD of 0.0005 mg/kg/d), developmental immunotoxicity in mice exposed pre- and
post-natally (candidate RfD of 0.0004 mg/kg/d), immunological effects in mice (lowest
candidate RfCs of 0.0003-0.003 ppm and lowest candidate RfDs of 0.0005-0.005 mg/kg/d), and
kidney effects in rats and mice (candidate RfCs of 0.0006-0.002 ppm and candidate RfDs of
0.0003-0.001 mg/kg/d). The most sensitive candidate reference values also generally have low
composite uncertainty factors (with the exception of some mouse immunological and kidney
effects), so are expected to be reflective of the most sensitive effects as well. Thus, the most
sensitive candidate references values for multiple effects span about an order of magnitude for
both inhalation (0.0003-0.003 ppm [0.002-0.02 mg/m3]) and oral (0.0004-0.005 mg/kg/d)
exposures. The most sensitive candidate references values for neurological and reproductive
effects are about an order of magnitude higher (lowest candidate RfCs of 0.007-0.02 ppm [0.04-
0.1 mg/m3] and lowest candidate RfDs of 0.009-0.02 mg/kg/d). Lastly, the liver effects have
6/22/2009
1211
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
candidate reference values that are another 2 orders of magnitude higher (candidate RfCs of 1-2
ppm [6-10 mg/m3] and candidate RfDs of 0.9-2 mg/kg/d).
6.2.1.4 Non-cancer reference values (Section 5.1.5)
6.2.1.4.1 Reference Concentration
The goal is to select an overall RfC that is well supported by the available data (i.e.,
without excessive uncertainty given the extensive database) and protective for all the candidate
critical effects, recognizing that individual candidate RfC values are by nature somewhat
imprecise. As discussed in Section 5.1 in Chapter 5, the lowest candidate RfC values within
each health effect category span a 3000-fold range from 0.0003-0.9 ppm (Table 5.1.21). One
approach to selecting a RfC would be to select the lowest calculated value of 0.0003 ppm for
decreased thymus weight in mice. However, six candidate RfCs (cRfCs and p-cRfCs) from both
oral and inhalation studies are in the relatively narrow range of 0.0003-0.003 ppm at the low end
of the overall range (Table 5.1.19). Given the somewhat imprecise nature of the individual
candidate RfC values, and the fact that multiple effects/studies lead to similar candidate RfC
values, the approach taken in this assessment is to select a RfC supported by multiple
effects/studies. The advantages of this approach, which is only possible when there is a
relatively large database of studies/effects and when multiple candidate values happen to fall
within a narrow range at the low end of the overall range, are that it leads to a more robust RfC
(less sensitive to limitations of individual studies) and that it provides the important
characterization that the RfC exposure level is similar for multiple non-cancer effects rather than
being based on a sole explicit critical effect.
Therefore, six critical studies/effects were chosen to support the RfC for TCE non-cancer
effects (see Table 5.1.23). Five of the lowest candidate candidate RfCs, ranging from 0.0003-
0.003 ppm for developmental, kidney, and immunologic effects, are values derived from route-
to-route extrapolation using the PBPK model. The lowest candidate RfC estimate from an
inhalation study is 0.001 ppm for kidney effects. For all six candidate RfCs, the PBPK model
was used for inter- and intra-species extrapolation, based on the preferred dose metric for each
endpoint. There is high confidence in the candidate RfCs for kidney effects for the following
reasons: they are based on clearly adverse effects, two of the values are derived from chronic
studies, and the extrapolation to humans is based on dose metrics clearly related to toxicity
estimated with high confidence with the PBPK model developed in Section 3.5. There is
somewhat less confidence in the lowest candidate RfC for developmental effects (heart
malformations) (see Section 5.1.2.8), and the lowest candidate RfC estimates for immunological
effects (see Section 5.1.2.5). Thus, we do not rely on any single estimate alone; however, each
estimate is supported by estimates of similar magnitude from other effects.
6/22/2009
1212
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
1 As a whole, the estimates support a preferred RfC estimate of 0.001 ppm (1 ppb or 5
2 (J-g/m3). This estimate is within approximately a factor of 3 of the lowest estimates of 0.0003
3 ppm for decreased thymus weight in mice, 0.0004 ppm for heart malformations in rats, 0.0006
4 ppm for toxic nephropathy in rats, 0.001 ppm for increased kidney weight in rats, 0.002 ppm for
5 toxic nephrosis in mice, and 0.003 ppm for increased anti-dsDNA antibodies in mice. Thus,
6 there is robust support for a RfC of 0.001 ppm provided by estimates for multiple effects from
7 multiple studies. The estimates are based on PBPK model-based estimates of internal dose for
8 inter-species, intra-species, and/or route-to-route extrapolation, and there is sufficient confidence
9 in the PBPK model, as well as support from mechanistic data for some of the dose metrics
10 (specifically total oxidative metabolism for the heart malformations and bioactivation of DCVC
11 and total GSH metabolism for toxic nephropathy) (see Section 5.1.3.1). Note that there is some
12 human evidence of developmental heart defects from TCE exposure in community studies (see
13 Section 4.7.3.1.1) and of kidney toxicity in TCE-exposed workers (Section 4.3.1).
14 In summary, the preferred RfC estimate is 0.001 ppm (1 ppb or 5 |ig/m3) based on route-
15 to-route extrapolated results from oral studies for the critical effects of heart malformations
16 (rats), immunotoxicity (mice), and toxic nephropathy (rats, mice), and an inhalation study for the
17 critical effect of increased kidney weight (rats).
