DRAFT - DO NOT CITE OR QUOTE
^CDA
mI
EPA/63 5/R-08/01 ID
June 2011
TOXICOLOGICAL REVIEW
OF
T etrachloroethylene
(Perchloroethylene)
(CAS No. 127-18-4)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
June 2011
NOTICE
This document is a Final Agency/Interagency Review Draft. This information is distributed
solely for the purpose of pre-dissemination peer review under applicable information quality
guidelines. It has not been formally disseminated by EPA. It does not represent and should not
be construed to represent any Agency determination or policy. It is being circulated for review
of its technical accuracy and science policy implications.
U.S. Environmental Protection Agency
Washington, DC
-------
DISCLAIMER
This document is a preliminary draft for review purposes only. This information is
distributed solely for the purpose of pre-dissemination peer review under applicable information
quality guidelines. It has not been formally disseminated by EPA. It does not represent and
should not be construed to represent any Agency determination or policy. Mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 ii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS
TOXICOLOGICAL REVIEW for TETRACHLOROETHYLENE
(PERCHLOROETHYLENE) (CAS No. 127-18-4)
CONTENTS iii
LIST OF TABLES xii
LIST OF FIGURES xv
LIST OF ABBREVIATIONS AND ACRONYMS xviii
FOREWORD xxi
1. INTRODUCTION 1-1
2. BACKGROUND 2-1
2.1. USES AND PHYSICAL/CHEMICAL PROPERTIES 2-1
2.2. OCCURRENCE AND EXPOSURE 2-1
2.2.1. Air 2-1
2.2.2. Water 2-4
2.2.3. Food 2-5
2.2.4. Soil 2-6
2.2.5. Breast Milk 2-6
2.2.6. Direct Ingestion 2-7
3. TOXICOKINETICS 3-1
3.1. ABSORPTION 3-1
3.1.1. Inhalation 3-1
3.1.2. Oral 3-2
3.1.3. Dermal 3-2
3.2. DISTRIBUTION AND BODY BURDEN 3-2
3.3. METABOLISM 3-4
3.3.1. Introduction 3-4
3.3.2. Extent of Metabolism 3-4
3.3.3. Pathways of Metabolism 3-7
3.3.3.1. Cytochrome P450-Dependent Oxidation 3-7
3.3.3.2. Glutathione (GSH) Conjugation Pathway 3-11
3.3.3.3. Relative Roles of the Cytochrome P450 (CYP) and
Glutathione (GSH) Pathways 3-16
3.3.4. Susceptibility 3-16
3.3.5. Comparison of Tetrachloroethylene Metabolism with
Trichloroethylene Metabolism 3-18
6/24/11
This document is a draft for review purposes only and does not constitute Agency policy
iii DRAFT—DO NOT CITE OR QUOTE
-------
3.3.5.1. Extent of Metabolism 3-18
3.3.5.2. Cytochrome P450 (CYP)-Mediated Oxidation 3-19
3.3.5.3. Glutathione (GSH) Conjugation Pathway 3-20
3.3.5.4. Summary 3-20
3.4. EXCRETION 3-21
3 .5. TOXICOKINETIC MODELING 3-23
3.5.1. Choice of Physiologically Based Pharmacokinetic (PBPK) Model for
Use in Dose-Response Modeling 3-23
3.5.1.1. Limitations of Previously Developed Physiologically Based
Pharmacokinetic (PBPK) Models 3-23
3.5.1.2. The Chiu and Ginsberg (In Press) Model 3-30
3.5.2. Age and Gender-Related Differences in Tetrachloroethylene
Pharmacokinetics 3-48
3.5.3. Metabolic Interactions with Other Chemicals 3-51
4. HAZARD IDENTIFICATION 4-1
4.1. NEUROTOXICITY 4-1
4.1.1. Human Studies 4-1
4.1.1.1. Chamber Studies 4-2
4.1.1.2. Chronic Exposure Studies 4-5
4.1.1.3. Summary of Neuropsychological Effects in Low- and
Moderate-Exposure Studies 4-33
4.1.2. Animal Studies 4-39
4.1.2.1. Inhalation Studies 4-40
4.1.2.2. Oral and Intraperitoneal Studies 4-47
4.1.3. Mode of Action for Neurotoxic Effects 4-51
4.1.3.1. Visual Function 4-52
4.1.3.2. Cognition 4-54
4.1.3.3. Reaction Time 4-54
4.1.4. Summary of Neurotoxic Effects in Humans and Animals 4-55
4.2. KIDNEY AND BLADDER TOXICITY AND CANCER 4-58
4.2.1. Human Studies 4-58
4.2.1.1. Kidney Toxicity in Humans 4-58
4.2.1.2. Kidney Cancer in Humans 4-63
4.2.1.3. Bladder Cancer in Humans 4-80
4.2.2. Animal Studies 4-99
4.2.2.1. Kidney Toxicity in Animals 4-99
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 iv DRAFT—DO NOT CITE OR QUOTE
-------
4.2.2.2. Kidney Cancer in Animals 4-108
4.2.3. Summary of Kidney Effects in Humans and Animals 4-112
4.2.4. Hypothesized Mode(s) of Action for Kidney Carcinogenicity 4-113
4.2.4.1. Role of Metabolism in Kidney Carcinogenicity 4-114
4.2.4.2. a2[j,-Globulin Accumulation 4-114
4.2.4.3. Genotoxicity 4-118
4.2.4.4. Peroxisome Proliferation 4-119
4.2.4.5. Cytotoxicity/Sustained Chronic Nephrotoxicity Not
Associated with a2[j,-Globulin Nephropathy 4-121
4.2.4.6. Summary 4-122
4.3. LIVER TOXICITY AND CANCER 4-123
4.3.1. Human Studies 4-123
4.3.1.1. Liver Damage 4-124
4.3.1.2. Liver Cancer 4-127
4.3.2. Animal Studies 4-139
4.3.2.1. Liver Toxicity 4-139
4.3.2.2. Liver Cancer 4-148
4.3.3. Summary of Liver Effects in Humans and Animals 4-149
4.3.4. Mode of Action for Hemangiosarcomas or Hemangiomas in Mice 4-151
4.3.5. Mode of Action for Murine Hepatocellular Tumors 4-151
4.3.5.1. Contribution of Tetrachloroethylene Metabolites to Mode of
Action and Carcinogenicity 4-152
4.3.5.2. Genotoxicity 4-156
4.3.5.3. AlteredDNAMethylation 4-157
4.3.5.4. Cytotoxicity and Secondary Oxidative Stress 4-158
4.3.5.5. Peroxisome Proliferator-Activated Receptor (PPAR)
Activation Mode of Action 4-160
4.3.5.6. Mode of Action Conclusions for Hepatocellular Tumors 4-178
4.4. ESOPHAGEAL CANCER 4-181
4.4.1. Consideration of Exposure-Assessment Methodology 4-181
4.4.2. Summary of Results 4-183
4.5. LUNG AND RESPIRATORY CANCER 4-191
4.5.1. Consideration of Exposure-Assessment Methodology 4-192
4.5.2. Summary of Results 4-193
4.6. IMMUNOTOXICITY, HEMATOLOGIC TOXICITY, AND CANCERS OF
THE IMMUNE SYSTEM 4-203
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 v DRAFT—DO NOT CITE OR QUOTE
-------
4.6.1. Human Studies 4-203
4.6.1.1. Noncancer Immune and Hematologic Effects 4-203
4.6.1.2. Cancers of the Immune System, Including Childhood
Leukemia 4-212
4.6.2. Animal Studies 4-248
4.6.2.1. Noncancer Effects 4-248
4.6.2.2. Cancer Effects 4-251
4.6.3. Summary and Conclusions 4-262
4.6.3.1. Immunotoxicity, Hematologic Toxicity, and Cancers of the
Immune System in Humans 4-262
4.6.3.2. Immunol ogi cal and Hematol ogi cal Toxi city and
Mononuclear Cell Leukemias in Rodents 4-264
4.7. DEVELOPMENTAL AND REPRODUCTIVE TOXICITY AND
REPRODUCTIVE CANCERS 4-268
4.7.1. Development 4-268
4.7.1.1. Human Developmental Toxicity Data 4-268
4.7.1.2. Animal Developmental Toxicity Studies 4-301
4.7.2. Reproduction 4-306
4.7.2.1. Human Reproduction Data 4-306
4.7.2.2. Animal Reproductive Toxicity Studies 4-326
4.7.2.3. Reproductive Cancers in Humans 4-329
4.7.3. Summary of Human and Animal Developmental/Reproductive Studies ..4-345
4.7.3.1. Summary of Human Data 4-345
4.7.3.2. Summary of Animal Data 4-348
4.7.4. Mode of Action for Developmental Effects 4-352
4.8. GENOTOXICITY 4-353
4.8.1. Tetrachloroethylene (PCE) 4-354
4.8.1.1. Mammalian Systems (Including Human Studies) 4-354
4.8.1.2. Drosophila melanogaster 4-367
4.8.1.3. Bacterial and Fungal Systems 4-368
4.8.1.4. Summary 4-369
4.8.2. Trichloroacetic Acid (TCA) 4-371
4.8.2.1. Mammalian Systems (Including Human Studies) 4-371
4.8.2.2. Bacterial Systems 4-376
4.8.2.3. Summary 4-377
4.8.3. Dichloroacetic Acid (DCA) 4-378
6/24/11
This document is a draft for review purposes only and does not constitute Agency policy
vi DRAFT—DO NOT CITE OR QUOTE
-------
4.8.3.1. Mammalian Systems 4-378
4.8.3.2. Bacterial Systems 4-382
4.8.3.3. Summary 4-383
4.8.4. Chloral Hydrate 4-384
4.8.4.1. Mammalian Systems (Including Human Studies) 4-384
4.8.4.2. Bacterial and Fungal Systems 4-393
4.8.4.3. Summary 4-394
4.8.5. Trichloroacetyl Chloride 4-394
4.8.5.1. Bacterial Systems 4-395
4.8.6. Tetrachloroethylene (PCE) Epoxide 4-395
4.8.6.1. Bacterial Systems 4-395
4.8.7. Trichloroethanol (TCOH) 4-395
4.8.7.1. Bacterial Systems 4-395
4.8.8. »Y-(1,2,2-Trichlorovinyl)-A-Cysteine (1,2-TCVC), »Y-Trichlorovinyl
Glutathione (TCVG), A-AcetykY-( l ,2,2-Trichlorovinyl)-A-Cysteine
(NAcTCVC) 4-396
4.8.8.1. Bacterial Systems 4-396
4.8.8.2. Mammalian Systems 4-396
4.8.9. TCVC Sulfoxide 4-399
4.8.10. Synthesis and Overall Summary 4-399
4.9. SUSCEPTIBLE POPULATIONS 4-403
4.9.1. Life-Stages 4-403
4.9.1.1. Early Life-Stages 4-403
4.9.1.2. Later Life-Stages 4-421
4.9.2. Other Susceptibility Factors 4-422
4.9.2.1. Gender 4-423
4.9.2.2. Race/Ethnicity 4-424
4.9.2.3. Genetics 4-425
4.9.2.4. Preexisting Disease 4-426
4.9.2.5. Lifestyle Factors and Nutrition Status 4-426
4.9.2.6. Socioeconomic Status 4-427
4.9.2.7. Multiple Exposures and Cumulative Risks 4-428
4.9.3. Uncertainty of Database and Research Needs for Susceptible
Populations 4-429
4.10. SUMMARY OF HAZARD IDENTIFICATION 4-430
4.10.1. Overview of Noncancer and Cancer Hazard 4-430
6/24/11
This document is a draft for review purposes only and does not constitute Agency policy
vii DRAFT—DO NOT CITE OR QUOTE
-------
4.10.2. Characterization of Noncancer Effects 4-431
4.10.2.1. Neurotoxicity 4-431
4.10.2.2. Kidney Toxicity 4-439
4.10.2.3. Liver Toxicity 4-439
4.10.2.4. Immunotoxicity and hematologic toxicity 4-440
4.10.2.5. Reproductive and Developmental Toxicity 4-441
4.10.2.6. Summary of Noncancer Toxicities and Identification of
Studies for Dose-Response Analyses 4-445
4.10.3. Characterization of Cancer Hazard 4-447
4.10.4. Synthesis of Epidemiologic Studies 4-448
4.10.5. Synthesis of Rodent Cancer Bioassay Findings 4-453
4.10.5.1. Carcinogenicity Findings in Rats 4-454
4.10.5.2. Carcinogenicity Findings in Mice 4-459
4.10.5.3. Carcinogenic Mode of Action Hypotheses 4-460
5. DOSE-RESPONSE EVALUATION 5-1
5.1. INHALATION REFERENCE CONCENTRATION (RfC) 5-1
5.1.1. Choice of Candidate Studies and Critical Effect 5-1
5.1.1.1. Choice of Critical Effect 5-1
5.1.1.2. Overview of Candidate Principal Studies 5-1
5.1.1.3. Selection of Principal Studies 5-2
5.1.2. Additional Analyses: Feasibility of Dose-Response Modeling 5-15
5.1.3. Reference Concentration (RfC) Derivation, Including Application of
Uncertainty Factors 5-16
5.1.4. Dose-Response Analyses for Comparison of Noncancer Effects Other
Than Critical Effects in Neurotoxicity 5-18
5.1.4.1. Sample Reference Concentrations (RfCs) for Kidney
Toxicity 5-20
5.1.4.2. Sample Reference Concentrations (RfCs) for Liver Toxicity 5-20
5.1.4.3. Sample Reference Concentrations (RfCs) for Immunotoxicity
and Hematologic Toxicity 5-21
5.1.4.4. Sample Reference Concentrations (RfCs) for Reproductive
and Developmental Toxicity 5-23
5.1.4.5. Summary of Sample Reference Concentrations (RfCs) for
Noncancer Endpoints Other Than the Critical Effect 5-23
5.1.5. Previous Inhalation Assessment 5-26
5.2. ORAL REFERENCE DOSE (RfD) 5-26
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 viii DRAFT—DO NOT CITE OR QUOTE
-------
5.2.1. Choice of Principal Study and Critical Effects 5-26
5.2.2. Additional Analyses: Route-to-Route Extrapolation Using PBPK
Modeling 5-27
5.2.3. Reference Dose (RfD) Derivation, Including Application of
Uncertainty Factors 5-28
5.2.4. Dose-response Analyses for Noncancer Effects Other Than Critical
Effect of Neurotoxicity 5-30
5.2.4.1. Sample Reference Doses (RfDs) for Kidney Toxicity 5-32
5.2.4.2. Sample Reference Doses (RfDs) for Liver Toxicity 5-33
5.2.4.3. Sample Reference Doses (RfDs) for Immunotoxicity and
Hematologic Toxicity 5-36
5.2.4.4. Sample Reference Doses (RfDs) for Reproductive and
Developmental Toxicity 5-36
5.2.4.5. Summary of Sample Reference Doses (RfDs) for Noncancer
Endpoints Other Than the Critical Effect 5-36
5.2.5. Previous Oral Assessment 5-37
5.3. UNCERTAINTIES IN INHALATION REFERENCE CONCENTRATION
(RfC) AND ORAL REFERENCE DOSE (RfD) 5-39
5.3.1. Point of Departure 5-40
5.3.2. Extrapolation from Laboratory Animal Studies to Humans 5-40
5.3.3. Human Variation 5-41
5.3.4. Database Uncertainties 5-42
5.4. CANCER DOSE-RESPONSE ASSESSMENT 5-43
5.4.1. Choice of Study /Data with Rationale and Justification 5-43
5.4.2. Dose-Response Data 5-45
5.4.2.1. Liver Tumors in Mice 5-45
5.4.2.2. Other Tumor Sites in Male Mice 5-48
5.4.2.3. Mononuclear Cell Leukemia in Rats 5-49
5.4.2.4. Other Tumor Sites in Rats 5-52
5.4.3. Dose Adjustments and Extrapolation Methods 5-55
5.4.3.1. Estimation of Dose Metrics for Dose-Response Modeling 5-55
5.4.3.2. Extrapolation Methods 5-60
5.4.4. Cancer Risk Values 5-67
5.4.4.1. Dose-Response Modeling Results 5-68
5.4.4.2. Choice of Data Set and Associated Uncertainties 5-94
5.4.4.3. Recommended Inhalation Unit Risk 5-97
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 ix DRAFT—DO NOT CITE OR QUOTE
-------
5.4.4.4. Recommended Oral Slope Factor 5-97
5.4.4.5. Uncertainties in Human Population Variability and
Quantitative Adjustment for Sensitive Populations (Age-
Dependent Ajustment Factors) 5-102
5.4.4.6. Concordance of Animal and Human Risk Estimates 5-102
5.4.5. Summary of Uncertainties in Cancer Risk Values 5-104
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND
DOSE-RESPONSE 6-1
6.1. HUMAN HAZARD POTENTIAL 6-1
6.1.1. Exposure (see Section 2) 6-1
6.1.2. Toxicokinetics and Physiologically Based Pharmacokinetic (PBPK)
Modeling (see Section 3) 6-2
6.1.3. Noncancer Toxicity (see Section 4.10.1) 6-4
6.1.4. Neurological Effects (see Section 4.1) 6-5
6.1.5. Summary of Other Noncancer Adverse Effects (see Sections 4.2, 4.3,
4.6, and 4.7) 6-6
6.1.5.1. Kidney Toxicity (see Section 4.2) 6-6
6.1.5.2. Liver Toxicity (see Section 4.3) 6-7
6.1.5.3. Immunologic and Hematopoetic Toxicity (see Section 4.6) 6-7
6.1.5.4. Reproductive Toxicity (see Section 4.7) 6-8
6.1.5.5. Developmental Toxicity (see Section 4.7) 6-9
6.1.6. Carcinogenicity (see Section 4.10.2) 6-10
6.1.7. Susceptibility (see Section 4.9) 6-11
6.2. DOSE-RESPONSE ASSESSMENT 6-12
6.2.1. Noncancer Effects (see Section 5.1) 6-13
6.2.2. Selection of Critical Effect and Principal Studies (see Section 5.1.1) 6-13
6.2.3. Uncertainties and Application of Uncertainty Factors (UFs) (see
Sections 5.1.3, 5.2.3, and 5.3) 6-14
6.2.3.1. Human Variation 6-14
6.2.3.2. LOAEL-to-NOAEL Uncertainty 6-14
6.2.3.3. Database Uncertainty 6-15
6.2.4. Reference Concentration (see Section 5.1.3) 6-15
6.2.5. Reference Dose (see Section 5.2) 6-16
6.2.6. Dose-Response Analyses for Noncancer Effects Other Than Critical
Effect of Neurotoxicity (see Sections 5.1.4 and 5.2.4) 6-16
6.2.7. Cancer (see Section 5.2) 6-17
6/24/11
This document is a draft for review purposes only and does not constitute Agency policy
x DRAFT—DO NOT CITE OR QUOTE
-------
6.2.8. Choice of Data Set for Use in Cancer Risk Estimation 6-18
6.2.9. Inhalation Unit Risk Estimate (see Section 5.4.4.3) 6-20
6.2.10. Oral Slope Factor Estimate (see Section 5.4.4.4) 6-21
6.2.11. Uncertainties in Cancer Dose-Response Assessment 6-21
6.3. OVERALL CHARACTERIZATION OF TETRACHLOROETHYLENE
HAZARD AND DOSE-RESPONSE 6-22
7. References 7-1
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES
Table 2-1. Physical and chemical properties of PCE 2-2
Table 3-1. Log-likelihood and parameters after calibration 3-34
Table 3-2. Predictions for area-under-the-curve of tetrachloroethylene in blood (mg-hr/L-day per
ppm in air or mg-hr/L-day per mg/kg-day oral intake) using posterior mode parameters 3-38
Table 3-3. Predictions for fraction of tetrachloroethylene in oxidized by cytochrome P450
(P450s) (mg/kg-day oxidized per mg/kg-day intake) using posterior mode parameters 3-40
Table 3-4. Predictions for fraction of tetrachloroethylene in conjugated with glutathione (GSH)
(mg/kg-day conjugated per mg/kg-day intake) using posterior mode parameters 3-42
Table 3-5. Predictions for Trichloroacetic acid (TCA) produced systemically (mg/kg-day
systemic TCA per ppm in air or mg/kg-day systemic TCA per mg/kg-day oral intake) using
posterior mode parameters 3-44
Table 3-6. Summary evaluation of the reliability of tetrachloroethylene dose metrics 3-46
Table 3-7. Ratio of average daily dose at various life-stages to the average daily dose for a
25-year-old adult: physiologically based pharmacokinetic (PBPK) simulations 3-50
Table 4-1. Summary of human neurotoxicity studies of occupational or residential exposures to
dry-cleaning facilities using tetrachloroethylene 4-6
Table 4-2. Summary of effects of chronic tetrachloroethylene exposure in humans seen in
studies of neuropsychological function" 4-34
Table 4-3. Summary of animal inhalation neurotoxicology studies 4-41
Table 4-4. Summary of oral neurotoxicity animal studies 4-48
Table 4-5. Summary of in vitro ion channel effects with tetrachloroethylene and other
chlorinated solvents 4-52
Table 4-6. Summary of human kidney toxicity marker studies of occupational exposures to dry-
cleaning facilities using tetrachloroethylene 4-60
Table 4-7. Summary of human studies on tetrachloroethylene exposure and kidney cancer
4-69
Table 4-8. Summary of human studies on tetrachloroethylene exposure and bladder
cancer 4-85
Table 4-9. Summary of rodent kidney toxicity studies 4-100
Table 4-10. Kidney tumor incidence in laboratory animals exposed to tetrachloroethylene
4-102
Table 4-11. Renal a2[j,-globulin formation in tetrachloroethylene-exposed rodents 4-116
Table 4-12. Renal peroxisome proliferation in tetrachloroethylene-exposed rodents 4-120
Table 4-13. Summary of studies of human liver toxicity 4-126
Table 4-14. Summary of human studies on tetrachloroethylene exposure and liver cancer ..4-131
Table 4-15. Summary of inhalation and oral rodent liver toxicity studies 4-140
Table 4-16. Incidence of hepatic tumors in rodents exposed to tetrachloroethylene 4-142
Table 4-17. Hepatocarcinogenicity of TCA in rodent drinking water studies 4-154
Table 4-18. Hepatocarcinogenicity of DC A in rodent drinking water studies 4-154
Table 4-19. Incidence of mouse liver tumors with drinking water administration of TCA
and DCA, alone and in combination 4-155
Table 4-20. Rodent studies of induction of peroxisome proliferation or its markers by
tetrachloroethylene 4-162
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xii DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-21. Potency indicators for mouse hepatocarcinogenicity and in vitro transactivation of
mouse PPARa for four PPARa agonists21 4-171
Table 4-22. Potency indicators for rat hepatocarcinogenicity and common short-term
markers of PPARa-agonism for four PPARa agonists3 4-172
Table 4-23. Summary of human studies on tetrachloroethylene exposure and esophageal
cancer 4-185
Table 4-24. Summary of human studies on tetrachloroethylene exposure and lung cancer... 4-195
Table 4-25. Immune and hematological parameters in studies of dry-cleaning workers or
tetrachloroethylene exposure in children 4-204
Table 4-26. Immune-related conditions in studies of dry cleaning or tetrachloroethylene
exposure in humansa 4-211
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and
hematopoietic cancers, including leukemia 4-215
Table 4-28. Summary of epidemiologic studies on tetrachloroethylene exposure and
childhood hematopoietic cancers, including leukemia 4-230
Table 4-29. Results of epidemiologic studies of potential tetrachloroethylene exposure and adult
lymphopoietic cancer and leukemia, by cancer type and study design 4-236
Table 4-30. Results of epidemiologic studies of potential tetrachloroethylene exposure and adult
non-Hodgkin lymphoma, by study design 4-238
Table 4-31. Results of epidemiologic studies of potential tetrachloroethylene exposure and adult
Hodgkin lymphoma and multiple myeloma, by study design 4-242
Table 4-32. Results of epidemiologic studies of potential tetrachloroethylene exposure and adult
lymphopoeitic cancers, with data pertaining to exposure-response gradients, by cancer type 4-245
Table 4-33. Mononuclear cell leukemia incidence in rats exposed to tetrachloroethylene.... 4-252
Table 4-34. Epidemiology studies on reproduction and development 4-271
Table 4-35. Exposure concentrations (ppm) at which effects occurred in a two-generation study
4-327
Table 4-36. Summary of human studies on tetrachloroethylene exposure and breast cancer 4-333
Table 4-37. Summary of human studies on tetrachloroethylene exposure and cervical
cancer 4-339
Table 4-38. Summary of mammalian developmental and reproductive toxicity studies for
tetrachloroethylene 4-349
Table 4-39. Genotoxicity of tetrachloroethylene—mammalian systems (in vitro and in vivo)a.. 4-
356
Table 4-40. Genotoxicity of tetrachloroethylene—bacterial, yeast, and fungal systems3 . . . 4-
359
Table 4-41. Genotoxicity of trichloroacetic acid (TCA)—mammalian systems (in vitro and in
vivo)a 4-372
Table 4-42. Genotoxicity of trichloroacetic acid (TCA)—bacterial systems3 4-374
Table 4-43. Genotoxicity of dichloroacetic acid (DCA)—mammalian systems (in vitro and in
vivo)a 4-378
Table 4-44. Genotoxicity of dichloroacetic acid (DCA)—bacterial systems3 4-381
Table 4-45. Genotoxicity of chloral hydrate—mammalian systems (in vitro)a 4-384
Table 4-46. Genotoxicity of chloral hydrate—mammalian systems (in vivo)3 4-387
Table 4-47. Genotoxicity of chloral hydrate—bacterial, yeast, and fungal systems3 4-389
Table 4-48. Genotoxicity of additional tetrachloroethylene metabolites—all systems 4-398
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xiii DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-49. Inhalation studies suitable for dose-response analyses 4-432
Table 4-50. NOAELs and LOAELs in selected studies involving oral exposure to
tetrachloroethylene 4-435
Table 4-51. Tumor incidence in rats exposed to tetrachloroethylene 4-456
Table 4-52. Tumor incidence in mice exposed to tetrachloroethylene 4-457
Table 4-53. Renal a2[j,-globulin accumulation in tetrachloroethylene-exposed rodents 4-463
Table 4-54. Renal peroxisome proliferation in tetrachloroethylene-exposed rodents 4-463
Table 4-55. Rodent studies of induction of hepatic peroxisome proliferation or its markers
by tetrachloroethylene 4-465
Table 5-1. Neurotoxicological inhalation studies considered in the development of an RfC .... 5-3
Table 5-2. Summary of rationale for identifying studies on tetrachloroethylene for RfC
development 5-7
Table 5-3. Application of uncertainty factors to four neurological endpoints from three studies
used to derive the RfC 5-16
Table 5-4. Sample RfCs for kidney effects 5-22
Table 5-5. Sample RfCs for liver effects 5-22
Table 5-6. Sample RfCs for immunological and hematological effects 5-24
Table 5-7. Sample RfCs for reproductive and developmental effects 5-24
Table 5-8. Application of uncertainty factors to neurological endpoints from three studies used
to derive the RfD 5-29
Table 5-9. Sample RfDs for kidney effects 5-34
Table 5-10. Sample RfDs for liver effects 5-34
Table 5-11. Sample RfDs for immunological and hematological effects 5-35
Table 5-12. Sample RfDs for reproductive and developmental effects 5-35
Table 5-13. Tumor incidence in mice exposed to tetrachloroethylene 5-46
Table 5-14. Historical control data of the Japan Bioassay Research Center, Crj/BDFl mouse,
104-week studies 5-47
Table 5-15. Incidence of mononuclear cell leukemia, kidney tumors, and brain gliomas in rats
exposed to tetrachloroethylene by inhalation 5-49
Table 5-16. Historical control data of the Japan Bioassay Research Center, F344/DuCrj
(Fischer) rat, 104-week studies 5-51
Table 5-17. Summary of PBPK-derived dose metric estimates used for dose-response analysis of
rodent tumor data 5-57
Table 5-18. Human equivalent unit risks, derived using PBPK-derived dose metrics and
multistage model; tumor incidence data from JISA (1993) and NTP (1986b) 5-73
Table 5-19. Dose-response summary and unit risk estimates using continuous equivalent
administered tetrachloroethylene levels as dose metric, from NTP (1986b) and JISA (1993) . 5-75
Table 5-20. Range of outputs from fitting different BMDS models using continuous equivalent
administered tetrachloroethylene levels as dose metric, from JISA (1973)a 5-77
Table 5-21. Human equivalent oral slope factors, derived using primary dose metrics and
multistage model; tumor incidence data from JISA (1993) and NTP (1986b) 5-100
Table 5-22. Summary of uncertainties in tetrachloroethylene cancer unit risk estimate 5-105
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xiv DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES
Figure 3-1. Postulated scheme for the metabolism of tetrachloroethylene by the cytochrome
P450 (P450) oxidative pathway and glutathione ^'-transferase (GST)-mediated glutathione (GSH)
conjugation pathway. PCE and identified (*) urinary metabolites: (1) PCE, (2) PCE-Fe-0
intermediate, (3) trichloroacetyl chloride, (4) trichloroacetic acid, (5) PCE oxide, (6) ethandioyl
dichloride, (7) oxalic acid, (8) »Y-(1,2,2-trichlorovinyl) glutathione (TCVG),
(9) »Y-(1,2,2-trichlorovinyl)-A-cysteine (TCVC), (10) A'-acetyl trichlorovinyl cysteine
(NAcTCVC), (11) dichloroacetic acid. Enzymes: cytochrome P450 (P450), GST,
gamma-glutamyltransferase (GGT), dipeptidase (DP), P-lyase, flavin mono-oxygenase-3
(FM03), iV-acetyl transferase (NAT) 3-5
Figure 3-2. Comparison of model predictions for blood concentration with experiment. PCE
inhaled concentration was 72 ppm. Predictions are at different ventilation-to-perfusion ratios
and at an alveolar ventilation rate of 7 L/minute (the geometric mean of values in the Monster
experiment). Standard deviations on the experimental data were very small (e.g., 0.025 mg/L
and 0.003 mg/L at 20 and 140 hours, respectively). Experimental data adapted from Monster
etal. (1979) 3-27
Figure 3-3. Comparison of model predictions for alveolar concentration of
tetrachloroethylene with experimental data on humans. Inhaled concentration is 100 ppm,
7 hours/day, for 5 days, and predictions assume alveolar ventilation rate of 5.02 L/minute and a
ventilation-to-perfusion ratio of 1.0. Experimental data show mean alveolar concentration in
subjects in Stewart et al. (1970). Some points early in the time course were deleted because of
difficulty in obtaining numerical values from the author's plot 3-28
Figure 3-4. Previously published estimates for the total amount of tetrachloroethylene
metabolized at 0.001 ppm (1 ppb) continuous inhalation exposure. All estimates are point
estimates except for Bois et al. (1996) and Chiu and Bois (2006), which are estimates of
combined uncertainty and population variability (95% confidence intervals [CIs]), and
Covington et al. (2007) and Qiu et al. (2009), which are estimates of uncertainty in the
population mean (90% CIs) 3-29
Figure 3-5. Overall structure of updated physiologically based pharmacokinetic (PBPK) model
for tetrachloroethylene and metabolites. Boxes with underlined labels are additions or
modifications of the Chiu et al. (2009) model for trichloroethylene 3-31
Figure 3-6. Comparison of mouse (A-B), rat (C-D), and human (E-F) rates of hepatic oxidation
(A, C, and E) or conjugation (B, D, and F) measured in vitro (symbols) and predicted by the
model (lines). Data shown consist of measurements of tetrachloroethylene in vitro oxidation and
conjugation [solid circle: Dekant et al. (1998), solid square: Green et al. (1990); solid diamond:
Lash et al. (1998); solid triangle: Lash et al. (2007); solid upside-down triangle: Reitz et al.
(1996)], reported fits of in vitro tetrachloroethylene Vmax and Km for oxidation [grey-filled circle:
Costa and Ivanetich (1980); grey-filled square: Costa and Ivanetich (1984); grey-filled diamond:
Lipscomb et al. (1998)TCE; grey-filled triangle: (Wheeler et al., 2001) CH2I2; grey-filled upside-
down triangle: (Wheeler et al., 2001) CH2CI2], and measurements of TCE in vitro conjugation
[open circle: Lash et al. (1998); open square: Lash et al. (1999); open diamond: Green et al.
(1997)]. Model predictions are using baseline parameters (dotted line), overall posterior mode
parameters (solid thick line), and alternative posterior mode parameters (grey lines) 3-36
Figure 3-7. Physiologically based pharmacokinetic (PBPK) simulations of variations with age
and gender in blood concentrations of tetrachloroethylene and its main metabolite trichloroacetic
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xv DRAFT—DO NOT CITE OR QUOTE
-------
acid (TCA). Simulations are for continuous lifetime oral exposure at a constant daily intake of
1 |ig/kg-day 3-50
Figure 4-1. Visual contrast sensitivity functions for control and exposed participants in a study
of workers in a day-care center located in a building with a dry-cleaning facility (Schreiber et al.,
2002). The X-axis represents the frequency of the stimulus bars, with finer bars toward the right.
The Y-axis represents the inverse of the contrast at which the subject could no longer distinguish
the orientation of the bars (threshold). Blue circles (top line) = controls; red circles (bottom line)
= exposed. For any frequency, a higher contrast sensitivity threshold represents better visual
function. Visual contrast sensitivity was significantly lower across all spatial frequencies in
exposed workers at a day-care center colocated with a dry-cleaning facility compared with their
matched controls 4-23
Figure 4-2. Visual contrast sensitivity functions for control and exposed participants in
residential exposure study (Schreiber et al., 2002). The X-axis represents the frequency of the
stimulus bars, with finer bars toward the right. The Y-axis represents the inverse of the contrast
at which the subject could no longer distinguish the orientation of the bars (threshold). Blue
circles (top line) = controls; red circles (bottom line) = exposed. For any frequency, a higher
contrast sensitivity threshold represents better visual function. Visual contrast sensitivity was
significantly lower across all spatial frequencies in exposed residents of apartments in building
with dry-cleaning facilities compared with their matched controls 4-28
Figure 4-3. Incidences of hepatocellular adenomas (A) and hepatocellular adenomas and
carcinomas (B) in mice exposed to DEHP. Ito et al. (2007) exposed PPARa null [-/-] and wild-
type [+/+] Sv/129 mice for 22 months; David et al. (1999) exposed B6C3Fi wild-type [+/+] mice
for up to 104 weeks. Data are presented as incidence +/- SD assuming a binomial distribution
for each group. Single asterisks (*) indicate a significant difference from controls of the same
genotype in the same study (Fisher exact test, p < 0.05). Double asterisks (**) indicate a
significant trend with dose in the study (Cochran Armitage test, p < 0.05). All pair-wise cross-
study comparisons between like dose groups (e.g., Ito et al. [-/-] 500 ppm compared with David
et al. [+/+] 500 ppm) were not significant (Fisher exact testp > 0.05). Because David et al.
(1999) reported only adenomas and carcinomas, the cholangiocellular carcinoma reported by Ito
et al. (2007) in DEHP-exposed PPARa null mice was excluded from analyses. Adapted from
Guyton et al. (2009) 4-169
Figure 5-1. Exposure-response array for neurotoxicological inhalation studies considered for
RfC development (listed in Table 5-1). PODs (HEC for LOAELs and NOAELs) are displayed
and labeled by study, effect, and duration. Principal studies selected for RfC derivation are
shaded in blue and the POD range (± 2SE) are presented 5-6
Figure 5-2. Reference concentration values for inhalation exposure to tetrachloroethylene. ..5-18
Figure 5-3. Comparison of candidate RfCs (black squares) supporting the RfC (grey vertical
line) and sample RfCs (open squares) for effects other the critical effect (CNS toxicity). Black
circles = study/endpoint LOAEL in terms of human equivalent concentrations. Open
circles = study/endpoint NOAEL in terms of human equivalent concentrations. Species in
each study is shown in brackets after the reference (mouse: M; rat: R; human: H) 5-25
Figure 5-4. Reference dose values from principal studies following exposure to
tetrachloroethylene 5-31
Figure 5-5. Comparison of candidate RfDs (black squares) supporting the RfD (grey vertical
line) and sample RfDs (open squares) for effects other the critical effect (CNS toxicity). Black
circles = study/endpoint LOAEL in terms of human equivalent dose. Open circles =
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xvi DRAFT—DO NOT CITE OR QUOTE
-------
study/endpoint NOAEL in terms of human equivalent dose. Species in each study is shown
in brackets after the reference (mouse: M; rat: R; human: H) 5-38
Figure 5-6. Mouse liver tumor responses (hepatocellular adenomas or carcinomas), as additional
risk, for two chronic inhalation bioassays (see Table 5-13), plotted against continuous equivalent
concentration (ppm), for male and female mice 5-48
Figure 5-7. Rat mononuclear cell leukemia responses (minus control) in two chronic bioassays
(see Table 5-15), plotted against continuous equivalent exposure (ppm) for (a) male and (b)
female rats 5-51
Figure 5-8. Sequence of steps for extrapolating from tetrachloroethylene bioassays in animals to
human-equivalent exposures expected to be associated with comparable cancer risk (combined
interspecies and route-to-route extrapolation). See Table 5-17 for units 5-54
Figure 5-9: Dose-response modeling of male mouse hepatocellular tumors associated with
inhalation exposure to tetrachloroethylene, in terms of liver total oxidative metabolites; response
data from JISA (1993). Details in Appendix D 5-71
Figure 5-10. Dose-response modeling of female mouse hepatocellular tumors associated with
inhalation exposure to tetrachloroethylene, in terms of liver total oxidative metabolites; response
data from JISA (1993). Details in Appendix D 5-72
Figure 5-11. Dose-response modeling of male mouse hemangiomas or hemangiosarcomas
associated with inhalation exposure to tetrachloroethylene, in terms of tetrachloroethylene AUC
in blood; response data from JISA (1993). Details in Appendix D 5-78
Figure 5-12. Dose-response modeling of female and male rat MCLs associated with inhalation
exposure to tetrachloroethylene, in terms of tetrachloroethylene AUC in blood; response data
from JISA (1993). Details in Appendix D 5-81
Figure 5-13. Dose-response modeling of male rat tumors—kidney, brain gliomas, interstitial cell
tumors, MCLs—associated with inhalation exposure to tetrachloroethylene, in terms of
tetrachloroethylene AUC in blood; response data from NTP (1986b). Details in Appendix D... 5-
87
Figure 5-14. Comparison of inhalation unit risks for tetrachloroethylene derived from rodent
bioassays using PBPK-based dose metrics and administered concentration. Symbols represent
results using the posterior mode PBPK model results, with filled symbols representing the
preferred dose metrics (Tables 5-18 and 5-19). Red-filled symbols use the multistage model
with all dose groups; green-filled symbols use a different dose-response approach in response to
NRC (2010) comments. Solid error bars show the range of estimates using the range of posterior
modes for the human PBPK model-based conversion to a human equivalent unit risk (Table 5-
18). Dashed error bars show the range of unit risk estimates (based on administered
concentration) using alternative dose-response models with goodness-of-fit p-values > 0.10
(Table 5-20) 5-93
Figure 5-15. Comparison of oral slope factors for tetrachloroethylene, derived from rodent
bioassays using PBPK-based dose metrics and route-to-route extrapolation. Symbols represent
results using the posterior mode PBPK model results, with filled symbols representing the
preferred dose metrics (Table 5-21). Red-filled symbols use the multistage model with all dose
groups; green-filled symbols use a different dose-response approach in response to NRC (2010)
comments. Solid error bars show the range of estimates using the range of posterior modes for
the human PBPK model-based conversion to a human equivalent unit risk (Table 5-21) 5-99
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xvii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS
8-OHdG
8-hydroxydeoxyguanosine
AAP
alanine aminopeptidase
ALT
alanine transferase
AST
aspartase amino transaminase
AT SDR
Agency for Toxic Substances and Disease Registry
AUC
area-under-the-curve
BMC
benchmark concentration
BMCL
lower bound benchmark concentration
BMD
benchmark dose
BMDL
lower bound benchmark dose
BMDS
Benchmark Dose Software
BMDU
95% upper bound benchmark dose
BUN
blood urea nitrogen
BW
body weight
CARB
California Air Resources Board
CASRN
Chemical Abstracts Service Registry Number
CCI
Color Confusion Index
CI
confidence interval
CLL
chronic lymphocytic leukemia
CNS
central nervous system
co2
carbon dioxide
CT
carbon tetrachloride
CYP P450
cytochrome P450
DCA
dichloroacetic acid
DEHP
di(2-ethylhexyl)phthalate
EEGs
electroencephalograms
EPA
U.S. Environmental Protection Agency
FDA
Food and Drug Administration
FM03
flavin-containing monooxygenase 3
GGT
gamma-glutamyltransferase
GSH
glutathione
GST
glutathione ^'-transferase
GSTx
glutathione ^'-transferase isoform, where x denotes different isoforms (such as M,
T, P, S, Z)
HEC
human equivalent concentration
HSIA
Halogenated Solvents Industry Alliance
i.p.
intraperitoneal
IAP
intestinal alkaline phosphatase
IARC
International Agency for Research on Cancer
IOM
Institute of Medicine
IPCS
International Programme on Chemical Safety
IRIS
Integrated Risk Information System
IUGR
intrauterine growth restriction
6/24/11
This document is a draft for review purposes only and does not constitute Agency policy
xviii DRAFT—DO NOT CITE OR QUOTE
-------
JISA
Japan Industrial Safety Association
Km
Michaelis-Menten constant
LECioS
95% lower confidence limits on the air concentrations associated with a 10%
extra risk of cancer incidence
LGL
large granular lymphocyte
LOAEL
lowest-observed-adverse-effect level
MLE
maximum likelihood estimate
MCA
monochloroacetic acid
MCL-5
microsomal epoxide hydrolase
MCL
mononuclear cell leukemia
MOA
mode of action
MRL
minimal risk level
NAG
/V-acetyl-P-D-glucosaminidase
NCI
National Cancer Institute
NHL
non-Hodgkin's lymphoma
NIOSH
National Institutes of Occupational Safety and Health
NK
natural killer
NOAEL
no-ob served-adverse-effect level
NRC
National Research Council
NTP
National Toxicology Program
NYSDOH
New York State Department of Health
NYSOAG
New York State Office of Attorney General
OR
odds ratio
P450
cytochrome P450
PBPK
physiologically based pharmacokinetic
PCO
palmitoyl CoA oxidation
PHG
public health goal
POD
point of departure
PPAR
peroxisome proliferater activated receptor
PPAR-a
peroxisome proliferater activated receptor, alpha isoform
PPAR-5
peroxisome proliferater activated receptor, delta isoform
RBP
retinol binding protein
REAL
revised European-American Lymphoma
RfC
reference concentration
RfD
reference dose
RfV
reference value
RR
relative risk
SAP
Scientific Advisory Panel
SCE
sister chromatid exchange
SES
socio-economic status
SGA
small for gestational age
SIR
standardized incidence ratio
SMR
standardized mortality ratio
SSB
single-strand break
TCA
trichloroacetic acid
TCE
tri chl oroethy 1 ene
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xix DRAFT—DO NOT CITE OR QUOTE
-------
TCOH trichloroethanol
TCVC »Y-(1,2,2,-trichlorovinyl)-L-cysteine
TCVCSO »V-(1,2,2,-trichlorovinyl)-L-cysteine sulfoxide
TCVG »Y-(1,2,2-trichlorovinyl) glutathione
TNAP tissue nonspecific alkaline phosphatase
TWA time-weighted average
U/L international units per liter
UDS unscheduled DNA synthesis
UF uncertainty factor
VCS visual contrast sensitivity
Ve ventilation rate
VEP visually evoked potential
Vmax maximum velocity
WHO World Health Organization
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xx DRAFT—DO NOT CITE OR QUOTE
-------
FOREWORD
The purpose of this Toxicological Review is to provide scientific support and rationale
for the hazard and dose-response assessment in IRIS pertaining to chronic exposure to
tetrachloroethylene. It is not intended to be a comprehensive treatise on the chemical or
toxicological nature of tetrachloroethylene.