18 6.2.1.4.2 Reference Dose
19 As with the RfC determination above, the goal is to select an overall RfD that is well
20 supported by the available data (i.e., without excessive uncertainty given the extensive database)
21 and protective for all the candidate critical effects, recognizing that individual candidate RfD
22 values are by nature somewhat imprecise. As discussed in Section 5.1 in Chapter 5, the lowest
23 candidate RfD values (cRfDs and p-cRfDs) within each health effect category span a nearly
24 3000-fold range from 0.0003-0.8 mg/kg/d (Table 5.1.21). However, four candidate RfDs from
25 oral studies are in the relatively narrow range of 0.0003-0.0005 mg/kg/d at the low end of the the
26 overall range. Given the somewhat imprecise nature of the individual candidate RfD values, and
27 the fact that multiple effects/studies lead to similar candidate RfD values, the approach taken in
28 this assessment is to select a RfD supported by multiple effects/studies. The advantages of this
29 approach, which is only possible when there is a relatively large database of studies/effects and
30 when multiple candidate values happen to fall within a narrow range at the low end of the overall
31 range, are that it leads to a more robust RfD (less sensitive to limitations of individual studies)
32 and that it provides the important characterization that the RfD exposure level is similar for
33 multiple non-cancer effects rather than being based on a sole explicit critical effect.
6/22/2009
1213
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Therefore, four critical studies/effects were chosen to support the RfD for TCE non-
cancer effects (see Table 5.1.24). Three of the lowest candidate RfDs - 0.0003 mg/kg/d for toxic
nephropathy in rats, and 0.0005 mg/kg/d for heart malformations in rats and decreased thymus
weights in mice - are derived using the PBPK model for inter- and intra-species extrapolation,
based on the preferred dose metric for each endpoint. The other of these lowest candidate RfDs
- 0.0004 mg/kg/d for developmental immunotoxicity (decreased PFC response and increased
delayed-type hypersensitivity) in mice - is based on applied dose. There is high confidence in
the candidate RfD for kidney effects(see Section 5.1.2.2), which is based on clearly adverse
effects, derived from a chronic study, and extrapolated to humans based on a dose metric clearly
related to toxicity estimated with high confidence with the PBPK model developed in Section
3.5. There is somewhat less confidence in the candidate RfDs for decreased thymus weights (see
Section 5.1.2.5) and heart malformations and developmental immunological effects (see Section
5.1.2.8). Thus, we do not rely on any single estimate alone; however, each estimate is supported
by estimates of similar magnitude from other effects.
As a whole, the estimates support a preferred RfD of 0.0004 mg/kg/d. This estimate is
within 25% of the lowest estimates of 0.0003 for toxic nephropathy in rats, 0.0004 mg/kg/d for
developmental immunotoxicity (decreased PFC and increased delayed-type hypersensitivity) in
mice, and 0.0005 mg/kg/d for heart malformations in rats and decreased thymus weights in mice.
Thus, there is strong, robust support for a RfD of 0.0004 mg/kg/d provided by the concordance
of estimates derived from multiple effects from multiple studies. The estimates for kidney
effects, thymus effects, and developmental heart malformations are based on PBPK model-based
estimates of internal dose for inter-species and intra-species extrapolation, and there is sufficient
confidence in the PBPK model, as well as support from mechanistic data for some of the dose
metrics (specifically total oxidative metabolism for the heart malformations and bioactivation of
DCVC for toxic nephropathy) (see Section 5.1.3.1). Note that there is some human evidence of
developmental heart defects from TCE exposure in community studies (see Section 4.7.3.1.1)
and of kidney toxicity in TCE-exposed workers (Section 4.3.1).
In summary, the preferred RfD estimate is 0.0004 mg/kg/d based on the critical effects of
heart malformations (rats), adult immunological effects (mice), developmental immunotoxicity
(mice), and toxic nephropathy (rats).
6/22/2009
1214
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
6.2.2 Cancer (Section 5.2)
6.2.2.1 Background and methods (Rodent: Section 5.2.1.1; Human: 5.2.2.1)
As summarized above, following U.S. EPA (2005a) Guidelines for Carcinogen Risk
Assessment, TCE is characterized as carcinogenic in humans by all routes of exposure, based on
convincing evidence of a causal association between TCE exposure in humans and kidney
cancer, but there is also human evidence of TCE carcinogenicity in the liver and lymphoid
tissues. This conclusion is further supported by rodent bioassay data indicating carcinogenicity
of TCE in rats and mice at tumor sites that include those identified in human epidemiologic
studies. Therefore, both human epidemiologic studies as well as rodent bioassays were
considered for deriving PODs for dose-response assessment of cancer endpoints. For PODs
derived from rodent bioassays, default dosimetry procedures were applied to convert applied
rodent doses to human equivalent doses. Essentially, for inhalation exposures, "ppm
equivalence" across species was assumed. For oral doses, 3/4-power body-weight scaling was
used, with a default average human body weight of 70 kg. In addition to applied doses, several
internal dose metrics estimated using a PBPK model for TCE and its metabolites were used in
the dose-response modeling for each tumor type. In general, an attempt was made to use tissue-
specific dose metrics representing particular pathways or metabolites identified from available
data as having a likely role in the induction of a tissue-specific cancer. Where insufficient
information was available to establish particular metabolites or pathways of likely relevance to a
tissue-specific cancer, more general "upstream" metrics had to be used. In addition, the selection
of dose metrics was limited to metrics that could be adequately estimated by the PBPK model.
Regarding low-dose extrapolation, a key consideration in determining what extrapolation
approach to use is the MOA(s). However, MOA data are lacking or limited for each of the
cancer responses associated with TCE exposure, with the exception of the kidney tumors. For
the kidney tumors, the weight of the available evidence supports the conclusion that a mutagenic
MOA is operative; this MOA supports linear low-dose extrapolation. For the other TCE-induced
tumors, the MOA(s) is unknown. When the MOA(s) cannot be clearly defined, EPA generally
uses a linear approach to estimate low-dose risk (U.S. EPA, 2005a), based on the following
general principles:
— A chemical's carcinogenic effects may act additively to ongoing biological processes, given
that diverse human populations are already exposed to other agents and have substantial
background incidences of various cancers.
— A broadening of the dose-response curve (i.e., less rapid fall-off of response with decreasing
dose) in diverse human populations and, accordingly, a greater potential for risks from low-
dose exposures (Ziese et al., 1987; Lutz et al., 2005) is expected for two reasons: First, even
6/22/2009
1215
-------
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
if there is a "threshold" concentration for effects at the cellular level, that threshold is
expected to differ across individuals. Second, greater variability in response to exposures
would be anticipated in heterogeneous populations than in inbred laboratory species under
controlled conditions (due to, e.g., genetic variability, disease status, age, nutrition, and
smoking status).
The general use of linear extrapolation provides reasonable upper-bound estimates that are
believed to be health-protective (U.S. EPA, 2005a) and also provides consistency across
assessments.
6/22/2009
1216
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
6.2.2.2 Inhalation Unit Risk Estimate (Rodent: Section 5.2.1.3; Human: Section 5.2.2.1 and
5.2.2.2)
The inhalation unit risk for TCE is defined as a plausible upper bound lifetime extra risk
of cancer from chronic inhalation of TCE per unit of air concentration. The preferred estimate of
the inhalation unit risk for TCE is 2.20 x 10~2 per ppm (2 x 10~2 per ppm [4 x 10~6 per jig/m3]
rounded to 1 significant figure), based on human kidney cancer risks reported by Charbotel et al.
(2006) and adjusted for potential risk for tumors at multiple sites. This estimate is based on
good-quality human data, thus avoiding the uncertainties inherent in interspecies extrapolation.
The Charbotel et al. (2006) case-control study of 86 incident renal cell carcinoma (RCC) cases
and 316 age- and sex-matched controls, with individual cumulative exposure estimates for TCE
inhalation for each subject, provides a sufficient human dataset for deriving quantitative cancer
risk estimates for RCC in humans. The study is a high-quality study which used a detailed
exposure assessment (Fevotte et al., 2006) and took numerous potential confounding factors,
including exposure to other chemicals, into account. A significant dose-response relationship
was reported for cumulative TCE exposure and RCC (Charbotel et al., 2006). Human data on
TCE exposure and cancer risk sufficient for dose-response modeling are only available for RCC,
yet human and rodent data suggest that TCE exposure increases the risk of cancer at other sites
as well. In particular, there is evidence from human (and rodent) studies for increased risks of
lymphoma and liver cancer. Therefore, the inhalation unit risk estimate derived from human
data for RCC incidence was adjusted to account for potential increased risk of those tumor types.
To make this adjustment, a factor accounting for the relative contributions to the extra risk for
cancer incidence from TCE exposure for these three tumor types combined versus the extra risk
for RCC alone was estimated, and this factor was applied to the unit risk estimate for RCC to
obtain a unit risk estimate for the three tumor types combined (i.e., lifetime extra risk for
developing any of the 3 types of tumor). This estimate is considered a better estimate of total
cancer risk from TCE exposure than the estimate for RCC alone. Although only the Charbotel et
al. (2006) study was found adequate for direct estimation of inhalation unit risks, the available
epidemiologic data provide sufficient information for estimating the relative potency of TCE
across tumor sites. In particular, the relative contributions to extra risk (for cancer incidence)
were calculated from two different datasets to derive the adjustment factor for adjusting the unit
risk estimate for RCC to a unit risk estimate for the 3 types of cancers (RCC, lymphoma, and
liver) combined. The first calculation is based on the results of the meta-analyses of human
epidemiologic data for the 3 tumor types; the second calculation is based on the results of the
6/22/2009
1217
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
Raaschou-Nielsen et al. (2003) study, the largest single human epidemiologic study by far with
RR estimates for all 3 tumor types. Both calculations support a 4-fold adjustment factor.