The intent of Section 6, Major Conclusions in the Characterization of Hazard and Dose-
Response, is to present the significant 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 by
addressing the quality of data and related uncertainties. The discussion is intended to convey the
limitations of the assessment and to aid and guide the risk assessor in the ensuing steps of the
risk assessment process.
For other general information about this assessment or other questions relating to IRIS,
refer to EPA's IRIS Hotline at (202) 566-1676 (phone), (202) 566-1749 (fax), or
hotline.iris@epa.gov (e-mail address).
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xxi DRAFT—DO NOT CITE OR QUOTE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGERS
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Kathryn Z. Guyton
Karen A. Hogan
AUTHORS
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
AmbujaBale
Stanley Bar one
Weihsueh A. Chiu
Glinda Cooper
Rebecca Brown Dzubow1
Glinda Cooper
Maureen R. Gwinn
Leonid Kopylev
Susan Makris
Cheryl Siegel Scott
Ravi Subramaniam
CONTRIBUTORS
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
David Bussard
Jane Caldwell
Robert McGaughy1'2
Jean Parker1
Deborah Rice1
Bob Sonawane
Paul White
Larry Lavcovic1
formerly with National Center for Environmental Assessment
2Former Chemical Manager.
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xxii DRAFT—DO NOT CITE OR QUOTE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS (continued)
REVIEWERS
This document has been provided for review to EPA scientists, interagency reviewers
from other federal agencies and White House offices, and the public, and peer reviewed by
independent scientists external to EPA. Appendix A provides a summary and EPA's disposition
of the comments received from the independent external peer reviewers and from the public.
INTERNAL PEER REVIEWERS
Hugh Barton, U.S. Environmental Protection Agency, National Health and Environmental
Effects Research Laboratory, Research Triangle Park, NC
Robert Benson, U.S. Environmental Protection Agency, Office of Partnerships and Regulatory
Assistance, Region 8, Denver, CO
William Boyes, U.S. Environmental Protection Agency, National Health and Environmental
Effects Research Laboratory, Research Triangle Park, NC
Jim Cogliano, U.S. Environmental Protection Agency, National Center for Environmental
Assessment, Washington, DC
Herman Gibb, formerly with U.S. Environmental Protection Agency, National Center for
Environmental Assessment, Washington, DC
John Lipscomb, U.S. Environmental Protection Agency, National Center for Environmental
Assessment, Cincinnati, OH
Elizabeth Margosches, U.S. Environmental Protection Agency, Office of Pollution, Prevention,
and Toxics, Washington, DC
Deirdre Murphy, U.S. Environmental Protection Agency, Office of Air Quality and Planning and
Standards, Research Triangle Park, NC
Onyemaechi Nweke, U.S. Environmental Protection Agency, Office of Policy, Economics, and
Innovation, Washington, DC
Brenda Foos, U.S. Environmental Protection Agency, Office of the Administrator, Office of
Children's Health Protection, Washington, DC
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xxiii DRAFT—DO NOT CITE OR QUOTE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS (continued)
Bruce Rodan, U.S. Environmental Protection Agency, National Center for Environmental
Assessment, Washington, DC
Diana M. Wong, U.S. Environmental Protection Agency, Office of Science and Technology,
Office of Water, Washington, DC
Tracey Woodruff, formerly with U.S. Environmental Protection Agency, Office of Policy,
Economics, and Innovation, Washington, DC
CONSULTANTS
Anne Aschengrau, Department of Epidemiology, Boston University School of Public Health,
Boston, MA
Matt Bogdanffy, Lincoln University, Lincoln University, PA
George Lucier, formerly with National Institute of Environmental Health Sciences, Research
Triangle Park, NC
Robert Park, National Institute for Occupational Safety and Health, Education and Information
Division, Cincinnati, OH
Val Schaeffer, Occupational Safety and Health Administration, Directorate for Health Standards,
Washington, DC
NEUROTOXICITY EXPERT REVIEW PANEL REVIEWERS: Public Workshop,
February 25,2004
Kent Anger (Chair), Center for Research on Occupational and Environmental Toxicology,
Oregon Health and Science University, Portland, OR
Rosmarie Bowler, San Francisco State University, San Francisco, CA
Diana Echeverria, Battelle Center for Public Health Research and Evaluation, Seattle, WA
Fabriziomaria Gobba, Dipartimento di Scienze Igienistiche, Universita di Modena e Reggio
Emilia, Modena, Italy
William Merigan, Department of Ophthalmology and Center for Visual Science,
University of Rochester School of Medicine and Dentistry, Rochester, NY
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xxiv DRAFT—DO NOT CITE OR QUOTE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS (continued)
EXTERNAL PEER REVIEWERS
Sam Kacew (Chair)
University of Ottawa, Ontario, Canada
Bruce H. Alexander
University of Minnesota School of Public Health, Minneapolis, MN
Margit L. Bleecker
Center for Occupational and Environmental Neurology, Baltimore, MD
Gary P. Carlson
Purdue University, West Lafayette, IN
Linda D. Cowan
University of Oklahoma Health Sciences Center, Oklahoma City, OK
Mary E. Davis
West Virginia University, Morgantown, WV
H. Christopher Frey
North Carolina State University, Raleigh NC
Joseph R. Landolph
University of Southern California, Los Angeles, CA
M.E. (Bette) Meek
University of Ottawa, Ontario, Canada
David C. McMillan
University of Nebraska Medical Center, Omaha, NE
M. Christopher Newland
Auburn University, Auburn, AL
Julia Quint
California Department of Public Health (retired), Berkeley, CA
Gary L. Rosner
University of Texas M.D. Anderson Cancer Center, Houston, TX
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xxv DRAFT—DO NOT CITE OR QUOTE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS (continued)
Ivan Rusyn
University of North Carolina, Chapel Hill, NC
Rolf Schulte-Hermann
Medical University of Vienna, Austria
Irvin R. Schultz
Battelle Pacific Northwest Division, Sequim, WA
Robert Snyder
Rutgers, the State University of New Jersey, Piscataway, NJ
Roberta F. White
Boston University School of Public Health, Boston, MA
Luoping Zhang
University of California, Berkeley, CA
Yiliang Zhu
University of South Florida, Tampa, FL
ACKNOWLEDGMENTS
The authors would like to acknowledge the contributions of the following individuals:
Terri Konoza of NCEA who managed the document production activities; Ellen Lorang, who
provided HERO literature database support; Cristopher Broyles of IntelliTech Systems, Inc. and
Heidi Glick of ECFlex, Inc., who provided editorial support; Lana Wood, Debbie Kleiser, and
Crystal Lewis of ECFlex, Inc. and Stacey Herron of IntelliTech Systems, Inc., who provided
word processing support; Patricia von Brook of KBM Group, who provided editorial support for
the previous draft; and Christine Chang of KBM Group, who provided word processing support
for the previous draft.
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 xxvi DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
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
tetrachloroethylene. 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. The RfD (expressed in units of mg/kg-day) is defined as an estimate (with
uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human
population (including sensitive subgroups) that is likely to be without an appreciable risk of
deleterious effects during a lifetime. The inhalation RfC (expressed in units of mg/m3) is
analogous to the oral RfD, but provides a continuous inhalation exposure estimate. The 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 |ig/m3 air breathed.
Development of these hazard identification and dose-response assessments for
tetrachloroethylene has followed the general guidelines for risk assessment set forth by the
National Research Council (NRC. 1983. 1994). 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.
1986c). Guidelines for Mutagenicity Risk Assessment (U.S. EPA. 1986b). Recommendations for
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 1-1 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
and Documentation of Biological Values for Use in Risk Assessment (U.S. EPA. 1988b).
Guidelines for Developmental Toxicity Risk Assessment (U.S. EPA. 1991a). Interim Policy for
Particle Size and Limit Concentration Issues in Inhalation Toxicity (Kaufman et al.. 2009).
Methods for Derivation of Inhalation Reference Concentrations and Application of Inhalation
Dosimetry (U.S. EPA. 1994). Use of the Benchmark Dose Approach in Health Risk Assessment
(U.S. EPA. 1995). Guidelines for Reproductive Toxicity Risk Assessment (U.S. EPA. 1996a).
Guidelines for Neurotoxicity Risk Assessment (U.S. EPA. 1998b). Science Policy Council
Handbook. Risk Characterization (U.S. EPA. 2000b). Benchmark Dose Technical Guidance
Document (U.S. EPA. 2000a). 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.
2006c). and A Framework for Assessing Health Risks of Environmental Exposures to Children
(U.S. EPA. 2006b).
The literature search strategy employed for tetrachloroethylene 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. A comprehensive literature review was carried
out through October 2010.
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 1 -2 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
2. BACKGROUND
2 1 USES AND PHYSICAL/CHEMICAL PROPERTIES
Tetrachloroethylene is a widely used solvent that is produced commercially for use in dry
cleaning, textile processing, and metal-cleaning operations. It has the following use pattern:
55% as a chemical intermediate, 25% for metal cleaning and vapor degreasing, 15% for dry
cleaning and textile processing, and 5% for other unspecified uses (ATSDR. 1997a).
Table 2-1 lists the physical and chemical properties of tetrachloroethylene (ATSDR.
1997a). The reference citations can be found in the Agency for Toxic Substances and Disease
Registry (ATSDR) document and are not included in the reference list for this document.
2 2 OCCURRENCE AND EXPOSURE
Tetrachloroethylene has been detected in ground water and surface water as well as in air,
soil, food, and breast milk. The primary exposure routes of concern are inhalation of vapor and
ingestion of contaminated water. Although dermal exposure is possible via contaminated tap
water during showering, bathing, or swimming, this is generally not considered a major route of
exposure.
2.2.1. Air
Because of its high volatility, there is considerable potential for release of
tetrachloroethylene into the atmosphere. Once in the air, it is not susceptible to wet deposition
because of its hydrophobicity. The primary method for removal is photooxidation to
trichloroacetyl chloride, trichloroacetic acid (TCA), carbon monoxide, ozone, and phosgene
(U.S. EPA. 1982). However, this reaction is very slow, so tetrachloroethylene is not implicated
in the buildup of any of the reaction products in the troposphere. Though the half-life of
perchloroethylene can vary based on season and environmental conditions, it has been estimated
at 96 days under typical conditions (ATSDR. 1997a).
Ambient tetrachloroethylene concentrations vary from source to source and with
proximity to the source. Outdoors, the high volatility of perc leads to increased ambient air
concentrations near points of use (ATSDR. 1997a; U.S. EPA. 1996b). Specific to early lifestage
exposure scenarios, elevated ambient air concentrations include measurements taken outside of a
daycare center adjacent to a dry cleaner (NYSDOH. 2005c) and on a playground near a factory
(Monster and Smolders. 1984a). It should be noted that outdoor concentrations can vary widely
within a period of a few hours as a function of wind velocity and direction, precipitation,
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 2-1 DRAFT—DO NOT CITE OR QUOTE
-------
1 humidity, and sunlight. ATSDR (1997a) reported mean tetrachloroethylene concentrations of
2 8.8 |ig/m3 in areas close to points of release.
Table 2-1. Physical and chemical properties of tetrachloroethylene
Property
Information
Reference
Molecular weight
165.83
Lide (1990)
Color
Colorless
Sax and Lewis (1987)
Physical state
Liquid (at room temperature)
Sax and Lewis (1987)
Melting point
! 19EC
Lide (1990)
Boiling point
121EC
Lide (1990)
Density at 20EC
1.6227 g/mL
Lide (1990)
Density at 25EC
No data
Odor
Ethereal
HSDB (1996)
Odor threshold: water
0.3 ppm
U.S. EPA (1988)
Odor threshold: air
1 ppm
U.S. EPA (1988)
Solubility: water at 25EC
150 mg/L
HSDB (1996)
Solubility: organic solvent(s)
Miscible with alcohol, ether,
chloroform, benzene, solvent
hexane, and most of the fixed
and volatile oils
HSDB (1996)
Partition coefficients: Log K0w
3.4
HSDB (1996)
Partition coefficients: Log K0c
2.2B2.7
Seip et al. (2008)
Zytner et al. (2009b)
Vapor pressure at 25EC
18.47 mm Hg
HSDB (1996)
Henry's law constant at 25EC
1.8 H 10"2 atm-m3/mol
Gossett (1987)
Autoignition temperature
No data
Flashpoint
None
HSDB (1996)
Flammability limits
Nonflammable
HSDB (1996)
Conversion factors, air
1 mg/L = 141.4 ppm
1 ppm = 6.78 mg/m3
HSDB (1996)
Explosive limits
No data
Source: ATSDR (1997a).
3 EPA has carried out modeling to characterize the geographic distribution of
4 tetrachloroethylene for its National-Scale Air Toxics Assessment database (U.S. EPA. 1996b).
5 Median census tract-based tetrachloroethylene concentrations across the United States were
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 2-2 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
estimated at about 0.3 |ig/m3 for urban areas and 0.1 |ig/m3 for rural areas (75% upper
percentiles of 0.4 and 0.2 |ig/m3, respectively). The California Air Resources Board (CARB.
1998) reported a statewide median air concentration of 0.3 |ig/m3 in 2001, which represents the
lowest value in what has been a decreasing trend since 1990. Note that these averages, which are
based on geographic areas, only characterize the likely exposure of individuals who spend an
equal amount of time in all parts of the defined area, and they may, therefore, significantly
underestimate the exposure of individuals who consistently spend time in subareas that have
higher tetrachloroethylene concentrations.
Near points of use, such as dry cleaners or industrial facilities, indoor exposure to
tetrachloroethylene is more significant than outdoor exposure (U.S. EPA. 2001a). Adgate and
colleagues measured tetrachloroethylene in outside and indoor air at school, indoor air at home,
and using personal samples on children, and demonstrated that perc levels are lower in homes
with greater ventilation (Adgate et al.. 2004b) and in homes in non-urban settings (Adgate et al..
2004a; Adgate et al.. 2004b). Indoor air concentrations in an apartment above a dry cleaning
shop have been measured at up to 4.9 mg/m3 (Verberkand Scheffers. 1980) (see also Altmann et
al.. 1995; Garetano and Gochfeld. 2000; McDermott et al.. 2005; Schreiber et al.. 1993;
Schreiber et al.. 2002). Measurements have also been made in a daycare center adjacent to a dry
cleaners (NYSDOH. 2005a. b, c), and in a classroom exposed to perc from an air -emission from
a small chemical factory" (Monster and Smolders. 1984a). Mean concentrations inside dry
cleaning facilities were reported to be 454 - 1390 mg/m3 in the United States and 164 mg/m3 in
Nordic countries during the 1960s and 1970s. Overall levels declined from 95 - 210 mg/m3 in
the 1980s to 20 - 70 mg/m3 over the next decades in these countries (Gold et al.. 2008; Lynge et
al.. 2006; Lynge et al.. 2011).
The off-gassing of garments that have recently been dry-cleaned may be of concern (see
also Thomas et al.. 1991; Tichenor et al.. 1990). In the home, tetrachloroethylene vapors may
off-gas from the clothes of occupationally exposed individuals, or they may come directly from
the exhaled breath of exposed workers (ATSDR. 1997a) (see also Aggazzotti et al.. 1994a;
Aggazzotti et al.. 1994b). Relatively high tetrachloroethylene air concentrations have been
measured in the proximity of freshly dry-cleaned clothing stored in small, close spaces. A
residential closet storing newly dry-cleaned clothing had an air concentration of 2.9 mg/m3 after
1 day, which rapidly declined to 0.5 mg/m3 and persisted for several days (Tichenor et al.. 1990).
There is one documented mortality case: a 2-year-old boy was found dead after being put to
sleep in a room with curtains that had been incorrectly dry-cleaned (Gamier et al.. 1996).
Dry-cleaned garments transported in an automobile may also lead to unexpectedly high
levels of exposure. Park et al. (1998) used simulated driving cycles to estimate the
concentrations of several contaminants emitted from in-vehicle sources; see also see (Gulvas and
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 2-3 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Hemmerling. 1990). Using dry-cleaned clothes as a source, tetrachloroethylene levels inside a
stationary vehicle after 30 minutes reached 0.230 mg/m3. Approximating these exposures is not
easy because specific exposure levels would depend on many factors: car velocity, wind speed,
ventilation, and time spent in the automobile. Another study demonstrating exposure in a car
found that transporting a freshly dry-cleaned down jacket in a car resulted in a cabin air
concentration of 24.8 mg/m3 after 108 minutes (Chien. 1997).
Air exposure may also occur during showering or bathing as dissolved
tetrachloroethylene in the warm tap water is volatilized. Rao and Brown (1993) used an adult
physiologically based pharmacokinetic (PBPK) model combined with a microenvironmental
exposure model to estimate the dose received by inhalation exposure during showering and
bathing as well as by dermal exposure to the water. The tap water concentration of
tetrachloroethylene was 1 mg/L, which is probably a higher concentration than exists in most
water supplies. They also demonstrated that a majority of the tetrachloroethylene in the blood,
as a result of their bathing scenario, resulted from inhalation exposure, while about 15% resulted
from dermal absorption.
2.2.2. Water
Because of its relatively low aqueous solubility (see Table 2-1), it is not likely that
volatilized tetrachloroethylene will enter surface or rain water. However, it has been detected in
drinking water, ground water, and surface water (ATSDR. 1997a; Canada and Health Canada.
1993; Lagakos et al.. 1986; U.S. EPA. 2001a). Most of this contamination is probably due to
release in water following industrial use or by public use of consumer products.
Unless a surface water body is in the vicinity of a highly contaminated site, surface
waters are expected to have a lower concentration of tetrachloroethylene than ground water. In
an estimate of drinking water contamination in California, McKone and Bogen (1992) assumed
that surface water would have a negligible contribution to the concentration of
tetrachloroethylene measured in drinking water. Based on data from wells in California, they
estimated an average drinking water concentration of 0.3 |ig/L, with a standard deviation of 0.35
l-ig/L.
In areas near sources of contamination, ground water, and surface water concentrations
can be considerably higher than average. Because the density of tetrachloroethylene is about
60% higher than that of water, tetrachloroethylene is expected to accumulate near the bottom of a
stagnant receiving water body after a large-volume point discharge. Water samples collected
near the bottom of the St. Clair River near Sarnia, Ontario, downstream from several petroleum-
based production facilities, contained tetrachloroethylene concentrations ranging from 0.002 to
34.6 |ig/L (Canada and Health Canada. 1993). The concentrations in 17 samples of surface
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 2-4 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
water from the lower Niagara River in New York State in 1981 averaged 0.036 |ig/L (with a
maximum of 0.134 |ig/L; (Canada and Health Canada. 1993).
Exposure models have been developed to predict the fate and transport of organic
compounds such as tetrachloroethylene in environmental media, including air, water, and soil.
The outputs from two similar but independently developed environmental exposure models,
CalTOX and Fug30NT, were compared for a scenario designed to reproduce a residential area
near an industrial contamination site (Maddalena et al.. 1995). in which 75 moles/day are
released into the air and 0.7 moles/day are released into surface water. Although the soil
predictions differed, the predictions of tetrachloroethylene in air and ground water were similar,
with the concentration of air predicted by CalTOX approximately 6 |ig/m3 and the surface water
concentration 82 |ig/L. It should be noted that agreement of the models does not confirm the
validity of either one, but lends some support to the usefulness of the results.