The preferred estimate of the inhalation unit risk based on human epidemiologic data is
supported by inhalation unit risk estimates from multiple rodent bioassays, the most sensitive of
which range from 1 x 10~2 to 2 x 10-1 per ppm [2 x 10-6 to 3 x 10~5 per jng/m3| From the
inhalation bioassays selected for analysis in section 5.2.1.1, and using the preferred PBPK
model-based dose metrics, the inhalation unit risk estimate for the most sensitive sex/species is 8
x 10~2 per ppm [2 x 10~5 per |ig/m3], based on kidney adenomas and carcinomas reported by
Maltoni et al. (1986) for male Sprague-Dawley rats. Leukemias and Leydig cell tumors were
also increased in these rats, and, although a combined analysis for these tumor types which
incorporated the different site-specific preferred dose metrics was not performed, the result of
such an analysis is expected to be similar, about 9 x 10~2 per ppm [2 x 10~5 per |ig/m3]. The
next most sensitive sex/species from the inhalation bioassays is the female mouse, for which
lymphomas were reported by Henschler et al. (1980); these data yield a unit risk estimate of
1.0 x 10~2 per ppm [2 x 10-6 per |ig/m3]. In addition, the 90% confidence intervals reported in
Table 5.2.10 for male rat kidney tumors from Maltoni et al. (1986) and female mouse
lymphomas from Henschler et al. (1980), derived from the quantitative analysis of PBPK model
uncertainty, both included the estimate based on human data of 2 x 10 2 per ppm. Furthermore,
PBPK model-based route-to-route extrapolation of the results for the most sensitive sex/species
from the oral bioassays, kidney tumors in male Osborne-Mendel rats and testicular tumors in
Marshall rats (NTP 1988), leads to inhalation unit risk estimates of 2 x 10 1 per ppm [3 x 10 5
per |ig/m3] and 4 x 10 2 per ppm [8 x 10 6 per |ig/m3], respectively, with the preferred estimate
based on human data falling within the route-to-route extrapolation of the 90% confidence
intervals reported in Table 5.2.11. Finally, for all these estimates, the ratios of BMDs to the
BMDLs did not exceed a value of 3, indicating that the uncertainties in the dose-response
modeling for determining the POD in the observable range are small.
Although there are uncertainties in these various estimates, confidence in the proposed
inhalation unit risk estimate of 2 x 10 2 per ppm [4 x 10 6 per |ig/m3], based on human kidney
cancer risks reported by Charbotel et al. (2006) and adjusted for potential risk for tumors at
multiple sites (as summarized above in Section 6.1.4), is further increased by the similarity of
this estimate to estimates based on multiple rodent datasets. Application of the ADAF for kidney
cancer risks due to the weight of evidence supporting a mutagenic MOA for this endpoint is
summarized below in Section 6.2.2.5.
6/22/2009
1218
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
6.2.2.3 Oral Unit Risk Estimate (Rodent: Section 5.2.1.3; Human: Section 5.2.2.3)
The oral unit risk (or slope factor) for TCE is defined as a plausible upper bound lifetime
extra risk of cancer from chronic ingestion of TCE per mg/kg/d oral dose. The preferred
estimate of the oral unit risk is 4.63 x 10 2 per mg/kg/d (5 x 10~2 per mg/kg/d rounded to 1
significant figure), resulting from PBPK model-based route-to-route extrapolation of the
inhalation unit risk estimate based on the human kidney cancer risks reported in Charbotel et al.
(2006) and adjusted for potential risk for tumors at multiple sites. This estimate is based on
good-quality human data, thus avoiding uncertainties inherent in interspecies extrapolation. In
addition, uncertainty in the PBPK model-based route-to-route extrapolation is relatively low
(Chiu and White 2006, Chiu 2006). In this particular case, extrapolation using different dose
metrics yielded expected population mean risks within about a 2-fold range, and, for any
particular dose metric, the 95% confidence interval for the extrapolated population mean risks
for each site spanned a range of no more than about 3-fold.
This value is supported by oral unit risk estimates from multiple rodent bioassays, the
most sensitive of which range from 3 x 10~2 to 3 x 10_1 per mg/kg/d. From the oral bioassays
selected for analysis in section 5.2.1.1, and using the preferred PBPK model-based dose metrics,
the oral unit risk estimate for the most sensitive sex/species is 3 x 10 1 per mg/kg/d, based on
kidney tumors in male Osborne-Mendel rats (NTP 1988). The oral unit risk estimate for
testicular tumors in male Marshall rats (NTP 1988) is somewhat lower at 7 x 10 2 per mg/kg/d.
The next most sensitive sex/species result from the oral studies is for male mouse liver tumors
(NCI 1976), with an oral unit risk estimate of 3 x 10 2 per mg/kg/d. In addition, the 90%
confidence intervals reported in Table 5.2.11 for male Osborne-Mendel rat kidney tumors (NTP
1988), male F344 rat kidney tumors (NTP 1990), and male Marshall rat testicular tumors (NTP
1988), derived from the quantitative analysis of PBPK model uncertainty, all included the
estimate based on human data of 5 x 10 2 per mg/kg/d, while the upper 95% confidence bound
for male mouse liver tumors from NCI (1976) was slightly below this value at 4 x 10 2 per
mg/kg/d. Furthermore, PBPK model-based route-to-route extrapolation of the most sensitive
endpoint from the inhalation bioassays, male rat kidney tumors from Maltoni et al. (1986), leads
to an oral unit risk estimate of 1 x 10 1 per mg/kg/d, with the preferred estimate based on human
data falling within the route-to-route extrapolation of the 90% confidence interval reported in
Table 5.2.10. Finally, for all these estimates, the ratios of BMDs to the BMDLs did not exceed a
value of 3, indicating that the uncertainties in the dose-response modeling for determining the
POD in the observable range are small.
Although there are uncertainties in these various estimates, confidence in the proposed
oral unit risk estimate of 5 x 10 2 per mg/kg/d, resulting from PBPK model-based route-to-route
6/22/2009
1219
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
extrapolation of the inhalation unit risk estimate based on the human kidney cancer risks reported
in Charbotel et al. (2006) and adjusted for potential risk for tumors at multiple sites (as
summarized above in Section 6.1.4), is further increased by the similarity of this estimate to
estimates based on multiple rodent datasets. Application of the ADAF for kidney cancer risks
due to the weight of evidence supporting a mutagenic MOA for this endpoint is summarized
below in Section 6.2.2.5.
6.2.2.4 Uncertainties in cancer dose-response assessment
6.2.2.4.1 Uncertainties in estimates based on human epidemiologic data (Section 5.2.2.1.3)
All risk assessments involve uncertainty, as study data are extrapolated to make
inferences about potential effects in humans from environmental exposure. The preferred values
for the unit risk estimates are based on good quality human data, which avoids interspecies
extrapolation, one of the major sources of uncertainty in quantitative cancer risk estimates.
A remaining major uncertainty in the unit risk estimate for RCC incidence derived from
the Charbotel et al. (2006) is the extrapolation from occupational exposures to lower
environmental exposures. There was some evidence of a contribution to increased RCC risk
from peak exposures; however, there remained an apparent dose-response relationship for RCC
risk with increasing cumulative exposure without peaks, and the OR for exposure with peaks
compared to exposure without peaks was not significantly elevated (Charbotel et al., 2006).
Although the actual exposure-response relationship at low exposure levels is unknown, the
conclusion that a mutagenic MOA is operative for TCE-induced kidney tumors supports the
linear low-dose extrapolation that was used (U.S. EPA, 2005a). Additional support for use of
linear extrapolation is discussed above in Section 6.2.2.1.
In addition, because a linear model was used in the observable range of the human data
and the POD was within the low-dose linear range for extra risk as a function of exposure, linear
extrapolation below the LEC01 is virtually a straight continuation of the 95% UCL on the linear
model used above the LEC01. Thus, the use of linear extrapolation from the POD differed
negligibly from extrapolation of the dose-response model itself to low dose.
With respect to uncertainties in the dose-response modeling, the two-step approach of
modeling only in the observable range, as put forth in EPA's cancer assessment guidelines (U.S.
EPA, 2005a), is designed in part to minimize model dependence. The ratio of the EC01 to the
LEC01, which gives some indication of the uncertainties in the dose-response modeling, was
about a factor of 2. Thus, overall, modeling uncertainties in the observable range are considered
to be negligible.
An important source of uncertainty in the underlying Charbotel et al. (2006) study is the
retrospective estimation of TCE exposures in the study subjects. This case-control study was
6/22/2009
1220
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
conducted in the Arve Valley in France, a region with a high concentration of screw cutting
workshops using TCE and other degreasing agents. Since the 1960s, occupational physicians of
the region have collected a large quantity of well-documented measurements, including TCE air
concentrations and urinary metabolite levels (Fevotte et al., 2006). The study investigators
conducted a comprehensive exposure assessment to estimate cumulative TCE exposures for the
individual study subjects, using a detailed occupational questionnaire with a customized task-
exposure matrix for the screw-cutting workers and a more general occupational questionnaire for
workers exposed to TCE in other industries (Fevotte et al., 2006). The exposure assessment also
attempted to take dermal exposure from hand-dipping practices into account by equating it with
an equivalent airborne concentration based on biological monitoring data. Despite the
appreciable effort of the investigators, considerable uncertainty associated with any retrospective
exposure assessment is inevitable, and some exposure misclassification is unavoidable. Such
exposure misclassification was most likely for the 19 deceased cases and their matched controls,
for which proxy respondents were used, and for exposures outside the screw-cutting industry
(295 of 1486 identified job periods involved TCE exposure; 120 of these were not in the screw-
cutting industry).