The off-gassing of tetrachloroethylene from a drinking water supply can result in
exposure. In 1976, EPA measured tetrachloroethylene levels ranging from 800 to 2,000 |ig/L in
drinking water samples in Massachusetts (Paulu et al.. 1999). Similar levels were reported
elsewhere in New England. These concentrations were attributed to the vinyl-lined asbestos-
cement pipes that were used to carry water in this area (Webler and Brown. 1993). Letkiewicz et
al. (1982) estimated that 53% of newborn infants are formula-fed from drinking water sources
and the other 47% receive all of their fluid from breast milk. Taking into account volatilization
during boiling of water, they indicate that the uptake of tetrachloroethylene in formula-fed
infants on a mg/kg-day basis is 10 times higher than in adults with the same level of drinking
water contamination. In addition, incidental water consumption may occur for children when
swimming or bathing (U.S. EPA. 2008).
Although dermal exposure is possible via contaminated tap water during showering,
bathing, or swimming, this is generally not considered a major route of exposure (Nakai et al..
1999; Poet et al.. 2002; Stewart and Dodd. 1964). Rao and Brown (1993) demonstrated that only
15%) of the tetrachloroethylene in the blood resulted from dermal exposure as compared to
inhalation of vapors.
2.2.3. Food
Certain foods have been found to be contaminated with tetrachloroethylene (U.S. EPA.
2001a); (also see Daft. 1988; Heikes and Hopper. 1986; McConnell etal.. 1975; U.S. EPA.
2001a). Because of the lipophilic nature of tetrachloroethylene, it may bind to lipid molecules in
such foods as margarine, oils, meats, and other fatty foods stored in areas where there is
tetrachloroethylene in the air (Schreiber. 1997; U.S. EPA. 2001a). In 1988, elevated
tetrachloroethylene levels were seen in margarine and butter samples obtained from grocery
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 2-5 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
stores located near dry cleaning facilities (Entz and Diachenko. 1988); see also (TJhler and
Miller. 1988). Further studies confirmed that close proximity to a dry cleaning facility was
associated with elevated tetrachloroethylene levels in butter samples (Kacew and Lambert.
1997). Nonetheless, food is not considered to be a major exposure pathway. Other sources of
information about tetrachloroethylene in foods are the Food and Drug Administration (FDA.
2003) and (Fleming-Jones and Smith. 2003).
2.2.4. Soil
Where contamination occurs, perc can be measured in soil (U.S. EPA. 2001a). This
pathway for ingestion of perc has not been directly examined. A clear need exists to evaluate this
pathway particularly for children with pica who can ingest high quantities of contaminated soil
through hand-to-mouth activity, as has been shown for lead (U.S. EPA. 2008).
2.2.5. Breast Milk
Due to its lipid solubility, tetrachloroethylene can concentrate in human breast milk
(Bagnell and Ellenberger. 1977; NYSDOH. 2000; Pellizzari etal.. 1982; Schreiber. 1993. 1997;
Schreiber et al.. 2002; Sheldon et al.. 1985; U.S. EPA. 2001a). as well as in milk from cows
(Wanner et al.. 1982). goats (Hamada and Tanaka. 1995). and rats (Byczkowski and Fisher.
1994; Byczkowski et al.. 1994). Breast milk can contain high concentrations of
tetrachloroethylene and some of its toxic metabolites. Reported levels of tetrachloroethylene in
breast milk have ranged up to 43 |ig/L in the general population (U.S. EPA. 2001a). In one case
study, the breast milk of a woman was found to contain 10 mg/L of tetrachloroethylene 1 hr
following a visit to her husband at his work in a dry cleaning establishment. This concentration
dropped to 3 mg/L after 24 hrs. Her child suffered from obstructive jaundice and hepatomegaly,
but these conditions improved when breastfeeding was discontinued (Bagnell and Ellenberger.
1977).
Physiologically based pharmacokinetic (PBPK) models have been utilized to estimate
perc doses from milk to the human infant (Byczkowski and Fisher. 1995; Fisher et al.. 1997;
Schreiber. 1993). and rat (Byczkowski etal.. 1994). Schreiber (1993) used a PBPK model to
estimate the dose a nursing infant might receive from an exposed mother's breast milk. Using
different exposure scenarios, Schreiber (1993) predicted that human breast milk concentrations
could range from 1.5 mg/L for a typical residential scenario, 16-3,000 mg/L for a residential
scenario near a dry cleaner, and to 857-8,440 mg/L for an occupational scenario. Assuming that
a 7.2-kg infant ingests 700mL of breast milk per day, Schreiber estimated the dose to the infant
could range from 0.0001 to 0.82mg/kg/day. This study showed that it is possible for the dose an
infant receives through breast milk to approach levels that could result in adverse health effects
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 2-6 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
and exceed the 1988 EPA RfD of 0.01 mg/kg-day (U.S. EPA. 1988a). Actual indoor air
concentrations (24-hr average), as measured in apartments in New York State, were used to
predict potential levels in breast milk in these modeling scenarios. The apartments included one
located above a dry cleaning facility that used an old dry-to-dry machine (average concentration,
45.8mg/m3), three located above facilities that used transfer machines (average concentration,
7.7mg/m3), and two located above facilities that used newer dry-to-dry machines (average
concentration 0.25 mg/m3) (Schreiber. 1993). The predicted breast milk concentrations in these
scenarios ranged from 16 to 3,000 |ig/L. Assuming that a 7.2 kg infant ingests 700 mL of breast
milk per day, Schreiber (1993) determined that the infant dose from milk could range from
0.0015 to 0.3 mg/kg-day.
Using the same exposure conditions as Schreiber (1993). Byczkowski and Fisher (1995)
predicted lower doses to the infant (0.0009-0.202 mg/kg-day), although these doses approached
levels that could result in adverse health effects. Exceedances of the RfD were seen only in
those apartments above old dry-to-dry machines (0.202 mg/kg-day) or above transfer machines
(0.029 mg/kg-day). Using milk production and suckling variables, Fisher et al. (1997) estimated
the dose that a human infant might receive after maternal occupational exposure to be 25
ppm/day.
Ingestion through breast milk and infant exposures is discussed further in Section 4.8.
However, Schreiber (1997) has suggested that if infants live adjacent to or in close proximity to
dry cleaning facilitates, the dose received through breast milk ingestion will be insignificant
when compared with that from their inhalation exposure.
2.2.6. Direct Ingestion
In rare circumstances, direct ingestion of tetrachloroethylene has been documented. A
6-year-old boy who directly ingested 12-16 g tetrachloroethylene experienced drowsiness,
vertigo, agitation, and hallucinations. He then lost consciousness and went into a coma, and later
recovered (Koppel et al.. 1985). Follow-up testing on the boy was not reported, so any potential
long-term effects of the exposure are unknown.
This document is a draft for review purposes only and does not constitute Agency policy
6/24/11 2-7 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
3. TOXICOKINETICS
3 1 ABSORPTION
Tetrachloroethylene is rapidly absorbed into the bloodstream following oral and
inhalation exposures. It can also be absorbed across the skin following dermal exposure to either
pure or diluted solvent or vapors (Nakai et al.. 1999; Poet et al.. 2002; Stewart and Dodd. 1964).
3.1.1. Inhalation
The major exposure route for tetrachloroethylene is considered to be inhalation (I ARC.
1995; U.S. EPA. 1985b). Pulmonary uptake of tetrachloroethylene is rapid; however, complete
tissue equilibrium occurs only after several hours. Absorption into the systemic circulation
through pulmonary uptake is proportional to the ventilation rate, the duration of exposure, and, at
lower ambient concentrations to which humans are likely to be exposed, the concentration in the
inspired air (Hake and Stewart. 1977; Monster et al.. 1979).
Chiu et al. (2007) reported that peak levels of tetrachloroethylene in venous blood and air
occurred near the end of a 6-hour inhalation exposure to 1 ppm and declined thereafter. In the
Monster et al. (1979) study, uptake after 4 hours was 75% of its value at the onset of exposure.
Increased physical activity increases uptake but lowers the alveolar partial pressure, thus
removing more tetrachloroethylene from the alveoli, resulting in a longer time to reach tissue
equilibrium (Pezzagno et al.. 1988).
The blood:gas partition coefficient for tetrachloroethylene describes how the chemical
will partition itself between the two phases. Specifically, it is the ratio of concentrations at
steady state; i.e., when all rates are constant after equilibrium has been reached. Reported values
for the coefficient in humans range from around 10-20 (e.g., Byczkowski and Fisher. 1994; Droz
and Guillemin. 1986; Gearhart et al.. 1993; Hattis et al.. 1990; Reitz et al.. 1996; Ward et al..
1988). meaning that if tetrachloroethylene is in equilibrium, the concentration in blood will be
10-20 times higher than the concentration in the alveoli.
Opdam and Smolders (1986) determined concentrations of tetrachloroethylene in alveolar
air for 1-60-second residence times (the time interval from the beginning of an inhalation to the
end of the next inhalation) for six volunteers exposed to 0.5-9.8 ppm of chemical for
1-60 minutes. These investigators found the concentrations of tetrachloroethylene in alveolar air
to decrease with residence times for breaths during exposure periods but to increase during
postexposure for residence times less than 10 seconds. Alveolar air tetrachloroethylene
concentration correlated with the concentrations in pulmonary artery mixed venous blood.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-1 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Like the studies in humans, inhalation studies in laboratory animals provide clear
evidence that tetrachloroethylene is readily absorbed via the lungs into the systemic circulation
(Dallas et al.. 1994a; Pegg et al.. 1979).
3.1.2. Oral
Gastric absorption of tetrachloroethylene occurs at a relatively rapid rate and is
essentially complete. Close to 100% of oral doses are absorbed from the gut, according to
reports of several studies conducted in mice, rats, and dogs (Dallas et al.. 1994a. 1995; Frantz
andWatanabe. 1983; Pegg et al.. 1979; Schumann et al.. 1980). Absorption into the systemic
circulation was indicated by blood tetrachloroethylene levels of 21.5 jag/m L following accidental
ingestion of the chemical by a 6-year-old boy (Koppel et al.. 1985).
3.1.3. Dermal
Absorption of tetrachloroethylene by humans following dermal exposure to vapors of the
chemical has been reported to be relatively insignificant (only 1%) when compared with
absorption via inhalation of vapors (Nakai et al.. 1999; Riihimaki and Pfaffli. 1978). The
amount of chemical absorbed during the immersion of one thumb in liquid tetrachloroethylene is
equivalent to the uptake during inhalation of 10-15 ppm of the compound for the same time
period (Stewart and Dodd. 1964).
Studies in animals confirm that dermal uptake of tetrachloroethylene following vapor
exposure is minimal when compared with pulmonary uptake (McDougal et al.. 1990; Tsuruta.
1989). whereas dermal uptake is greater following direct skin application (Jakobson et al.. 1982).
Notably, the conclusions of Bogen et al. (1992). based on the results of their study in hairless
guinea pigs, indicate that dermal absorption of tetrachloroethylene from contaminated water
supplies could be an important route of exposure for humans. These investigators estimated that
a standard 70-kg man with 80% of his body immersed in water would completely absorb the
amount of tetrachloroethylene in 2 L of that water.
3 2 DISTRIBUTION AND BODY BURDEN
Once absorbed, tetrachloroethylene is distributed by first-order diffusion processes to all
tissues in the mammalian body. The highest concentrations of tetrachloroethylene are found in
adipose tissue due to the lipophilicity of the compound (U.S. EPA. 1985b). Concentrations of
tetrachloroethylene reach higher levels in brain and liver than in many other tissues (Gamier et
al.. 1996; Levine et al.. 1981; Lukaszewski. 1979). Absolute tissue concentrations are directly
proportional to the body burden or exposure dose. Due to its lipid solubility, tetrachloroethylene
is also concentrated in milk, and it has been measured in human breast milk (NYSDOH. 2000;
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-2 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Schreiber. 1993. 1997; Schreiber et al.. 2002). Higher concentrations occur in milk having
higher fat content; e.g., a noticeable difference exists between the milk:blood partition
coefficients for rats (12) and for humans (Byczkowski and Fisher. 1994). reflecting the higher fat
content of rat milk. Tetrachloroethylene readily crosses both the blood:brain barrier and the
placenta. Partition coefficients for various tissues, relative to blood or air, have been reported by
several investigators (Byczkowski and Fisher. 1994; Dallas et al.. 1994a; Gearhart et al.. 1993;
Ward et al.. 1988). Section 3.5 presents examples of these.
Repeated daily inhalation exposures of human volunteers to tetrachloroethylene indicate
accumulation of the compound in the body, which is thought to be due to its high lipid solubility.
Because of its long residence time in adipose tissue, repeated daily exposure results in an
accumulated concentration; tetrachloroethylene from new exposures adds to the residual
concentration from previous exposures until steady state is reached. Blood levels of
tetrachloroethylene increase over several days with continued daily exposures. Following
cessation of these exposures, it is still present in the blood. Exhalation of the compound
continues over a number of days due to its slow release from the adipose tissue (Altmann et al..
1990; Skender et al.. 1991; Stewart et al.. 1977). For a given concentration in blood or air, the
half-time—the time necessary to equilibrate the adipose tissue to 50% of its final
concentration—is about 25 hours (Fernandez et al.. 1976; Monster. 1979). Therefore, during a
single 8-hour exposure, adipose tissue does not reach steady-state equilibrium.
Tetrachloroethylene uptake by fatty tissue during the working hours of the week is
countered by the elimination that occurs during nonexposure times of nights and weekends; thus,
for persons exposed to tetrachloroethylene on a 5-day-a-week work schedule, an equilibrium is
eventually established, but it requires a time period of 3-4 weeks of exposure for adipose tissue
to reach plateau concentrations (U.S. EPA. 1985b).
Animal studies provide clear evidence that tetrachloroethylene distributes widely to all
tissues of the body, readily crossing the blood:brain barrier and the placenta (Dallas et al.. 1994b;
Ghantous et al.. 1986; Savolainen etal.. 1977b; Schumann etal.. 1980). Following exposure of
rats to tetrachloroethylene, the compound has been measured in blood, fat, brain, lungs, liver,
kidneys, heart, and skeletal muscle (Dallas et al.. 1994b; Savolainen et al.. 1977b). The highest
tissue concentrations were found in adipose tissue (60 or more times blood level) and in the brain
and liver (4 and 5 times that found in the blood, respectively), as can be calculated from the rat
tissue-distribution data of (Dallas et al.. 1994b; Savolainen etal.. 1977b) found the concentration
of tetrachloroethylene in fat to be 9-18 times the concentrations found in nonfat tissues. Skeletal
muscle contained the lowest concentration. In one human fatality case, the concentration of
tetrachloroethylene in the brain was 120 times higher than concentrations measured in the lung.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-3 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
In another case, the concentrations in the liver were 8, 3.4, and 3.5 times higher, respectively,
than concentrations measured in the lung, kidney, and brain (Levine et al.. 1981).
3 3 METABOLISM
This section describes the metabolism of tetrachloroethylene, identifying metabolites
thought to be causally associated with toxic responses as well as those used to evaluate the flux
of parent compound through the known metabolic pathways. Sex- and species-dependent
differences in the metabolism of tetrachloroethylene and potential contributors to interindividual
differences are identified. See Section 4.9 for further discussion of how these factors affect
variability and susceptibility.
3.3.1. Introduction
The metabolism of tetrachloroethylene has been studied mostly in mice, rats, and humans
(for reviews, see Anders et al.. 1988; Dekant et al.. 1987; IARC. 1995; Lash and Parker. 2001;
U.S. EPA. 1985b. 1986a. 1991b). Tetrachloroethylene is metabolized in laboratory animals and
in humans through at least two distinct pathways: oxidative metabolism via the cytochrome P450
(CYP [also abbreviated as P450]) mixed-function oxidase system and glutathione (GSH)
conjugation followed by subsequent further biotransformation and processing, either through the
cysteine conjugate P-lyase pathway or by other enzymes including flavin-containing
monooxygenase 3 (FM03) and CYP3A (Anders et al.. 1988; Birner et al.. 1996; Costa and
Ivanetich. 1980; Daniel. 1963; Dekant etal.. 1987; 1989; Filser and Bolt. 1979; IARC. 1995;
Lash and Parker. 2001; Lash etal.. 1998; Pegg etal.. 1979; U.S. EPA. 1985b. 1991b; Volkel et
al.. 1998). The conjugative pathway is toxicologically significant because it yields relatively
potent toxic metabolites (Anders et al.. 1988; 1986a; Dekant et al.. 1986d; Lash and Parker.
2001; Vamvakas et al.. 1987; 1989a. b; 1989c; Werner et al.. 1996). Figure 3-1 depicts the
overall scheme of tetrachloroethylene metabolism. Known metabolites presented in this figure
are identified by an asterisk.
3.3.2. Extent of Metabolism
Studies in both animals and humans indicate that overall metabolism of
tetrachloroethylene is relatively limited—particularly at higher exposures (reviewed in Lash and
Parker. 2001; U.S. EPA. 1985b. 1991b). as evidenced by the high percentage of absorbed dose
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-4 DRAFT-DO NOT CITE OR QUOTE
-------
\
NH,+
CL C
FM03
P450
CL C
CL C
cl c
4* OH
[Reactive species]
CI, CH
P-lyase
CO, CO
[Reactive species]
Figure 3-1. Postulated scheme for the metabolism of tetrachloroethylene by
the cytochrome P450 (P450) oxidative pathway and glutathione ^-transferase
(GST)-mediated glutathione (GSH) conjugation pathway. PCE and identified
(*) urinary metabolites: (1) PCE, (2) PCE-Fe-0 intermediate, (3) trichloroacetyl
chloride, (4) trichloroacetic acid, (5) PCE oxide, (6) ethandioyl dichloride,
(7) oxalic acid, (8) ^-(l^^-trichlorovinyl) glutathione (TCVG),
(9) »Y-(1,2,2-trichlorovinyl)-A-cysteine (TCVC), (10) A-acetyl trichlorovinyl
cysteine (NAcTCVC), (11) dichloroacetic acid. Enzymes: cytochrome P450
(P450), GST, gamma-glutamyltransferase (GGT), dipeptidase (DP), P-lyase,
flavin mono-oxygenase-3 (FM03), A-acetyl transferase (NAT).
Sources: Adapted from Pegg et al. (1979). Costa and Ivanetich (1980). U.S. EPA (1985b). Dekant
et al. (1986d). Lash and Parker (2001). Yoshioka et al. (2002). Cliiu et al. (2007)
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/2011 3-5 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
excreted in the breath as the parent molecule (Boettner and Muranko. 1969; Bub en and
O'Flahertv. 1985; Chiu et al.. 2007; Daniel 1963; Essing et al.. 1973; Fernandez et al.. 1976;
Filser and Bolt 1979; Frantz and Watanabe. 1983; Ikeda and Ohtsuii. 1972; Monster et al.. 1983;
Monster. 1979; Monster et al.. 1979; Ohtsuki et al.. 1983; Pegg et al.. 1979; Schumann et al..
1980; Stewart et al.. 1970; Stewart et al.. 1961; Volkel et al.. 1998; Yllner. 1961). Because of its
high lipid solubility, tetrachloroethylene can be sequestered in fat and, thus, not all metabolism is
evident in short sampling time periods.
The extent of metabolism after inhalation exposure in humans has been estimated by
measuring trichloro-compounds excreted in the urine and exhalation of tetrachloroethylene in
expired air (Boettner and Muranko. 1969; Bolanowska and Golacka. 1972; Essing et al.. 1973;
Fernandez et al.. 1976; Ikeda et al.. 1972; May. 1976; 1983; Monster et al.. 1979; Monster and
Houtkooper. 1979; Stewart et al.. 1970; Stewart et al.. 1961). Several studies reported only
about 1-3% of the estimated amounts inhaled were metabolized to trichloroacetic acid (TCA)
and other chlorinated oxidation products, although additional tetrachloroethylene—as much as
20% or more of the dose—may be metabolized over a longer period (Bogen et al.. 1992; Bois et
al.. 1996; Monster et al.. 1979; U.S. EPA. 1985b. 1991b). For example, Chiu et al. (2007) noted
that although an average of 0.4% of tetrachloroethylene intake (1 ppm for 6 hours) was
recovered in urine as TCA, total recovery in urine and exhaled air accounted for on average only
82% of intake. This would imply that 18% were metabolized, but Chiu et al. (2007) noted
substantial uncertainty and variability in these calculations and concluded that they were
consistent with previous studies at higher exposures. Interestingly, Chiu et al. (2007) also noted
significant variability among the seven subjects and among the four occasions, contributing to
the uncertainty in measurements. A literature review published by Hattis et al. (1990) reported
estimates of the fraction of tetrachloroethylene metabolized at a low dose of 1 ppm to range from
2-86%. Based on data from Monster et al. (1979). Bois and colleagues (Bois et al.. 1996; Chiu
and Bois. 2006) used physiologically based pharmacokinetic (PBPK) modeling to predict that at
exposure levels near current occupational standards, a median of approximately 1.5% of inhaled
tetrachloroethylene would be metabolized, whereas, at ambient air levels (0.001 ppm), the
median estimate would be considerably higher (23—36%).