Another noteworthy source of uncertainty in the Charbotel et al. (2006) study is the
possible influence of potential confounding or modifying factors. This study population, with a
high prevalence of metal-working, also had relatively high prevalences of exposure to petroleum
oils, cadmium, petroleum solvents, welding fumes, and asbestos (Fevotte et al., 2006). Other
exposures assessed included other solvents (including other chlorinated solvents), lead, and
ionizing radiation. None of these exposures was found to be significantly associated with RCC
at a p=0.05 significance level. Cutting fluids and other petroleum oils were associated with RCC
at a p=0.1 significance level; however, further modeling suggested no association with RCC
when other significant factors were taken into account (Charbotel et al., 2006). The medical
questionnaire included familial kidney disease and medical history, such as kidney stones,
infection, chronic dialysis, hypertension, and use of anti-hypertensive drugs, diuretics, and
analgesics. Body mass index (BMI) was also calculated, and lifestyle information such as
smoking habits and coffee consumption was collected. Univariate analyses found high levels of
smoking and BMI to be associated with increased odds of RCC, and these two variables were
included in the conditional logistic regressions. Thus, although impacts of other factors are
possible, this study took great pains to attempt to account for potential confounding or modifying
factors.
Some other sources of uncertainty associated with the epidemiological data are the dose
metric and lag period. As discussed above, there was some evidence of a contribution to
increased RCC risk from peak TCE exposures; however, there appeared to be an independent
6/22/2009
1221
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
effect of cumulative exposure without peaks. Cumulative exposure is considered a good
measure of total exposure because it integrates exposure (levels) over time. If there is a
contributing effect of peak exposures, not already taken into account in the cumulative exposure
metric, the linear slope may be overestimated to some extent. Sometimes cancer data are
modeled with the inclusion of a lag period to discount more recent exposures not likely to have
contributed to the onset of cancer. In an unpublished report (Charbotel et al., 2005), Charbotel et
al. also present the results of a conditional logistic regression with a 10-year lag period, and these
results are very similar to the untagged results reported in their published paper, suggesting that
the lag period might not be an important factor in this study.
Some additional sources of uncertainty are not so much inherent in the exposure-response
modeling or in the epidemiologic data themselves but, rather, arise in the process of obtaining
more general Agency risk estimates from the epidemiologic results. EPA cancer risk estimates
are typically derived to represent an upper bound on increased risk of cancer incidence for all
sites affected by an agent for the general population. From experimental animal studies, this is
accomplished by using tumor incidence data and summing across all the tumor sites that
demonstrate significantly increased incidences, customarily for the most sensitive sex and
species, to attempt to be protective of the general human population. However, in estimating
comparable risks from the Charbotel et al. (2006) epidemiologic data, certain limitations are
encountered. For one thing, these epidemiology data represent a geographically limited (Arve
Valley, France) and likely not very diverse population of working adults. Thus, there is
uncertainty about the applicability of the results to a more diverse general population.
Additionally, the Charbotel et al. (2006) study was a study of RCC only, and so the risk
estimate derived from it does not represent all the tumor sites that may be affected by TCE. This
uncertainty was addressed by adjusting the RCC estimate to multiple sites, but there are also
uncertainties related to the assumptions inherent in the calculations for this adjustment. As
discussed in Section 5.2.2.2, adequate quantitative dose-response data were only available for
one cancer site in humans, so other human data were used to adjust the estimate derived for RCC
to include risk for other cancers with substantial human evidence of hazard (NHL and liver
cancer). The relative contributions to extra risk (for cancer incidence) were calculated from two
different datasets to derive an adjustment factor. The first calculation is based on the results of
the meta-analyses for the 3 tumor types; the second calculation is based on the results of the
Raaschou-Nielsen et al. (2003) study, the largest single study by far with RR estimates for all 3
tumor types. The fact that the calculations based on 2 different datasets yielded comparable
values for the adjustment factor provides more robust support for the use of the factor of 4.
Additional uncertainties pertain to the weight of evidence supporting the association of TCE
exposure with increased risk of cancer for the 3 tumor types. As discussed in Section 4.10.2, we
6/22/2009
1222
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
found that the weight of evidence for kidney cancer was sufficient to classify TCE as
"carcinogenic to humans". We also concluded that there was strong evidence that TCE causes
NHL as well, although the evidence for liver cancer was more limited. In addition, the rodent
studies demonstrate clear evidence of multi-site carcinogenicity, with tumor types including
those for which associations with TCE exposure are observed in human studies, i.e., liver and
kidney cancers and lymphomas. Overall, the evidence is sufficiently persuasive to support the
use of the adjustment factor of 4 based on these 3 tumor types. Alternatively, if one were to use
the factor based only on the 2 tumor types with the strongest evidence, the cancer inhalation unit
risk estimate would be only slightly reduced (25%).
Finally, the preferred value for the oral unit risk estimate was based on route-to-route
extrapolation of the inhalation unit risk based on human data using predictions from the PBPK
model. Because different internal dose metrics are preferred for each target tissue site, a separate
route-to-route extrapolation was performed for each site-specific unit risk estimate. As discussed
above, uncertainty in the PBPK model-based route-to-route extrapolation is relatively low (Chiu
and White 2006, Chiu 2006). In this particular case, extrapolation using different dose metrics
yielded expected population mean risks within about a 2-fold range, and, for any particular dose
metric, the 95% confidence interval for the extrapolated population mean risks for each site
spanned a range of no more than about 3-fold.