The extent of metabolism in animals has been estimated by conducting excretion-balance
studies using isotopically labeled tetrachloroethylene. In rodents, 2—88% of the dose was
metabolized, depending on dose level and species: the higher the dose, the smaller the percentage
metabolized. Rats metabolized a lower percentage of a given tetrachloroethylene body burden
than did mice (Daniel. 1963; Filser and Bolt. 1979; Frantz and Watanabe. 1983; Pegg et al..
1979; Schumann et al.. 1980; Yllner. 1961). As an example, using data from the Pegg et al.
(1979) and Schumann et al. (1980) studies in rats, U.S. Environmental Protection Agency (EPA)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-6 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
calculated that the percentage of body burdens excreted were unchanged following exposure to
10 and 600 ppm for 6 hours; they were 68 and 99%, respectively (U.S. EPA. 1985b). For
comparison, studies in mice exposed to 10 ppm for 6 hours found pulmonary excretion of only
12%, whereas 83% of the tetrachloroethylene was excreted by the pulmonary route for a body
burden of about 11 mg from oral administration (U.S. EPA. 1985b). As body burden is
increased, the proportion of tetrachloroethylene excreted unchanged increases, and the
percentage metabolized decreases. Urinary excretion data from studies by Filser and Bolt (1979)
and Buben and O'Flaherty (1985) suggest that metabolism of tetrachloroethylene is greater in
mice than in rats.
3.3.3. Pathways of Metabolism
The two known biotransformation pathways for tetrachloroethylene metabolism are
(1) oxidation by CYP enzymes and (2) conjugation with GSH followed by further processing of
the conjugate through various pathway bifurcation branches. As shown in Figure 3-1, the initial
step in the metabolism of tetrachloroethylene is formation of an Fe-0 intermediate for the
oxidative pathway or conjugation with GSH for the secondary pathway (Costa and Ivanetich.
1980; 1998; 1987; Dekantetal.. 1986a; Lash and Parker. 2001; Lash et al.. 1998; Miller and
Guengerich. 1982. 1983; Yoshioka et al.. 2002). It is possible that other yet unrecognized
pathways for tetrachloroethylene could exist in humans (Bois et al.. 1996; Monster et al.. 1979;
Sakamoto. 1976; U.S. EPA. 1985b. 1991b).
3.3.3.1. Cytochrome P450-Dependent Oxidation
3.3.3.1.1. Oxidative metabolites
In vivo, the major excretory metabolite of the oxidative pathway, TCA, is excreted in the
urine of all species tested (Birner et al.. 1996; Daniel. 1963; Dekant et al.. 1987; Leibman and
Ortiz. 1970. 1977; Ohtsuki et al.. 1983; Volkel et al.. 1998; Yllner. 1961). Oxalic acid has been
reported to be a relatively major urinary metabolite in rats (Dmitrieva. 1967; Pegg et al.. 1979).
Oxalic acid might either arise from action of microsomal epoxide hydrase on the epoxide
intermediate or may be a separate product from the initial Fe-0 intermediate. The oxalate
metabolite excretory product may also be derived from dichloroacetic acid (DCA) or
monochloroacetic acid (Tong et al.. 1998a. b). Pulmonary excretion of carbon dioxide (CO2)
amounting to <10% of the administered dose has been identified in exhaled breath from rodents
exposed to 14C-labeled tetrachloroethylene (Frantz and Watanabe. 1983; Pegg et al.. 1979;
Schumann etal.. 1980). accounting for less than either exhaled tetrachloroethylene or urinary
metabolites.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-7 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Trichloroethanol (TCOH) has been detected in the urine of subjects exposed to
tetrachloroethylene in some studies (Birner et al.. 1996; Ikeda and Ohtsuii. 1972; Ikeda et al..
1972; Monster et al.. 1983; Ogata et al.. 1962; 1971; Schreiber et al.. 2002; Tanaka and Ikeda.
1968; Weichardt and Lindner. 1975). but it could not be identified by others (Buben and
O'Flahertv. 1985; Chiu et al.. 2007; Costa and Ivanetich. 1980; Daniel. 1963; Fernandez et al..
1976; Hake and Stewart. 1977; Monster et al.. 1979; Volkel et al.. 1998; Yllner. 1961). Most of
the studies reporting TCOH have involved occupational or environmental exposures in which
there may be simultaneous exposure to trichloroethylene, for which TCOH is a major urinary
metabolite. In vitro, TCA—and, to a lesser extent—oxalic acid, but not TCOH, are detected in
incubations of tetrachloroethylene in rat microsomal protein (e.g.. Yoshioka et al.. 2002). Thus,
it appears likely that the reports of TCOH following tetrachloroethylene exposure may have been
artifacts of the analytical methodology used or of simultaneous trichloroethylene exposure.
Because TCOH is clearly not a significant metabolite for tetrachloroethylene, very little, if any,
TCA produced from tetrachloroethylene metabolism is likely to come through chloral, either
directly or indirectly through TCOH (Lash and Parker. 2001).
It was initially proposed that the first step in the oxidation of tetrachloroethylene is
hypothesized to yield 1,1,2,2-tetrachloroethylene oxide, a relatively unstable epoxide (Costa and
Ivanetich. 1980; Miller and Guengerich. 1982. 1983). Although an initial epoxide metabolite has
not been unequivocally demonstrated for tetrachloroethylene, evidence for this epoxide does
exist. The epoxide has been chemically synthesized (Bonse et al.. 1975; Frankel et al.. 1957)
(Kline et al.. 1978). The potential fates of tetrachloroethylene epoxide include trichloroacetyl
chloride, oxalate dichloride through tetrachloroethylene glycol, trichloroacetyl aminoethanol,
and, possibly, chloral hydrate (in equilibrium with chloral) (Bonse and Henschler. 1976;
Henschler and Bonse. 1977; Pegg et al.. 1979; U.S. EPA. 1985b. 1986a).
However, recent data (Yoshioka et al.. 2002) favor the hypothesis that the epoxide is not
an obligatory intermediate to formation of trichloroacetyl chloride. In particular, the pattern of
products of tetrachloroethylene oxide hydrolysis reported by Yoshioka et al. (2002) is dominated
by carbon monoxide (CO) and carbon dioxide (CO2), which differs markedly from the products
of oxidation in vivo or in vitro. Because TCA is believed to be derived primarily from
trichloroacetyl chloride (through hydrolysis or through reaction with amino groups of cellular
proteins), this would favor the hypothesis that the epoxide is a minor product of
tetrachloroethylene oxidation. Instead, the Fe-0 intermediate is postulated to collapse via
chlorine migration to yield predominantly trichloroacetyl chloride (Yoshioka et al.. 2002).
DCA has been identified as a tetrachloroethylene urinary metabolite (Dekant et al.. 1987;
Volkel et al.. 1998; Yllner. 1961). and may arise as a product of further metabolism of TCA or as
a result of P-lyase bioactivation of GSH conjugation metabolites. The major organ site of DCA
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-8 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
production is likely to differ for each pathway, with DCA arising from oxidative metabolism
primarily in the liver and from GSH-dependent metabolism products mostly in the kidney.
Dechlorination of TCA to DCA is catalyzed by gut contents (ingested food and bacteria) of the
rat and mouse (Moghaddam et al.. 1996); isolated mouse microflora have been shown to convert
TCA to DCA (Moghaddam et al.. 1997). However, data indicate that this does not contribute to
DCA detected systemically after trichloroethylene exposure, and a similar conclusion is
reasonable for tetrachloroethylene, given the lower rate of formation of TCA from
tetrachloroethylene as compared to trichloroethylene. In addition, data from trichloroethylene
suggest that for that compound, DCA formation is likely dominated by hydrolysis of
dichloroacetyl chloride—rather than dechlorination of TCA. As compared to tetrachloroethylene
exposure, trichloroethylene exposure leads to higher amounts of TCA, in conjunction with the
lower amounts of DCA, detectable in blood or urine. This is inconsistent with dechlorination of
TCA being the origin of DCA detected in urine after tetrachloroethylene exposure, and supports
the hypothesis that DCA is derived predominantly from GSH conjugation of tetrachloroethylene.
3.3.3.1.2. Species differences
Although thought to be qualitatively similar, there are clear differences among species in
the quantitative aspects of tetrachloroethylene metabolism (Ikeda and Ohtsuii. 1972; Lash and
Parker. 2001; Schumann et al.. 1980; U.S. EPA. 1991b; Volkel et al.. 1998). These differences
are in the relative yields and kinetic behavior of metabolites (Green et al.. 1990; Ohtsuki et al..
1983; U.S. EPA. 1985b. 1991b; Volkel et al.. 1998). Rodents and humans differ in relative rates
of tetrachloroethylene metabolism in key target organs, in the doses at which saturation of
metabolism occurs, and in the half-times in the body.
The rate of metabolism of tetrachloroethylene is faster in rodents than in humans, and
higher metabolite concentrations in blood are obtained in rodents as compared with humans.
The higher blood levels of metabolites in rodents are particularly noticeable at the higher
tetrachloroethylene exposure levels because saturation is approached at lower exposure levels in
humans than in rodents. The half-life in the body of these metabolites is, however, noticeably
longer for humans than for rodents (144 hours in humans vs. approximately 10 hours or less in
rodents (see U.S. EPA. 1985b). It is for this reason that examinations of tetrachloroethylene
concentration and toxicity associations must reflect both blood concentration and time-integrated
dose metrics such as area-under-the-curve (AUC).
A study of species differences in tetrachloroethylene metabolism conducted by Dekant
and colleagues is presented in (Volkel et al.. 1998). These investigators compared both oxidative
and GSH-dependent metabolism in rats and humans exposed for 6 hours to 10-, 20-, or 40-ppm
tetrachloroethylene by inhalation. Rats were also exposed to 400-ppm concentrations. TCA was
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-9 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
the major urinary excretion product in both species; however, the elimination half-time was more
than four times slower in humans than in rats. Blood plasma concentrations of the metabolite
were higher (three-eightfold, depending on the dose) in rats than in humans exposed to identical
air concentration levels. These observations are in agreement with metabolic rates in general,
which are higher in mice than in rats; rats, in turn, have higher metabolic rates than do larger
animals, including humans. Dekant and his coworkers also reported urinary excretion of DCA
by rats—but not humans. They concluded most of the DCA resulted from GSH-dependent
metabolism. DCA, however, is further metabolized by P450 enzymes, which, in turn, limits its
detectability in urine.
3.3.3.1.3. Cytochrome P450 (CYP) isoforms and genetic polymorphisms
Oxidative metabolism of tetrachloroethylene, irrespective of the route of administration,
occurs predominantly in the liver but also at other sites. For example, the kidneys exhibit
cytochrome P450 enzyme activities, mostly in the proximal tubules, although total activity is
markedly less than in the liver (Lash and Parker. 2001; Lash et al.. 2001). CYP enzymes
occurring in other extrahepatic tissues—brain and lungs, for example—may also contribute to
oxidative metabolism of tetrachloroethylene.
Relatively few studies provide information about which specific CYP isoforms play a
role in tetrachloroethylene oxidative metabolism. CYP2E1 is presumed to have an important
role in tetrachloroethylene metabolism (Lash and Parker. 2001); however, the chemical-specific
related data are too sparse to provide strong support for this assumption (Dohertv et al.. 1996).
CYP2B1/2 may also be important for the metabolism of tetrachloroethylene. CYP3A
isoenzymes may contribute to the generation of reactive sulfoxides from metabolites of the GSH
pathway (see below). Costa and Ivanetich (1980) showed increased hepatic metabolism
following treatment with agents now known to induce these isoenzymes specifically.
Genetic polymorphisms are DNA sequence variations that result in changes in protein
sequence of an enzyme that can alter the enzyme's ability to catalyze a reaction or alter the
expression of an allele. Polymorphisms are known for most of the CYP enzymes including
CYP2E1 (Hu et al.. 1999; McCarver et al.. 1998) and CYP3A4 (Sata et al.. 2000; Westlind et al..
1999).
Metabolism of tetrachloroethylene to its putative epoxide is likely affected by CYP
enzymes. The metabolism of the putative metabolite chloral hydrate to TCOH and TCA may be
catalyzed by both alcohol dehydrogenase and CYP2E1. Oxidation of TCOH is catalyzed by
P450 enzymes, with CYP2E1 the likely predominant isoform involved, although other
isoenzymes may also play a role— even substituting for CYP2E1 in processing
tetrachloroethylene. The rat kidney expresses CYP2E1 (Cummings et al.. 1999; Speerschneider
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-10 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
andDekant. 1995). although the human kidney has not been shown to do so (Ametetal.. 1997;
Cummings et al.. 2000a). Therefore, renal CYP metabolism by this isoform in the rat kidney
would be relevant only insofar as the involvement of other isoenzymes in metabolizing
tetrachloroethylene via this route.
3.3.3.2. Glutathione (GSH) Conjugation Pathway
The GSH conjugation pathway was recognized much later than was the oxidative
pathway, yet it may be toxicologically influential (IARC. 1995; Lash and Parker. 2001; U.S.
EPA. 1991b). Similar to trichloroethylene, GSH conjugation of tetrachloroethylene is associated
with renal toxicity (Anders et al.. 1988; Dekant et al.. 1989; IARC. 1995; Lash et al.. 2000; Lash
and Parker. 2001; U.S. EPA. 1991b).
3.3.3.2.1. Glutathione (GSH) conjugation metabolites
The initial conjugation of tetrachloroethylene with GSH occurs mainly in the liver
(Dekant et al.. 1987; Green et al.. 1990; Vamvakas et al.. 1987; 1989b). with transport of the
conjugate and its cysteine counterpart to the kidney target organ for further processing. This first
step also occurs within the kidney (Lash et al.. 1998). As shown in Figure 3-1,
tetrachloroethylene is initially conjugated with GSH to form »Y-(1,2,2-trichlorovinyl) glutathione
(TCVG). This reaction, which is catalyzed by the GSH-.Y-transferase (GSTs) enzymes, a group
of enzyme isoforms, was traditionally considered to be a detoxification reaction, leading to more
water-soluble compounds that are more readily excreted. In many cases, however, as with
certain halogenated alkanes and alkenes such as tetrachloroethylene, GSH conjugation can be
important for bioactivation. TCVG is then processed through the cysteinylglycine conjugate
£-(1,2,2 trichlorovinyl)-A-cysteinylglycine to »Y-(1,2,2-trichlorovinyl)-A-cysteine (trichlorovinyl
cysteine, or TCVC) by the enzymatic removal of glutamyl and glycine residues by gamma-
glutamyltransferase (GGT) and various membrane-bound dipeptidases known as
cysteinylglycine dipeptidase (reviewed by Anders et al.. 1988; Dekant et al.. 1989; Lash and
Parker. 2001; U.S. EPA. 1991b). These enzymes reside in tissues other than the kidneys (e.g.,
the brain), indicating a potential for toxic reactive metabolite formation in those tissues as well.
Conversion of TCVG to TCVC by these cleavage enzymes leads to a critical bifurcation point of
the GSH pathway because the TCVC may be processed by certain enzymes to yield reactive,
toxic chemical species, although it may be metabolized via a different route to yield an excretory
product (Lash and Parker. 2001).
Importantly, the TCVC metabolite may also act as a substrate for renal P-lyases (Anders
et al.. 1988; Dekant et al.. 1988 reviewed by; Dekant etal.. 1989; Lash et al.. 2000; Lash and
Parker. 2001; U.S. EPA. 1991b). Renal P-lyases are known to cleave TCVC to yield an unstable
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-11 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
thiol, 1,2,2-trichlorovinylthiol, that may give rise to cytotoxic and mutagenic reactive chemical
species that can form covalent adducts with cellular nucleophiles, including DNA and proteins
(Volkel et al.. 1999). In addition, DCA is a downstream urinary excretion product of P-lyase
bioactivation of TCVC, and has been detected in urine of rats exposed to tetrachloroethylene
(Volkel et al.. 1998). P-lyases are a family of pyridoxal phosphate-containing enzymes that are
located in several tissues besides the kidneys, including liver and brain, and in intestinal flora,
although their substrate specificities may vary. Hepatic P-lyase is distinct from renal P-lyase and
has not been found to have a role in TCVC metabolism. P-lyase activity is higher in the rat
kidney than in the human kidney (Cooper. 1994; Lash et al.. 1990). which is consistent with
overall metabolic rates being higher in smaller versus larger mammalian species.
In addition to activation by P-lyases, TCVC may be metabolized by a flavin-containing
monooxygenase, FM03, or CYP enzymes to TCVC sulfoxide (TCVCSO), another reactive
metabolite (Ripp et al.. 1997). TCVCSO is a more potent nephrotoxicant than TCVC (Elfarra
and Krause. 2007). These TCVC sulfoxide and P-lyase cleavage products rearrange, forming a
thioketene (Dekant et al.. 1988; Ripp et al.. 1997). which is a potent acylating agent capable of
binding to cellular macromolecules, including DNA (Birner et al.. 1996; Pahleretal.. 1999b)
(1999a; Volkel et al.. 1999). Interestingly, the thioketene can degrade to form DCA, potentially
making this metabolite a product of both tetrachloroethylene metabolism pathways (Dekant et
al.. 1987; Volkel etal.. 1998).
In addition to P-lyase and FM03/CYP activation of TCVC, reactive sulfoxides can also
be produced by further CYP3A metabolism of iV-acetyl-5'-(l,2,2 -trichlorovinyl)-/.-cysteine
(NAcTCVC; Werner et al.. 1996). This tetrachloroethylene-derived mercapturate metabolite
results from TCVC being acetylated via a reversible reaction (Bartels. 1994; Birner et al.. 1996;
Duffel and Jakobv. 1982). iV-acetyl-5'-(l,2,2 trichlorovinyl)-/.-cysteine may be excreted in the
urine. However, in addition to its activation to sulfoxides via CYP3 A metabolism, it can also be
transported to other organs and deacetylated intracellularly, regenerating the cysteine conjugate
TCVC and, thus, making it available to other enzymes for activation (Uttamsingh et al.. 1998).
It should be noted that the A'-acetylation reaction is catalyzed by an enzyme located in the
endoplasmic reticulum that is distinct from the cytosolic enzymes that are polymorphic in
humans (Lash and Parker. 2001).
Some controversy surrounds the importance of the GSH conjugation pathway with regard
to tetrachloroethylene metabolism in humans. As noted above, the GSH pathway for
tetrachloroethylene was originally demonstrated only in rodents, and interpretation of the
then-existing data led some scientists to conclude that the pathway was not operative in humans
(Green et al.. 1990; U.S. EPA. 1991b). More recent data clearly demonstrate the existence of the
pathway in humans (Birner et al.. 1996; Schreiber et al.. 2002; Volkel et al.. 1998).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-12 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Quantitatively, urinary mercapturates comprise from 1% to as little as 0.03% of total recovered
urinary metabolites, but this does not reflect the total flux through the GSH pathway; rather it
reflects only the portion that is excreted. In particular, the amount of the mercapturate product
excreted in the urine also does not reflect the amount of the more important portion that is
converted to toxic by-products through further metabolism. However, there are discrepancies
regarding reported rates of tetrachloroethylene GSH metabolism (Dekant et al.. 1998; Green et
al.. 1990; Lash and Parker. 2001; Lash et al.. 2007; Lash et al.. 1998). These differences may be
due, in part, to different chemical assay methodology or to problems resulting from the stability
of the chemical product being measured or both (Lash and Parker. 2001). Some of the published
in vitro findings concerning TCVG production would predict similar susceptibility for humans
than for rodents with regard to renal toxicity (Lash et al.. 1998). while others appear to predict
much less susceptibility (Dekant et al.. 1998; Green et al.. 1990).
3.3.3.2.2. Species differences in gamma-glutamyltransferase (GGT) and p-lyase
Species-dependent differences in GGT (Hinchman and Ballatori. 1990) also are not
thought to be limiting because renal activity is present at high enough levels even in humans so
that GGT activity is not the rate-limiting step in the metabolism. Species-dependent differences
in this enzyme (described below) would have only a very small quantitative effect on the overall
metabolism of TCVG and other similar GSH conjugates. Species differences in GGT activities,
therefore, would not have a major role in species differences in renal toxicity (Lash and Parker.
2001) in affecting transformation of TCVG to TCVC, and, thus, should not be important to
differences in susceptibility to tetrachloroethylene-induced renal toxicity.
GGT is the only enzyme that can split the gamma-glutamyl bond in the GSH conjugates
to form cysteine conjugates (Lash and Parker. 2001). It is this reaction that creates TCVC, the
substrate for the enzymes that generate the toxic metabolites. Therefore, the distribution of GGT
is important. Renal proximal tubular cells have the highest activities of GGT of all tissues,
although GGT activity also occurs in the liver, and the kidney-to-liver ratio of this enzyme varies
among species. In the rat, the specific activity ratio is 875 (Hinchman and Ballatori. 1990). The
ratio is lower in other species that have been studied. The tissue distribution and relative activity
have not been fully studied in humans, but it is known that GGT activity is considerably higher
in the human liver than in the rodent liver (Lash and Parker. 2001). The kidney-to-liver ratio of
GGT for humans is thought to be closer to those of pigs (2) and Macaques (5) than to those of
rats or mice (423). For this reason, use of a rodent model for the processing of the
tetrachloroethylene GSH conjugate to the corresponding cysteine conjugate would overestimate
the contribution of the kidneys and underestimate the contribution of the liver in cleaving TCVG
to TCVC. Even so, the liver excretes most of the cysteine conjugates such as TCVC into the bile
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-13 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
or plasma, where it is cycled to the kidneys and taken up into renal epithelial cells. So, the
TCVC will still end up in the kidneys.