6.2.2.4.2 Uncertainties in estimates based on rodent bioassays (Section 5.2.1.4)
With respect to rodent-based cancer risk estimates, the cancer risk is typically estimated
from the total cancer burden from all sites that demonstrate an increased tumor incidence for the
most sensitive experimental species and sex. It is expected that this approach is protective of the
human population, which is more diverse but is exposed to lower exposure levels. In the case of
TCE, the impact of selection of the bioassay is limited, since, as discussed in Sections 5.2.1.3
and 5.2.3, estimates based on the two or three most sensitive bioassays are within an order of
magnitude of each other, and are consistent across routes of exposure when extrapolated using
the PBPK model.
Another source of uncertainty in the TCE rodent-based cancer risk estimates is
interspecies extrapolation. Several plausible PBPK model-based dose metrics were used for
extrapolation of toxicokinetics, but the cancer unit risk estimates obtained using the preferred
dose metrics were generally similar (within about 3-fold) to those derived using default
dosimetry assumptions, with the exception of the bioactivated DCVC dose metric for rat kidney
tumors and the metric for the amount of TCE oxidized in the respiratory tract for mouse lung
tumors occurring from oral exposure. However, there is greater biological support for these
selected dose metrics. The uncertainty in the PBPK model predictions themselves were analyzed
quantitatively through an analysis of the impact of parameter uncertainties in the PBPK model.
6/22/2009
1223
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
The 95% lower bounds on the BMD including parameter uncertainties in the PBPK model were
no more than 4-fold lower than those based on central estimates of the PBPK model predictions.
The greatest uncertainty was for unit risks derived from rat kidney tumors, primarily reflecting
the substantial uncertainty in the rat internal dose.
Regarding low-dose extrapolation, a key consideration in determining what extrapolation
approach to use is the MOA(s). However, MOA data are lacking or limited for each of the
cancer responses associated with TCE exposure, with the exception of the kidney tumors. For
the kidney tumors, the weight of the available evidence supports the conclustion that a mutagenic
MOA is operative; this MOA supports linear low-dose extrapolation. For the other TCE-induced
tumors, the MOA(s) is unknown. When the MOA(s) cannot be clearly defined, EPA generally
uses a linear approach to estimate low-dose risk (U.S. EPA, 2005a), based on the general
principles discussed above.
With respect to uncertainties in the dose-response modeling, the two-step approach of
modeling only in the observable range, as put forth in EPA's cancer assessment guidelines (U.S.
EPA, 2005a), is designed in part to minimize model dependence. The ratios of the BMDs to the
BMDLs, which give some indication of the uncertainties in the dose-response modeling, did not
exceed a value of 2.5 for all the primary analyses used in this assessment. Thus, overall,
modeling uncertainties in the observable range are considered to be negligible. Some additional
uncertainty is conveyed by uncertainties in the survival adjustments made to some of the
bioassay data; however, a comparison of the results of two different survival adjustment methods
suggest that their impact is minimal relative to the uncertainties already discussed.
6.2.2.5 Application of age-dependent adjustment factors (Section 5.2.3.3)
When there is sufficient weight of evidence to conclude that a carcinogen operates
through a mutagenic MOA, and in the absence of chemical-specific data on age-specific
susceptibility, U.S. EPA's Supplemental Guidance for Assessing Susceptibility from Early-Life
Exposure to Carcinogens (U.S. EPA, 2005b) recommends the application of default age-
dependent adjustment factors (ADAFs) to adjust for potential increased susceptibility from early-
life exposure. See the Supplemental Guidance for detailed information on the general
application of these adjustment factors. In brief, the Supplemental Guidance establishes ADAFs
for three specific age groups. The current ADAFs and their age groupings are 10 for <2 years, 3
for 2 to <16 years, and 1 for 16 years and above (U.S. EPA, 2005b). For risk assessments based
on specific exposure assessments, the 10-fold and 3-fold adjustments to the unit risk estimates
are to be combined with age-specific exposure estimates when estimating cancer risks from
early-life (<16 years age) exposure. The ADAFs and their age groups may be revised over time.
6/22/2009
1224
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
The most current information on the application of ADAFs for cancer risk assessment can be
found at www.epa.gov/cancerguidelines.
In the case of TCE, the inhalation and oral unit risk estimates reflect lifetime risk for
cancer at multiple sites, and a mutagenic MOA has been established for one of these sites, the
kidney. As illustrated in the example calculations in Section 5.2.3.3, application of the ADAFs
to the kidney cancer inhalation and oral unit risk estimates for TCE is likely to have minimal
impact on the total cancer risk except when exposure is primarily during early life.