The P-lyase enzyme is among the most important activator of toxic products in the
conjugation pathway—a fact particularly well documented in the kidney. There are some data,
however, that indicate that renal P-lyase-dependent metabolism is greater in rats than in mice or
in humans and greater in male than in female rats (Green et al.. 1990; Lash et al.. 1990; Volkel et
al.. 1998). This is not entirely in keeping with metabolic rates in general, which are higher in
mice than in rats, and rats, in turn, have higher metabolic rates than do larger animals, including
humans. Studies that measured only cytoplasmic P-lyase activity did not consider the
importance of mitochondrial P-lyase activity, which may be key to tetrachloroethylene
metabolite toxicity (Lash et al.. 2001).
In contrast, it must also be noted that species comparisons of tetrachloroethylene
metabolism in chronic exposures on a surface area- or metabolic-rate basis rather than on a direct
body-weight basis, particularly when including the total AUC for amount metabolized, indicate
that metabolite production in rats and humans may not differ significantly (Calabrese. 1983;
Rhomberg. 1992; U.S. EPA. 1986a). The fact is that metabolic rates and the amounts
metabolized are not the same thing. Metabolic rates are always faster in smaller species. Total
AUC may or may not be similar among species. Even if AUC is the same, the peak blood levels
may differ greatly from species to species. In other words, the pharmacokinetics are not the
same.
The higher percentage of mercapturate found in rat versus human urine does not indicate
a higher level of production of toxic products in the rat, because excreted mercapturate allows no
estimate of the amount of TCVC or A-acetyl TCVC being processed through alternate routes
(Lash and Parker. 2001). The relatively higher percentage of DCA in the urine may, however,
indicate relatively higher P-lyase enzyme activity and higher thioketene production in rats if the
DCA is indeed largely the product of the GST pathway rather than the oxidative pathway
(Volkel et al.. 1998). It is not known whether sex-dependent variation of P-lyase activity exists
in humans as it does in rats (Volkel et al.. 1998).
And finally, it is important to note that because the enzymes involved in this activation
pathway are also present in other tissues (Alberati-Giani et al.. 1995; Dohn and Anders. 1982;
Larsen. 1985; Larsen and Stevens. 1986; Malherbe et al.. 1995; Stevens and Jakobv. 1983;
Stevens. 1985; Tateishi et al.. 1978; 1986; Tomisawa et al.. 1984). there exists a potential for
formation of the reactive metabolites at sites other than the kidney, e.g., in the brain. In
one carcinogenicity bioassay of tetrachloroethylene, a biologically significant elevation of
gliomas in the rat brain was reported (NTP. 1986b). Whether or not toxic metabolites resulting
from P-lyase activity in the brain play a role in the development of the gliomas in the rat has not
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-14 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
been studied. The possibility that such tetrachloroethylene metabolites could be involved in the
mode of tumorigenic action producing gliomas is not unrealistic.
3.3.3.2.2.1. Glutathione S-transferase (GST) isoenzymes/polymorphisms
GSTs are a family of isoenzymes (Mannervik. 1985). found in cytoplasm. A distinct
microsomal GST isoenzyme also exists in most mammalian tissues (Otieno et al.. 1997).
Although GST activity occurs in most cell types, the liver is by far the predominant site of GSH
conjugation. GSTa, designated as GSTA in humans, is the predominate isoenzyme expressed in
normal kidney from rodents and humans (Campbell et al.. 1991; Cummings et al.. 2000b;
Mitchell et al.. 1997; Overbv et al.. 1994; Rodilla et al.. 1998). Available data thus far do not
indicate that variability in activity of this isoenzyme is important to differences in individual
susceptibility to toxicity. GSTS (GSTZ) catalyzes the oxidative metabolism of DCA to
glyoxylate (Board et al.. 1997; Tong et al.. 1998a. b), however, the tetrachloroethylene
metabolite DCA has been shown to be a potent, irreversible inhibitor of GSTZ activity (Tzeng et
al.. 2000Y
There are five human polymorphic variants of this GSTZ isoenzyme (Tzeng et al.. 2000;
U.S. EPA. 1998a). These genetic polymorphisms may influence tetrachloroethylene metabolism
although human data regarding this hypothesis are lacking. There are some species differences
in the other three cytoplasmic GSTs relevant to liver and kidney. GSTP expression is the most
variable and appears to be polymorphic in humans (Rodilla et al.. 1998). It has been found in rat
liver (Cummings et al.. 1999) but only in biliary ducts in humans (Terrier et al.. 1990) et
(Campbell etal.. 1991). GSTtc (GSTP) has been detected within the human kidney in various
cell types (Terrier et al.. 1990) but has not been isolated from rat kidney cells (Cummings et al..
1999). although GSTP has also been detected in the rabbit kidney (Cummings et al.. 1999).
Two homodimeric GST9 (GSTT) isoenzymes have been identified in the human kidney
(Cummings et al.. 2000a; Veitch et al.. 1997). GSTT has been detected in rat and mouse liver
and in mouse but not rat kidney (Cummings et al.. 1999; Quondamatteo et al.. 1998). GST|i
(GSTM) has been detected in rat kidney distal tubule cells (Cummings et al.. 2000b) and in
mouse and rabbit liver and kidney (Mitchell et al.. 1997; Overbv et al.. 1994)—but it was not
detected in human kidney (Cummings et al.. 2000a). It is not clear just how the differences in
these isoenzymes are related to species differences in tetrachloroethylene toxicity because the
isoenzyme specificity and reaction rates have not yet been studied with regard to
tetrachloroethylene (Lash and Parker. 2001).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-15 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
3.3.3.3. Relative Roles of the Cytochrome P450 (CYP) and Glutathione (GSH) Pathways
Although it is clear that at laboratory and occupational exposures, the oxidative CYP
pathway is quantitatively more important than the GSH conjugation pathway, the interorgan
patterns for some of the intermediate metabolites, as well as the relative toxicity of certain key
metabolites generated from these pathways, influence the relative importance of the
two pathways in determining toxicity. It is still not certain which metabolites, alone or in
combination, are explicitly responsible for specific tetrachloroethylene toxicities, and it is likely
that different metabolites contribute to toxicity at different target sites. In general, CYP
metabolism is associated with tetrachloroethylene-induced liver toxicity, whereas GSH
conjugation followed by further processing by P-lyase and other enzymes is associated with
tetrachloroethylene-induced renal toxicity. There is a possibility that P-lyase products could
contribute to toxicity in the brain, for example, and be a factor in the gliomas observed in rats.
The parent compound, itself, is also likely to be a contributing factor to tetrachloroethylene
neurotoxicity, particularly central nervous system effects.
Data from experiments designed to assess the effects of enzyme modulation suggest
competition between the two pathways (Dekant et al.. 1987; Lash et al.. 1999; Lash et al.. 2001;
Volkel et al.. 1998). Other data show relatively low urinary excretion of mercapturates as
compared to CYP-derived products. On the basis of these findings, some researchers have
concluded that there is a lack of toxicological significance for the low-affinity, low-activity GSH
pathway except when the high-affinity CYP pathway approaches saturation (Green et al.. 1990)
(1997; Volkel et al.. 1998). However, this conclusion does not consider the relative toxicological
potency or chemical reactivity of the metabolites from the two pathways or the fact that the
amount of mercapturate excreted is not a valid quantitative indicator of the extent of conjugative
pathway metabolism (Lash and Parker. 2001).
Specific tetrachloroethylene metabolites are known to be associated with certain
toxicities when they are administered directly. Exactly how these same compounds, as
metabolites of tetrachloroethylene, contribute to the various toxicities associated with exposure
to the parent compound is not yet well understood.
3.3.4. Susceptibility
Differences in enzyme activity may lead to variations among individuals in their
sensitivity to tetrachloroethylene toxicities. A 10-fold difference in CYP enzyme metabolic
capacity among humans is a generally accepted norm. Although individual variations in the
CYP2E1 enzymatic activity as high as 20-50-fold have been reported (Lieber. 1997; Stephens et
al.. 1994; Yoo et al.. 1988). these in vitro measurements would be taken out of physiological
context if used to estimate in vivo interindividual variations. Measurable and obvious
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-16 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
differences in CYP enzymatic activity are observed among various ethnic groups and age groups
(Goldstein et al.. 1969; Raunio et al.. 1995). No chemical-specific data regarding the manner in
which CYP enzyme isoforms might affect susceptibility to adverse effects are available for
tetrachl oroethy lene.
Diagnosis of polymorphisms in carcinogen-activating and -inactivating enzymes and
cancer susceptibility have been noted (Raucv. 1995; Stephens et al.. 1994; Yoo et al.. 1988).
Potential strain-dependent differences among rodents and human genetic polymorphisms in
metabolizing enzymes involved in biotransformation of tetrachl oroethy lene are now known to
exist. Whether CYP polymorphisms could account for interindividual variation in
tetrachloroethylene metabolism among humans—and, thus, differences in susceptibility to
tetrachloroethylene-induced toxicities—is not known.
The GSTs involved in tetrachloroethylene metabolism are described in Section 3.3.3.2.
A potential exists for interindividual variation to occur in tetrachloroethylene metabolism as a
result of variability in GST enzyme expression. It is important to note that GST polymorphism
has been associated with increased risk of kidney cancer in people exposed to trichl oroethy lene
(Moore and Harrington-Brock. 2000). There are no direct, chemical-distinctive data with regard
to the specific isoenzyme family responsible for TCVG formation in the metabolism of
tetrachloroethylene. There are species-dependent differences as to which isozymes occur in liver
and kidney, although it is unknown how the various enzymes are related to differences in the
metabolism of tetrachloroethylene. The compound is likely a good substrate for GSTA (Lash
and Parker. 2001). GSTT and GSTP occur in human kidney, as does GSTA, the primary
isozyme in human kidney, meaning that there is a potential for differences in the ability to
produce TCVG. GSTZ transforms the tetrachloroethylene metabolite DCA. DCA has also been
shown to have a potent irreversible inhibitory effect on the GSTZ isoenzyme, which is known to
have at least four polymorphic variations.
Inhibition or induction of the enzymes responsible for tetrachloroethylene metabolism
can, and likely does, alter susceptibility to toxicity (IARC. 1995; Lash and Parker. 2001; U.S.
EPA. 1985b). Numerous environmental pollutants and therapeutic agents alike have the
potential to induce or inhibit tetrachloroethylene-metabolizing enzymes. For example,
tetrachloroethylene metabolism is increased by inducers of cytochrome CYP enzymes such as
toluene, phenobarbital, and pregnenolone-16a-carbonitrile, whereas CYP inhibitors such as SKF
525A, metyrapone, and carbon monoxide decrease tetrachloroethylene metabolism (Costa and
Ivanetich. 1980; Ikeda and Imamura. 1973; Moslen et al.. 1977). Chronic exposure to
tetrachloroethylene has been shown to cause self-induction of metabolism (Kaemmerer et al..
1982; Savolainen et al.. 1977b; Vainio et al.. 1976). Other factors, such as health status or
disease state, activity patterns, or concomitant exposure to other chemicals, can potentially
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-17 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
influence tetrachloroethylene metabolism and its resulting toxicity. Section 4.9 addresses
coexposures and cumulative risk in greater detail.
3.3.5. Comparison of Tetrachloroethylene Metabolism with Trichloroethylene Metabolism
3.3.5.1. Extent of Metabolism
The available data indicate that, overall, tetrachloroethylene is less extensively
metabolized than is the closely related chemical, trichloroethylene. The difference is due to the
fact that a lower fraction of a tetrachloroethylene dose is metabolized via the major oxidative
CYP pathway when compared with an equivalent dose of the trichloroethylene congener (Lash
and Parker. 2001; Ohtsuki et al.. 1983; Volkel et al.. 1998). For example, in balance studies of
humans, only about 1-3% of the estimated amounts of tetrachloroethylene inhaled were shown
to be metabolized to TCA and other chlorinated metabolites, although these studies fail to
account for total dose (see above for further discussion). These amounts can be compared to the
40-75% of trichloroethylene shown to be metabolized in various human balance studies similar
to the ones conducted for tetrachloroethylene (U.S. EPA. 1985b).
Because of its higher lipid solubility, tetrachloroethylene may appear to be less well
metabolized than trichloroethylene, at least to a certain degree, simply because it is more slowly
metabolized due to fat sequestration. However, the animal data from studies of the
two compounds provide results similar to those of the human studies regarding the relative extent
of metabolism. For example, using data from laboratory animal studies of tetrachloroethylene
(Pegg et al.. 1979; Schumann et al.. 1980). EPA reported the percentage of tetrachloroethylene
body burdens excreted as unchanged parent compound following exposure to 10 and 600 ppm
for 6 hours to be 68 and 99%, respectively (U.S. EPA. 1985b). By comparison, rats and mice
exposed to equivalent 10- and 600-ppm trichloroethylene doses (Stott et al.. 1982) metabolized a
higher percentage of this compound, with mice metabolizing essentially all of the inhaled dose
and rats metabolizing 98 and 79% of the low and high inhaled doses, respectively.
Saturation of metabolism occurs at a higher dose for trichloroethylene than for
tetrachloroethylene; thus, at certain dose levels, the differences in the amounts of the
two compounds metabolized are relatively greater than at other dose levels. Tetrachloroethylene
appears to be a lower-affinity substrate for CYP enzymes than trichloroethylene (Ohtsuki et al..
1983; Volkel et al.. 1998). In vitro, the Michaelis-Menten constant (Km) value for
tetrachloroethylene is reported to be higher than the Km value for trichloroethylene (Lipscomb et
al.. 1998).
Both tetrachloroethylene and trichloroethylene are liver toxicants and cause liver
hepatocellular carcinomas in mice. The liver toxicity, including carcinogenicity, of these
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-18 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
compounds is thought to be due to metabolites. It is interesting to note that although
trichloroethylene appears to be more extensively metabolized—due to greater CYP metabolism
in the liver—the relative cancer potency for liver tumors is similar for the two compounds.
Comparisons of potencies for kidney cancer are more difficult because there is a lack of studies
with both compounds using comparable species/strains and routes of exposure.
3.3.5.2. Cytochrome P450 (CYP)-Mediated Oxidation
TCA, DCA, chloral, and TCOH are reported biotransformation products of both
tetrachloroethylene and trichloroethylene; however, the relative amounts produced and the
precursor intermediates are different for the two compounds. TCA is the major urinary
metabolite for tetrachloroethylene, and it is also an excretion product of trichloroethylene,
whereas TCOH is the major trichloroethylene urinary excretion product. As discussed
previously in Section 3.3.3.1, the formation of chloral and TCOH in metabolism of
tetrachloroethylene is not likely to be significant. Therefore, very little, if any, TCA produced
from tetrachloroethylene metabolism comes through chloral—either directly or indirectly
through TCOH (Lash and Parker. 2001). The TCA from tetrachloroethylene comes through
trichloroacetyl chloride, possibly via the epoxide, but more likely directly from chlorine
migration of the Fe-0 intermediate. On the other hand, the TCA produced from
trichloroethylene metabolism is thought to come through chloral—both directly and through
TCOH enterohepatic circulation (Lash et al.. 2000).
DCA is a biotransformation product of both tetrachloroethylene and trichloroethylene,
although it is believed that a greater portion of DCA coming from tetrachloroethylene
metabolism does not arise from CYP metabolism, but rather results from further processing of
TCVC, whereas the DCA coming from trichloroethylene metabolism results from CYP
oxidation.
Quantitatively, the liver is by far the predominant site of tetrachloroethylene and
trichloroethylene oxidative metabolism; although most other tissues contain the CYPs that could
conceivably metabolize these compounds. CYP2E1 has been shown to be important in rodent
metabolism of trichloroethylene; however, the chemical-specific data are sparse with regard to
its role in tetrachloroethylene metabolism (Dohertv et al.. 1996). Still, assuming that CYP2E1 is
important to tetrachloroethylene metabolism is not unreasonable. CYP3 A isoenzymes and
especially CYP2B1/2 may be important for tetrachloroethylene. Costa and Ivanetich (1980)
showed increased/decreased hepatic metabolism following treatment with agents now known to
selectively induce/inhibit CYP3A and/or CY2B specifically.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-19 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
3.3.5.3. Glutathione (GSH) Conjugation Pathway
The GSH-dependent pathway for tetrachloroethylene exists in both rodents and humans,
and the pathway is also operative for trichloroethylene in these species (Birner et al.. 1996;
Volkel et al.. 1998). The flux through this pathway at experimental or occupational exposures is
thought to be quantitatively less than that through the P450 pathway. Toxic metabolites can arise
from several sources in the pathway; however, for tetrachloroethylene, as well as for
trichloroethylene, the GSH pathway is associated with renal toxicity (Anders et al.. 1988; Dekant
et al.. 1989; IARC. 1995; Lash et al.. 2000; Lash and Parker. 2001; U.S. EPA. 1991b). For both
compounds, recovery of urinary mercapturates, the stable end-products of the GSH pathway,
comprises 1% or less of the total dose (Dekant et al.. 1986d; Lash and Parker. 2001). but this
does not reflect the total flux through the GSH pathway. In particular, the TCVC metabolite and
the corresponding dichlorovinyl cysteine and their respective A'-acetylated forms derived from
trichloroethylene might also act as substrates for renal P-lyases and other enzymes such as FM03
and CYP3A (Anders et al.. 1988; Dekant et al.. 1988; reviewed by; Dekant et al.. 1989; Lash et
al.. 2000; Lash and Parker. 2001; U.S. EPA. 1991b) (see Section 3.3.3). It should be noted that a
higher cysteine »Y-conjugate-to-mercapturate ratio exists for tetrachloroethylene when compared
to trichloroethylene, which could influence the relative bioactivation and nephrotoxicity of
these two compounds (Lash and Parker. 2001).
3.3.5.4. Summary
Tetrachloroethylene is closely related structurally to trichloroethylene, and the
two chemicals cause similar toxic effects, many of which are attributed to metabolic activation of
the parent compounds. Interestingly, although tetrachloroethylene is not as extensively oxidized
as trichloroethylene, they have similar potency for liver tumors, with which oxidative
metabolism is associated. TCA, DCA, chloral, and TCOH are reported P450 biotransformation
products of both tetrachloroethylene and trichloroethylene; however, only TCA predominates for
tetrachloroethylene whereas TCOH predominates for trichloroethylene. In addition, DCA is
likely formed via GSH conjugation for tetrachloroethylene and via oxidation for
trichloroethylene. The fact that the two compounds produce different reactive intermediate P450
metabolites is also important to consider. Excretion of urinary mercapturates suggests that,
relative to P450 oxidation, tetrachloroethylene is more extensively metabolized via GSH
conjugation than is trichloroethylene. However, these urinary excretion products do not reflect
the total flux through the GSH pathway since the glutathione and cysteine conjugates of both
chemicals have been shown to undergo further processing to products that are highly reactive.
Thus, regardless of similarities, both the qualitative and the quantitative differences between
tetrachloroethylene and trichloroethylene in metabolite production could have bearing on toxicity
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-20 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
and tumor induction, and the relative importance of various mechanisms and different modes of
action contributing to their toxic effects, including tumorigenesis, may vary between the
two parent compounds. Recognizing similarities and differences is important in attempting to
understand how each of these two compounds causes its toxic effects.
3 4 EXCRETION
Tetrachloroethylene is excreted from the body by pulmonary excretion of the parent
compound and urinary excretion of metabolism products, with a small amount of pulmonary
excretion of metabolism products. Tetrachloroethylene that is not metabolized is exhaled
unchanged, and this process is the primary pathway of tetrachloroethylene excretion in humans
for all routes of administration (Guberan and Fernandez. 1974; Koppel et al.. 1985; Monster et
al.. 1979; Opdam and Smolders. 1986; 1970; Stewart and Dodd. 1964; Stewart et al.. 1961;
1974; 1977). Pulmonary excretion of (unchanged) parent compound is also important in animals
(Bogen et al.. 1992; Frantz and Watanabe. 1983; Pegg et al.. 1979; Schumann et al.. 1980;
Yllner. 1961). A very small amount of tetrachloroethylene has been shown to be excreted
through the skin (Bolanowska and Golacka. 1972); however, it represents an insignificant
percentage of total tetrachloroethylene disposition.
Pulmonary excretion of unchanged tetrachloroethylene and other volatile compounds is
related to ventilation rate, cardiac output, and the solubility of the compound in blood and tissue.
The lung clearance of tetrachloroethylene in six adults exposed at rest to 72 ppm and 144 ppm of
tetrachloroethylene averaged 6.1 L/minute initially and decreased to 3.8 L/minute after 4 hours
(Monster et al.. 1979). Lung clearance represents the volume of air from which all
tetrachloroethylene can be removed per unit time. Normal ventilation rates in adults range from
5-8 L air/minute at rest. Pulmonary excretion of unchanged tetrachloroethylene 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 inverted equivalent of its uptake from the lungs. Pulmonary excretion
occurs in three first-order phases of desaturation of blood vessel-rich tissues, muscle tissue, and
adipose tissues (Guberan and Fernandez. 1974; Monster et al.. 1979). For humans, the
half-times of elimination from these three tissue groups are 12-16, 30-40, and 55-65 hours,
respectively (Monster et al.. 1979).
The long half-time of tetrachloroethylene elimination from adipose tissue, due to the high
adipose tissue:blood partition coefficient and the low rate of blood perfusion of the fat tissue
(EgerEI. 1963). is independent of the body burden of tetrachloroethylene, indicated by parallel
blood and exhaled air concentration decay curves (U.S. EPA. 1985b). However, the exhaled air
or end alveolar air concentrations and the blood concentrations after exposure and throughout
desaturation are proportional to the acquired body burden or exposure concentration and
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-21 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
duration, and they can serve as a means of estimating body burdens. The half-life of
tetrachloroethylene in the human body, measured as the inverse of the slope of the
log-concentration versus the time curve of the exhaled chemical, varies from 5-20 minutes for
the first phase of elimination up to approximately 50 hours during its extended phase (Chien.
1997; Monster et al.. 1979). The long half-time of tetrachloroethylene pulmonary excretion
indicates that a considerable time is necessary to completely eliminate the compound. This time
is greater than five times the half-life, or about 2 weeks for humans. For the rat, the half-time of
pulmonary elimination is about 7 hours.
Urinary and pulmonary clearances of metabolism products of tetrachloroethylene provide
other means of excretion. The mean half-time of urinary excretion for total trichloro-compounds
for 13 subjects exposed to tetrachloroethylene was determined to be 144 hours (Ikeda and
Imamura. 1973). When TCA is administered directly, however, the half-life is not that long.
The longer half-life of TCA from tetrachloroethylene metabolism is likely due to constant
metabolic conversion of the parent compound to TCA as tetrachloroethylene is cycled to the
liver over the period of time it is released from adipose tissue.
The urinary excretion of tetrachloroethylene biotransformation products, primarily TCA,
has been thought to represent only a small percentage of the total absorbed dose of
tetrachloroethylene in humans (ATSDR. 1997a; U.S. EPA. 1985b; Volkel et al.. 1998). Urinary
excretion of TCA (or total trichloro-compounds) was estimated to be only 1-3% in balance
studies conducted in humans (Boettner and Muranko. 1969; Chiu et al.. 2007; Essing et al.. 1973;
Fernandez et al.. 1976; Ikeda et al.. 1972; Monster et al.. 1983; Monster et al.. 1979; Monster and
Houtkooper. 1979; Stewart et al.. 1970; Stewart et al.. 1961). with urinary excretion of
GSH-derived metabolism products representing an even smaller fraction (Volkel et al.. 1998).
However, these studies did not follow urinary excretion for more than 3-7 days, and it is
possible that a larger percentage of the tetrachloroethylene dose was eventually excreted in urine.
In studies that also measured pulmonary excretion, the entire dose was not always accounted for
in the sum of exhaled tetrachloroethylene and urinary excretion of TCA (Chiu et al.. 2007;
Monster et al.. 1979). Part of the dose may be metabolized to biotransformation products that
were not measured, including oxidative products such as carbon monoxide, carbon dioxide, or
oxalic acid, and GSH conjugation products such as sulfoxides and reactive thiols (see
Section 3.3). In addition, the lowest exposures in these studies were around 1 ppm in air (Chiu et
al.. 2007). which is several orders of magnitude higher than ambient environmental exposures.
In laboratory animals, there is both a species- and dose-dependence to the amount of
pulmonary excretion of unchanged tetrachloroethylene that reflects the degree of metabolism
(Bogen et al.. 1992; Dallas et al.. 1994a; Pegg et al.. 1979; Schumann et al.. 1980). As the body
burden of tetrachloroethylene is increased in the rat or mouse, the percentage excreted as
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-22 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
unchanged parent compound increases. Conversely, as metabolism is the other principal route of
elimination of tetrachloroethylene, when the body burden increases, the percentage of the burden
metabolized decreases, although the absolute amount metabolized increases (Pegg et al.. 1979;
Schumann etal.. 1980). These observations suggest that, in the rodent, metabolism of
tetrachloroethylene and urinary excretion of its metabolites are rate limited and dose dependent,
whereas pulmonary excretion is a first-order process and is dose independent, with half-time and
rate constant being independent of the dose. Data from studies by Filser and Bolt (1979) and
Buben and O'Flaherty (1985) suggest that metabolism of tetrachloroethylene is greater in mice
than in rats, so conversely the amount of pulmonary excretion is greater in rats than in mice.
3 5 TOXICOKINETIC MODELING
Understanding tetrachloroethylene toxicokinetics is critical to both the qualitative and
quantitative assessment of human health risks from environmental exposures. A number of the
neurotoxic effects of tetrachloroethylene appear well correlated with parent compound
concentrations at the target site (Bushnell et al.. 2005). so characterizing tetrachloroethylene
blood or tissue concentrations can aid in performing risk assessment-related extrapolations, such
as between rodents and humans or between exposure routes. In addition, understanding
tetrachloroethylene metabolism is especially important toxicologically because specific
metabolites or metabolic pathways are associated with a number of endpoints of observed
toxicity. A more detailed discussion of the evidence for these associations, the specific
metabolites involved, and identification of the most appropriate dose metric are provided in
Section 5.
3.5.1. Choice of Physiologically Based Pharmacokinetic (PBPK) Model for Use in
Dose-Response Modeling
3.5.1.1. Limitations of Previously Developed Physiologically Based Pharmacokinetic
(PBPK) Models
A large number of PBPK models have been developed for tetrachloroethylene
toxicokinetics in both rodents and humans for various purposes. PBPK models can provide
estimates of tissue concentration as well as total metabolism of tetrachloroethylene. Provided
below is an overview of the models in literature— the aim of which is not to exhaustively cover
all of the models in the literature—but rather to capture the different assumptions made, the
range of data that has been used, and to indicate that these assumptions limit the ability of the
models to predict relevant tetrachloroethylene metabolite levels in humans.
Chen and Blancato (1987) developed a PBPK model for rats, mice, and humans. The
metabolic parameters maximum velocity (Vmax) and Km were derived by fitting the model to the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-23 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
total amount of metabolized tetrachloroethylene. Experimental data on total metabolite were
available for rodents. However, for humans, it was assumed that the urinary metabolite TCA, as
measured byMonster et al. (1979), accounted for 30% of the total metabolite. This percentage
was chosen because it resulted in a better fit.
Reitz et al. (1996) developed a PBPK model for rats, mice, and humans that describes the
total metabolism of tetrachloroethylene using Michaelis-Menten kinetics. The partition
coefficients for the five tissue compartments were measured independently. For rats and mice,
the metabolic parameters Vmax and Km, as well as the volume and blood flow rates of the fat
compartment, were obtained by simultaneously optimizing the fit to three sets of in vivo data
gathered in 6-hour inhalation radiolabeled tetrachloroethylene exposure studies. These data were
(a) concentration of tetrachloroethylene in exhaled breath, (b) radioactive body burden present in
animals at end of exposure, and (c) total postexposure radioactive metabolites recovered from all
excreta and carcass homogenates. The metabolic parameters for humans were estimated using a
—paedlelogram approach" (Reitz et al.. 1989). First-order constants for the rate of metabolism
were measured in vitro using isolated liver microsomes of all three species. The ratio of these in
vivo and in vitro metabolic rates was assumed to be nearly constant across species, as was found
to be the case for rats and mice. Using this constant ratio, the human in vivo metabolic rate
constant per gram of liver could be determined from the human in vitro value. Km was assumed
to be invariant across species because it is derived solely from the reaction rate constants for the
enzyme-catalyzed metabolic reactions. Reitz et al. (1996) also used a second method for
estimating Vmax, which was based on extrapolation from in vivo animal studies of other
chemicals metabolized by cytochrome P450 enzymes. Vmax, so estimated, was allometrically
scaled to humans. The values obtained by Reitz et al. (1996) through both these independent
methods were comparable.
Rao and Brown (1993) developed a human PBPK model for the purpose of investigating
neurotoxicological endpoints. The predictions of the model were fit to total metabolite levels
measured in rats and mice (Pegg et al.. 1979; Schumann et al.. 1980) to obtain Vmax
(allometrically scaled by body-weight3 4), and Km (considered invariant across species). Other
parameters were derived from various experimental data reported in the literature. The value of
Vmax for humans was determined by fitting the predicted total metabolite level to that estimated
from urinary metabolite measurements in humans (Fernandez et al.. 1976. combined; Monster et
al.. 1979 and), assuming that the ratio of urinary to total metabolites would be the same in
humans as that observed in rats (equal to 0.71).
Other authors have developed models for tetrachloroethylene that specifically describe
the kinetics of its major metabolite, TCA. Gearhart et al. (1993) developed a model for
tetrachloroethylene that also included the kinetics of TCA, assuming that TCA comprised 60%
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-24 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
of the total tetrachloroethylene metabolized in the rodent and using similar parameters for TCA
as in a model for trichloroethylene. Tetrachloroethylene metabolism parameters for mice were
estimated by fitting the model to the time course of tetrachloroethylene chamber concentration in
gas uptake studies. The model was independently validated at low oral doses (acute oral gavage
of tetrachloroethylene in corn oil) by comparing the time course of blood concentrations of
tetrachloroethylene and TCA in mice. 1 The parameters for describing tetrachloroethylene
metabolism in humans were derived by fitting the model to urinary excretion of TCA in
two subjects in a study by Fernandez et al. (1976). assuming the same ratio of TCA to total
metabolite as in the rodent. This value was set to 0.6 and attributed to Dekant et al. (1986a).
The validity of using this value for humans has not been evaluated. Reitz et al. (1996). in their
radiolabeled tetrachloroethylene studies, determined the fraction of urinary to total metabolites to
range from 0.49-0.59 in rats and from 0.56-0.66 in mice for exposure concentrations that varied
by two orders of magnitude.
Clewell et al. (2005) evaluated and extended the Gearhart et al. (1993) model further,
using tetrachloroethylene blood concentrations and urinary and blood TCA data gathered by
Volkel et al. (1998) on human subjects exposed to tetrachloroethylene concentrations of
10-40 ppm for 6 hours. They included metabolism of tetrachloroethylene in the kidney,
allowing for excretion directly into urine. By assuming metabolism in this organ to be at 10% of
the capacity of the liver, they obtained substantial improvement in the agreement with
experimental data on urinary excretion of TCA. An advantage in using the Volkel et al. (1998)
data is that they pertain to exposure concentrations that are lower than those in other studies
(e.g., 72-144 ppm in the (Monster et al.. 1979).
Loizou (2001) used a PBPK model that was structurally similar to that of Gearhart et al.
(1993). The model assumes a 15% stoichiometric yield for the total metabolite produced across
various dose levels (i.e., 15% of the parent compound in the liver is metabolized), but the basis
for these assumptions is not substantiated. The above yield is also assumed to hold for the
production of TCA because it is the major metabolite (personal communication from G. Loizou,
Health and Safety Laboratory, UK, to R. Subramaniam, U.S. EPA). Elimination rates of TCA
through blood and urine were chosen by calibrating the model to fit blood and urinary TCA
kinetics and exhaled tetrachloroethylene TCA concentration levels obtained from Monster et al.
(1979).
In addition, a number of PBPK models were developed only in humans, primarily to
characterize uncertainty and/or human variability. To assess intraindividual variability in uptake
and elimination over multiple exposure levels and scenarios, Chien (1997) collected exhaled
details pertaining to the derivation of parameters for metabolism in humans are not provided in the original paper
but are available in a review by Clewell et al. (20051.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-25 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
breath measurements on a single individual following four different exposure scenarios in a
controlled environmental facility (twelve, 30- or 90-minute exposures ranging from 0.5-3 ppm
in concentration) and following tetrachloroethylene exposure in 22 dry-cleaning facilities, where
ambient levels of tetrachloroethylene were recorded and exposures were carefully timed.
Bois et al. (1996), which was updated by Chiu and Bois (2006), used a Bayesian analysis
in conjunction with a PBPK model that was structurally similar to that used by Reitz et al.
(1996), and was only calibrated to the parent compound data (blood and exhaled breath) of the
individuals in Monster et al. (1979). The shape of the prior distribution was seen to have little
impact on final results. Model predictions were compared against alveolar concentrations of
subjects in the Opdam and Smolders (1986) study, and all data points were seen to fall within the
95th percentile envelope of predictions. The exposure concentrations in this study were
5-100 times lower than those used in the Monster et al. (1979) study; thus, this comparison
provides further weight to the strength of the model.
Covington et al. (2007) applied the same methodology to the Clewell et al. (2005) human
PBPK model, using additional data on the parent compound tetrachloroethylene and urinary
excretion data of its metabolite TCA (Fernandez et al.. 1976; Monster et al.. 1979). with a range
of exposure concentrations from 10-150 ppm. However, TCA blood concentrations from
Monster et al. (1979) were dropped from the analysis because the authors, in preliminary
calculations using a one-compartment PBPK model for TCE from Clewell et al. (2000). were
unable to reproduce the urinary excretion of TCA using the blood concentration data on TCA
from the same study. In addition, Covington et al. (2007) used only grouped data from both
these studies since the individual data were not available to them.
The Covington et al. (2007) analysis was revisited by Qiu et al. (2009). with the
following modifications:
1. A brain compartment was added.
2. Human kinetic data from Chiu et al. (2007) and Chien (1997) were used in addition to the
data used by Covington et al. (2007): namely Fernandez et al. (1976). Monster et al.
(1979). and Volkel et al. (1998). Thus, the human exposures used in the Qiu et al. (2009)
modeling range from 0.5-150 ppm.
3. Individual human data were used. However, blood TCA measurements from Volkel
et al. (1998) were not used, which the authors stated was because blood samples could
not be matched with individuals' urine samples and because there were not enough data
points to inform the time course for blood TCA. In addition, none of the TCA data from
Monster et al. (1979) were used.
4. Correlation between parameters (such as between cardiac output and alveolar ventilation
or between Vmax and Km) was addressed by reparameterization.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-26 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
5. Adjustment factors were added to maintain mass balance among fractional blood flows
and fractional tissue volumes.
These models provide substantially similar estimates of the tetrachloroethylene
concentration in the tissue. For example, as illustrated in Figures 3-2 and 3-3, estimated venous
blood concentration and alveolar concentration of tetrachloroethylene were in agreement to
within a factor of 2.0 among various models and experiment. However, the same models have
different approaches to estimating the metabolic parameters, thereby differing hugely in their
prediction of the amount metabolized at low dose—as shown in Figure 3-4 (adapted from Chiu
etj|Li_2007). These differences have major implications for the quantitative risk assessment and
represent the key controversy surrounding the application of PBPK models to
tetrachloroethylene toxicokinetics.
10
CD
E,
1 1
"E
CD
O
c
o
O
"O
o
m 01
Cfl
O
c
CD
>
0.01
Boismodel, VPR=1.2
Rao, Brown model, VPR=1.2
Reitz model, VPR=1.2
Bois model, VPR=1.6
Rao, Brown model, VPR=1.6
Reitz model, VPR=1.6
o Monster Experiment
-20
20 40 60 80 100 120
Post Exposure Time (hrs)
140
160
180
Figure 3-2. Comparison of model predictions for blood concentration with
experiment. PCE inhaled concentration was 72 ppm. Predictions are at different
ventilation-to-perfusion ratios and at an alveolar ventilation rate of 7 L/minute
(the geometric mean of values in the Monster experiment). Standard deviations
on the experimental data were very small (e.g., 0.025 mg/L and 0.003 mg/L at 20
and 140 hours, respectively). Experimental data adapted from Monster et al.
(1979).
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/2011 3-27 DRAFT-DO NOT CITE OR QUOTE
-------
10 -
Q.
Q.
O
£=
o
O
i_
TO
O
CD
>
<
1 -
%
O Bois model
v Rao & Brown mode|l
~ Reitzmodel
~ Stewart Expt
~
o
V
~
V Q»
o
o
~
V
#0
vv
Dn
4
~
V
~
~
o
~
On
5
o
S
o
m
o
&
o
¥
9
v
o
¥
~1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
50
100
150
200
250
300
350
Post-Exposure Time (hrs)
1 Figure 3-3. Comparison of model predictions for alveolar concentration of
2 tetrachloroethylene with experimental data on humans. Inhaled concentration is 100 ppm,
3 7 hours/day, for 5 days, and predictions assume alveolar ventilation rate of 5.02 L/minute and a
4 ventilation-to-perfusion ratio of 1.0. Experimental data show mean alveolar concentration in
5 subjects in Stewart et al. (1970). Some points early in the time course were deleted because of
6 difficulty in obtaining numerical values from the author's plot.
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/2011 3-28 DRAFT-DO NOT CITE OR QUOTE
-------
Chen and Blancato (1987)
Ward etal. (1988)
Bois etal. (1990) [1]
Bois etal. (1990) [2]
Rao and Brown (1993)
Reitzetal. (1996)
Bois etal. (1996)
Loizou (2001)
Clewell etal. (2005) [1]
Clewell et al. (2005) [2]
Chiu and Bois (2006)
Covington et al. (2007)
Qiu etal. (2009)
0.0% 20.0% 40.0% 60.0% 80.0%
Percent Metabolized at 1 ppb Inhalation Exposure
2.9%
4.4%
.9%
1.7%
2.3%
Figure 3-4. Previously published estimates for the total amount of
tetrachloroethylene metabolized at 0.001 ppm (1 ppb) continuous inhalation
exposure. All estimates are point estimates except for Bois et al. (1996) and Chiu
and Bois (2006), which are estimates of combined uncertainty and population
variability (95% confidence intervals [CIs]), and Covington et al. (2007) and Qiu
et al. (2009). which are estimates of uncertainty in the population mean
(90% CIs).
1 The various analyses in Figure 3-4 for tetrachloroethylene have a number of key
2 limitations. First, in no case have all the available data in mice, rats, and humans been
3 considered together in a single analysis. Thus, the extent to which different results reflect use of
4 different data sets and model structures is unclear. Moreover, while all the models estimate total
5 metabolism, those estimates are based on different types of data—in some cases, disappearance
6 of the parent compound, and in other cases, TCA and, therefore, oxidation—none of which
7 address GSH conjugation. These limitations and the above-mentioned controversy were also
8 noted in the National Research Council (NRC) report Review of the Environmental Protection
9 Agency's Draft IRIS Assessment of Tetrachloroethylene (NRC. 2010). In particular, NRC
10 concluded that, while a number of PBPK models have been developed for tetrachloroethylene ,
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/2011 3-29 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
they all have some key limitations that reduce the confidence with which they can be used for
risk assessment. NRC (2010) recommended the development of a —harmoniztf' PBPK model
that would integrate previous models and data. They pointed to the availability of in vitro and in
vivo data relevant to the GSH conjugation pathway, and they recommended exploring the
possibility of adding the GSH pathway to a harmonized PBPK model. This is important because
in the kidney, tetrachloroethylene causes tubular toxicity in mice and rats and is associated with
small increases in the incidences of kidney tumors reported in multiple strains of
tetrachloroethylene-exposed rats (JISA. 1993; NTP. 1986b). These effects are thought to be
associated with the tetrachloroethylene metabolism by GSH conjugation based on the production
in the kidney of nephrotoxic and genotoxic metabolites from this pathway (Lash and Parker.
2001).
3.5.1.2. The Chiu and Ginsberg (In Press) Model
In response to this advice from the NRC, another PBPK model was developed by Chiu
and Ginsberg (In Press). This model was developed to address many of the limitations of the
existing models for tetrachloroethylene, discussed above. Among the most important
improvements are (1) the utilization of all the available toxicokinetic data for tetrachloroethylene
and its metabolites in mice, rats, and humans; (2) the incorporation of available information on
the internal toxicokinetics of TCA derived from the most current PBPK modeling of
trichloroethylene and TCA; and (3) the separate estimation of oxidative and conjugation
metabolism pathways. Therefore, this assessment utilizes the Chiu and Ginsberg (In Press)
model to calculate relevant dose metrics to be used in dose-response modeling. An overview of
this model follows below.
In developing this model, first, a comprehensive literature search was made of relevant
toxicokinetic studies and the available toxicokinetic data digitized. These data were further
separated into —dabration" and "validation" data sets utilizing a wider range of data than any
previous analysis alone. Second, a harmonized PBPK model structure was developed that
separately tracked tetrachloroethylene oxidation and GSH conjugation. The Chiu and Ginsberg
(In Press) model includes a comprehensive analysis of TCA dosimetry originally developed by
the author for TCE, and it includes the urinary excretion kinetics of the metabolites NAcTCVC
and DCA. The Chiu and Ginsberg (In Press) model is described by the schematic below. The
reader is referred to Chiu and Ginsberg (In Press) for further details of the model structure.