In addition to the uncertainties discussed above for the inhalation and oral total cancer
unit risk estimates, there are uncertainties in the application of ADAFs to adjust for potential
increased early-life susceptibility. The adjustment is made only for the kidney cancer component
of total cancer risk because that is the tumor type for which the weight of evidence was sufficient
to conclude that TCE-induced carcinogenesis operates through a mutagenic MOA. However, it
may be that TCE operates through a mutagenic MOA for other tumor types as well or that it
operates through other MO As that might also convey increased early-life susceptibility.
Additionally, the ADAFs are general default factors, and it is uncertain to what extent they
reflect increased early-life susceptibility for exposure to TCE, if increased early-life
susceptibility occurs.
6/22/2009
1225
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
6.3 Overall Characterization of TCE Hazard and Dose-response
There is substantial potential for human exposure to TCE, as it has a widespread presence
in ambient air, indoor air, soil, and groundwater. At the same time, humans are likely to be
exposed to a variety of compounds that are either metabolites of TCE or which have common
metabolites or targets of toxicity. Once exposed, humans, as well as laboratory animal species,
rapidly absorb TCE, which is then distributed to tissues via systemic circulation, extensively
metabolized, and then excreted primarily in breath as unchanged TCE or C02, or in urine as
metabolites.
Based on the available human epidemiologic data and experimental and mechanistic
studies, it is concluded that TCE poses a potential human health hazard for non-cancer toxicity to
the central nervous system, the kidney, the liver, the immune system, the male reproductive
system, and the developing fetus. The evidence is more limited for TCE toxicity to the
respiratory tract and female reproductive system. Following U.S. EPA (2005a) Guidelines for
Carcinogen Risk Assessment, TCE is characterized as carcinogenic in humans by all routes of
exposure. This conclusion is based on convincing evidence of a causal association between TCE
exposure in humans and kidney cancer. The human evidence of carcinogenicity from
epidemiologic studies of TCE exposure is compelling for Non-Hodgkins Lymphoma (NHL) but
less convincing than for kidney cancer, and more limited for liver and biliary tract cancer.
Further support for the characterization of TCE as carcinogenic in humans by all routes of
exposure is derived from positive results in multiple rodent cancer bioassays in rats and mice of
both sexes, similar toxicokinetics between rodents and humans, mechanistic data supporting a
mutagenic MOA for kidney tumors, and the lack of mechanistic data supporting the conclusion
that any of the MOA(s) for TCE-induced rodent tumors are irrelevant to humans.
As TCE toxicity and carcinogenicity are generally associated with TCE metabolism,
susceptibility to TCE health effects may be modulated by factors affecting toxicokinetics,
including lifestage, gender, genetic polymorphisms, race/ethnicity, pre-existing health status,
lifestyle, and nutrition status. In addition, while these some of these factors are known risk
factors for effects associated with TCE exposure, it is not known how TCE interacts with known
risk factors for human diseases.
For non-cancer effects, the most sensitive types of effects, based either on human
equivalent concentrations/doses or on candidate RfCs/RfDs, appear to be developmental, kidney,
and immunological (adult and developmental) effects. The neurological and reproductive effects
appear to be about an order of magnitude less sensitive, with liver effects another two orders of
magnitude less sensitive. The preferred RfC estimate of 0.001 ppm (1 ppb or 5 (^g/m3) is based
on route-to-route extrapolated results from oral studies for the critical effects of heart
6/22/2009
1226
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
INTER-AGENCY REVIEW DRAFT - DO NOT QUOTE OR CITE
malformations (rats), immunotoxicity (mice), and toxic nephropathy (rats, mice), and an
inhalation study for the critical effect of increased kidney weight (rats). Similarly, the preferred
RfD estimate for non-cancer effects of 0.0004 mg/kg/d is based on the critical effects of heart
malformations (rats), adult immunological effects (mice), developmental immunotoxicity (mice),
and toxic nephropathy (rats). There is high confidence in these preferred non-cancer reference
values, as they are supported by moderate- to high-confidence estimates for multiple effects from
multiple studies.
For cancer, the preferred estimate of the inhalation unit risk is 2 x 10~2 per ppm [4 x 10~6
per jug/m31, based on human kidney cancer risks reported by Charbotel et al. (2006) and
adjusted, using human epidemiologic data, for potential risk for tumors at multiple sites. The
preferred estimate of the oral unit risk for cancer is 5 x 10~2 per mg/kg/d, resulting from PBPK
model-based route-to-route extrapolation of the inhalation unit risk estimate based on the human
kidney cancer risks reported in Charbotel et al. (2006) and adjusted, using human epidemiologic
data, for potential risk for tumors at multiple sites. There is high confidence in these unit risks
for cancer, as they are based on good quality human data, as well as being similar to unit risk
estimates based on multiple rodent bioassays. Because there is both sufficient weight of
evidence to conclude that TCE operates through a mutagenic MOA for kidney tumors and a lack
of TCE-specific quantitative data on early-life susceptibility, the default age-dependent
adjustment factors (ADAFs) can be applied for the kidney cancer component of the unit risks for
cancer; however, the application of ADAFs is likely to have a minimal impact on the total cancer
risk except when exposures are primarily during early life.
6/22/2009
1227
------- |