The model structure and parameters (shown in Figure 3-5) used in the Chiu and Ginsberg
(In Press) harmonized model differed from other human models along the following lines:
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-30 DRAFT-DO NOT CITE OR QUOTE
-------
^Jnhaled
n
t
PCE
I Exhaled air J
Respiratory
Tract Lumen
(Inhalation)
Respiratory
Tract Tissue
Venous
Blood
_ _ 1
Oxidation
Respiratory
Tract Lumen
(Exhalation)
— — ^xi oat ion ¦ r
(Dead space) M
...?3?.E>
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
The model explicitly addressed GSH conjugation of tetrachloroethylene in the liver and
kidney.
• The urinary data on DCA ("Volkel et al.. 1998) were included so as to be able to consider
separate P-lyase-dependent and P-lyase-independent pathways for the bioactivation of
TCVC in the GSH conjugation pathway.
• An empirical —dtay" parameter (whose value was —fiad") was added for urinary
excretion of DCA and NAcTCVC and represented a —tlmped" delay in the time course
due to the processes of formation, urinary excretion, and other clearance pathways.
• For tetrachloroethylene oxidation, metabolic parameters were obtained from four in vitro
studies. These consisted of data from microsomes or cells from the liver and microsomes
from the kidney (Costa and Ivanetich. 1980. 1984; Lash et al.. 2007; Reitz et al.. 1996).
• A fraction of tetrachloroethylene oxidation was assumed to form compounds other than
TCA. A baseline value of 10% was used for the fraction not forming TCA.1
• GSH conjugation metabolic parameters were obtained from four studies that measured
tetrachloroethylene GSH conjugation in vitro (Dekant et al.. 1998; Green et al.. 1990;
2007; Lash et al.. 1998). These studies were utilized to select a baseline value for
metabolic clearance along this pathway in all species.
• The model incorporated all in vivo data considered in the literature for the PBPK
modeling of tetrachloroethylene and metabolites, dividing these data into two groups, one
for model calibration, and the other for model validation. These data included short and
long dosing periods.
• A full Bayesian uncertainty/variability analysis was not performed. The limited Bayesian
analysis involving flat priors and making inferences only using posterior modes was used
for the estimation of a limited number of metabolism parameters for which there was
significant discrepancy between baseline predictions (using baseline values of these
parameters) and in vivo data related to metabolism [see Table A-l of Chiu and Ginsberg
and associated text for rationale]. The Markov Chain Monte Carlo (MCMC) approach
was used for optimization.2
• The model structure allowed it to be used to calculate internal dose metrics for inhaled
and oral exposure to tetrachloroethylene for mice, rats, and humans. Thus, the analysis
invito data measuring TCA alone; TCA along with chloral hydrate, TCOH, and DCA; and total water soluble
metabolites are all generally found to be consistent with each other.
2The Markov Chain Monte Carlo (MCMC) method provides an algorithm to sample from a desired probability
distribution—in this case, the likelihood function—the output of which is a sequence of samples—the -Markov
chain," or —bain" for short. Each —bain" has a random starting point. In order to capture the potential uncertainty
due to different starting points, 24 chains with different starting points were run for mice and rats, and 48 chains
were run for humans. The posterior mode from each chain was determined—i.e., the —bain-specific posterior
modes." Then, the highest posterior model among the 24 (or 48) chains was determined—i.e., the —ovrall posterior
model," or simply the -posterior mode."
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-32 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
could be used for route-to-route extrapolation or interspecies extrapolation, comparison
of parent and metabolite toxicity based on a common internal dose metric, and
investigation of the shape of the dose-response curve. The following dose metrics could
be determined using this model, and the confidence with which it can make predictions
for internal dose metrics of interest was further evaluated by the authors:
o Daily area-under-the-curve of tetrachloroethylene in blood
o Fraction of tetrachloroethylene intake metabolized by oxidation
o Fraction of tetrachloroethylene intake metabolized by GSH conjugation
o Equivalent daily production of TCA per kg body weightl
3.5.1.2.1. Estimated human parameter values for oxidation and conjugation in Chiu and
Ginsberg fin Press)
The results for all estimated parameters are shown in Table 3-1. The estimated
metabolism parameters for oxidation and conjugation are of particular interest, so we focus
briefly on those, referring the reader to the original paper for further details on these and other
parameter estimation. Figure 3-6 compares the in vivo predictions for hepatic metabolism with
available in vitro data. For oxidation, in mice and rats, the optimized values are about an order
of magnitude higher than baseline values, whereas in humans, the optimized values are quite
similar to baseline values. However, they do not appear unreasonable compared to the limited
data available from other related compounds (TCE and some halomethanes), as shown in
Figure 3-6. For example, the linear rates are lower than those for TCE, which is known to be
more extensively oxidized by P450s than tetrachloroethylene. At higher substrate concentrations
the predicted rate of oxidation of tetrachloroethylene in mice and humans is greater than that for
TCE, but this is an artifact of the assumption of a linear rate necessitated by KM being
unidentifiable. For GSH conjugation, the range of the in vitro data is quite wide, especially
when also taking into considering data from other compounds (see Figure 3-6). In mice and rats,
the in vitro data on tetrachloroethylene GSH conjugation (filled symbols in Figure 3-6) spans the
range of estimates from optimization to in vivo data. For humans, the in vitro data only consist
of nondetects from Dekant et al. (1998). which, if assumed to be half the detection limit, are
more consistent with the alternative posterior modes. Overall, however, the ranges of predicted
rates for tetrachloroethylene are consistent with the range inferred from halomethanes, and the in
vivo optimized values do not appear to be substantially outside the bounds suggested by
available in vitro data.
1 TCA produced in the kidney and excreted directly to urine was not included, since it does not reach any target
organ (i.e., the liver) or enter systemic circulation.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-33 DRAFT-DO NOT CITE OR QUOTE
-------
Table 3-1. Log-likelihood and parameters after calibration
Parameter
Baseline
Postcalibration
(posterior
mode)
GSD of
posterior
modes across
chains
Range of posterior
modes across chains
Mouse
Ln(Likelihood)
-
-1,780
-
-1808-1780
QP (L/hr)
2.09
2.89
1.03
2.86-3.22
Vmax (mg/hr) (saturable
oxidation pathway)
0.23
0.026
1.16
0.022-0.0369
Km (L/hr) (saturable
oxidation pathway)
88.6
0.417
1.28
0.338-0.892
VmaX2/Km2 (L/hr) (linear
oxidation pathway)
-
0.0188
1.05
0.0165-0.0207
VmaxT C V G/KmT C VG
(L/hr) (linear conjugation
pathway)
0.656
6.83E-05
3.83
3.05e-05-0.00179
kMetTCA (/hr)
1.48
0.638
1.05
0.56-0.695
kUrnTCA (/hr)
2.93
1.26
1.05
1.11-1.38
Rat
Ln(Likelihood)
-
-1314
-
-1321—1314
QP (L/hr)
10.2
6.31
1.02
6.28-6.68
Vmax (mg/hr) (saturable
oxidation pathway)
0.256
0.87
1.37
0.415-1.93
Km (L/hr) (saturable
oxidation pathway)
69.7
31.1
1.39
14.8-71.9
VmaxT C V G/KmT C VG
(L/hr) (linear conjugation
pathway)
2.22
0.00204
1.27
0.00131-0.00355
kDCA (/hr)
-
0.129
1.65
0.0758-0.451
FracNATUrn
-
0.0143
1.29
0.00919-0.0253
FracDCAUrn
-
0.702
1.26
0.43-0.98
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-34 DRAFT-DO NOT CITE OR QUOTE
-------
Table 3-1. Log-likelihood and parameters after calibration (continued)
Parameter
Baseline
Postcalibration
(posterior
mode)
GSD of
posterior
modes across
chains
Range of posterior
modes across chains
Human
Ln(Likelihood)
-
1,828
-
1,790-1,828
QP (L/hr)
372
476
1.1
450-640
VMax/KM (L/hr) (linear
oxidation pathway)
0.353
0.454
1.08
0.346-0.468
VMaxKi d/KMKi d (L/hr)
(linear oxidation pathway)
0.00076
0.0947
1.09
0.0702-0.105
VMaxTCVG/KMTCVG
(L/hr) (linear conjugation
pathway)
0.0196
5.26
17.1
0.00194-5.48
kNAT (/hr)
-
0.28
1.07
0.228-0.293
FracNATUrn
-
0.000482
15.8
0.000472-1
FracDCAUrn
-
0.00022
18.5
1.12e-05-0.442
GSD = geometric standard deviation.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-35 DRAFT-DO NOT CITE OR QUOTE
-------
A:Mouse Oxidation
O)
§
o
E
E
c
o
¦g
x
O
o
o
E
E
c
O
CO
CT)
ZJ
c
o
O
o
a)
icr2 10"'
CV liver (mmol/L)
D:Rat Conjugation
D)
JZ
o
E
E
c
o
(0
O)
u
c
o
O
o
~5
CC
10~3 10~2 10"'
CV liver (mmol/L)
C:Rat Oxidation
E o-=
10~3 10~2 10"'
CV liver (mmol/L)
E:Human Oxidation
F:Human Conjugation
E o-
TTTt] 1 I I 11lll| 1 I I I
1CT2 10"'
CV liver (mmol/L)
1CT3 1CT2 1CT1
CV liver (mmol/L)
Figure 3-6. Comparison of mouse (A-B), rat (C-D), and human (E-F) rates of
hepatic oxidation (A, C, and E) or conjugation (B, D, and F) measured in
vitro (symbols) and predicted by the model (lines). Data shown consist of
measurements of tetrachloroethylene in vitro oxidation and conjugation [solid
circle: Dekant et al. (1998). solid square: Green et al. (1990): solid diamond: Lash
et al. (1998): solid triangle: Lash et al. (2007): solid upside-down triangle: Reitz
et al. (1996)1. reported fits of in vitro tetrachloroethylene Vmax and Km for
oxidation [grey-filled circle: Costa and Ivanetich (1980): grey-filled square: Costa
and Ivanetich (1984): grey-filled diamond: Lipscomb et al. (1998)TCE; grey-
filled triangle: (Wheeler et al.. 2001) CH2I2; grey-filled upside-down triangle:
(Wheeler et al.. 2001) CH2C12], and measurements of TCE in vitro conjugation
[open circle: Lash et al. (1998): open square: Lash et al. (1999): open diamond:
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/2011 3-36 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Green et al. (1997)1. Model predictions are using baseline parameters (dotted
line), overall posterior mode parameters (solid thick line), and alternative
posterior mode parameters (grey lines).
Chiu and Ginsberg (In Press) observe that overall the fits to the data and validation were
within threefold of the observed data, and more consistently so for rats and humans, given the
inter- and intraindividual variability. The discrepancies in model fits reflected variability to a
large extent. There was difficulty in fitting the time course of TCA in mice and the fraction of
retained tetrachloroethylene exhaled.
3.5.1.2.2. Dose metric predictions based on posterior modes
Tables 3-2-3-5 summarize the PBPK model dose metric predictions (listed in the
previous subsection) based on the baseline, overall posterior mode, and chain-specific posterior
mode parameters. The uncertainty due to the distribution of chain-specific posterior modes
contributes to the overall uncertainty in the predicted dose metric. The blood tetrachloroethylene
dose metric has by far the least amount of this —saipling" uncertainty. This appears to be true
across all species, routes of exposure, and exposure levels. The dose metrics with the next lower
amount of sampling uncertainty are tetrachloroethylene oxidation and TCA formation. The
predictions for GSH conjugation are more uncertain. In the rat, the ranges of chain-specific
posterior modes span up to twofold, and in mice up to 10-fold. However, in humans, the ranges
spans about 3,000-fold, discussed above.
3.5.1.2.3. Overall pertinent conclusions on tetrachloroethylene dosimetry
Chiu and Ginsberg also presented detailed sensitivity analyses that enable determination
of the confidence with which a particular dose metric can be estimated (see Table 9 and
Supplementary Materials in their paper). These have to be analyzed together with the residuals
for error in calibration and validation (see Table 10 of their paper) and the ranges in the values of
the predicted dose-metrics (presented above in Tables 3-2-3-5) to obtain perspective on the
overall uncertainty in the PBPK model predictions. Table 3-6 summarizes the various measures
that may contribute to this overall uncertainty.
The highest confidence dose metric in the Chiu and Ginsberg (In Press) analysis is the
AUC of tetrachloroethylene in blood. The main source of uncertainty in this case is the residual
difference between the model predictions and the calibration and validation data—a factor of
about twofold for each species. Therefore, this dose metric should be considered reliable for use
in risk assessment with the acknowledgement of a possible twofold residual error.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-37 DRAFT-DO NOT CITE OR QUOTE
-------
Table 3-2. Predictions for area-under-the-curve of tetrachloroethylene in
blood (mg-hr/L-day per ppm in air or mg-hr/L-day per mg/kg-day oral
intake) using posterior mode parameters
Species continuous
exposure
Baseline
Posterior
mode
GSD of
posterior
modes across
chains
Range of posterior
modes across chains
Mouse
0.01 ppm
1.2
2.13
1.03
2.11-2.42
0.1 ppm
1.2
2.13
1.03
2.12-2.42
1 ppm
1.26
2.18
1.02
2.16-2.44
10 ppm
1.73
2.43
1.01
2.39-2.53
100 ppm
2.8
2.64
1
2.64-2.68
1,000 ppm
2.98
2.68
1
2.67-2.72
0.01 mg/kg-day
0.0217
0.104
1.06
0.0957-0.126
0.1 mg/kg-day
0.0218
0.104
1.06
0.0958-0.126
1 mg/kg-day
0.0221
0.105
1.06
0.0965-0.127
10 mg/kg-day
0.0265
0.112
1.05
0.103-0.129
100 mg/kg-day
0.168
0.152
1.03
0.138-0.152
1,000 mg/kg-day
0.296
0.178
1.03
0.159-0.18
Rat
0.01 ppm
1.03
2.25
1
2.25-2.27
0.1 ppm
1.03
2.25
1
2.25-2.27
1 ppm
1.04
2.25
1
2.25-2.27
10 ppm
1.11
2.25
1
2.25-2.27
100 ppm
2
2.29
1
2.28-2.32
1,000 ppm
2.4
2.39
1.01
2.36-2.42
0.01 mg/kg-day
0.0737
0.852
1.02
0.807-0.86
0.1 mg/kg-day
0.0738
0.852
1.02
0.807-0.86
1 mg/kg-day
0.0744
0.852
1.02
0.807-0.86
10 mg/kg-day
0.0816
0.854
1.02
0.809-0.861
100 mg/kg-day
0.23
0.864
1.02
0.821-0.869
1,000 mg/kg-day
0.543
0.912
1.02
0.869-0.919
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-38 DRAFT-DO NOT CITE OR QUOTE
-------
Table 3-2. Predictions for area-under-the-curve of tetrachloroethylene in
blood (mg-hour/L-day per ppm in air or mg-hour/L-day per mg/kg-day oral
intake) using posterior mode parameters (continued)
Species continuous
exposure
Baseline
Posterior
mode
GSD of
posterior
modes across
chains
Range of posterior
modes across chains
Human
0.01 ppm
2.35
2.03
1.05
2.01-2.36
0.1 ppm
2.35
2.03
1.05
2.01-2.36
1 ppm
2.35
2.03
1.05
2.01-2.36
100 ppm
2.35
2.03
1.05
2.01-2.36
1,000 ppm
2.35
2.03
1.05
2.01-2.36
0.01 mg/kg-day
2.37
2.04
1.05
2.01-2.36
0.1 mg/kg-day
2.71
1.74
1.03
1.58-1.82
1 mg/kg-day
2.71
1.74
1.03
1.58-1.82
10 mg/kg-day
2.71
1.74
1.03
1.58-1.82
100 mg/kg-day
2.71
1.74
1.03
1.58-1.82
1,000 mg/kg-day
2.72
1.74
1.03
1.58-1.82
GSD = geometric standard deviation.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-39 DRAFT-DO NOT CITE OR QUOTE
-------
Table 3-3. Predictions for fraction of tetrachloroethylene in oxidized by
cytochrome P450 (P450s) (mg/kg-day oxidized per mg/kg-day intake) using
posterior mode parameters
Species
continuous exposure
Baseline
Posterior
mode
GSD of
posterior
modes across
chains
Range of posterior
modes across chains
Mouse
0.01 ppm
0.00252
0.188
1.1
0.12-0.192
0.1 ppm
0.00254
0.187
1.09
0.12-0.191
1 ppm
0.00269
0.174
1.08
0.115-0.179
10 ppm
0.0062
0.118
1.06
0.0934-0.124
100 ppm
0.0141
0.0732
1.04
0.0632-0.075
1,000 ppm
0.00716
0.0664
1.05
0.0574-0.0688
0.01 mg/kg-day
0.00367
0.561
1.08
0.395-0.574
0.1 mg/kg-day
0.00368
0.561
1.08
0.395-0.574
1 mg/kg-day
0.00374
0.557
1.07
0.394-0.57
10 mg/kg-day
0.00445
0.524
1.07
0.38-0.535
100 mg/kg-day
0.0253
0.35
1.04
0.308-0.367
1,000 mg/kg-day
0.0361
0.239
1.03
0.216-0.25
Rat
0.01 ppm
0.000501
0.0419
1.02
0.0387-0.042
0.1 ppm
0.000502
0.0419
1.02
0.0387-0.042
1 ppm
0.000514
0.0418
1.02
0.0386-0.0419
10 ppm
0.000662
0.0409
1.02
0.0379-0.0409
100 ppm
0.0025
0.0331
1.07
0.0263-0.0358
1,000 ppm
0.00153
0.011
1.27
0.00587-0.0181
0.01 mg/kg-day
0.00143
0.106
1.02
0.0988-0.107
0.1 mg/kg-day
0.00144
0.106
1.02
0.0988-0.107
1 mg/kg-day
0.00145
0.106
1.02
0.0987-0.107
10 mg/kg-day
0.00158
0.105
1.02
0.0977-0.105
100 mg/kg-day
0.00431
0.0934
1.04
0.0817-0.096
1,000 mg/kg-day
0.00686
0.0434
1.2
0.026-0.0631
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-40 DRAFT-DO NOT CITE OR QUOTE
-------
Table 3-3. Predictions for fraction of tetrachloroethylene in oxidized by
cytochrome P450 (P450s) (mg/kg-day oxidized per mg/kg-day intake) using
posterior mode parameters (continued)
Species
continuous exposure
Baseline
Posterior
mode
GSD of
posterior
modes across
chains
Range of posterior
modes across chains
Human
0.01 ppm
0.00971
0.0098
1.12
0.00694-0.0104
0.1 ppm
0.00971
0.0098
1.12
0.00694-0.0104
1 ppm
0.00969
0.0098
1.12
0.00694-0.0104
10 ppm
0.00955
0.0098
1.12
0.00694-0.0104
100 ppm
0.00828
0.0098
1.12
0.00694-0.0104
1,000 ppm
0.00355
0.0098
1.12
0.00693-0.0104
0.01 mg/kg-day
0.0173
0.0175
1.09
0.0134-0.0184
0.1 mg/kg-day
0.0173
0.0175
1.09
0.0134-0.0184
1 mg/kg-day
0.0173
0.0175
1.09
0.0134-0.0184
10 mg/kg-day
0.0169
0.0175
1.09
0.0134-0.0184
100 mg/kg-day
0.0138
0.0175
1.09
0.0134-0.0184
1,000 mg/kg-day
0.00492
0.0175
1.09
0.0133-0.0184
GSD = geometric standard deviation.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-41 DRAFT-DO NOT CITE OR QUOTE
-------
Table 3-4. Predictions for fraction of tetrachloroethylene in conjugated with
glutathione (GSH) (mg/kg-day conjugated per mg/kg-day intake) using
posterior mode parameters
Species
continuous exposure
Baseline
Posterior
mode
GSD of
posterior
modes across
chains
Range of posterior
modes across chains
Mouse
0.01 ppm
0.348
0.000151
3.87
6.39e-05-0.00415
0.1 ppm
0.347
0.000152
3.87
6.43e-05-0.00417
1 ppm
0.337
0.000159
3.86
6.83e-05-0.0043
10 ppm
0.244
0.000207
3.81
8.95e-05-0.00523
100 ppm
0.0299
0.000251
3.79
0.000109-0.00642
1,000 ppm
0.00301
0.000258
3.79
0.000111-0.00663
0.01 mg/kg-day
0.929
0.000481
3.89
0.000208-0.0134
0.1 mg/kg-day
0.929
0.000481
3.89
0.000208-0.0134
1 mg/kg-day
0.928
0.000485
3.89
0.00021-0.0135
10 mg/kg-day
0.914
0.000521
3.87
0.000229-0.0141
100 mg/kg-day
0.454
0.000706
3.82
0.00031-0.0181
1,000 mg/kg-day
0.0485
0.000821
3.81
0.000362-0.0212
Rat
0.01 ppm
0.303
0.00308
1.27
0.00195-0.00519
0.1 ppm
0.303
0.00308
1.27
0.00195-0.00519
1 ppm
0.301
0.00309
1.27
0.00195-0.0052
10 ppm
0.286
0.00309
1.27
0.00196-0.00521
100 ppm
0.0939
0.00316
1.27
0.002-0.00529
1,000 ppm
0.0099
0.00335
1.27
0.00213-0.00559
0.01 mg/kg-day
0.874
0.00783
1.27
0.00498-0.0133
0.1 mg/kg-day
0.874
0.00783
1.27
0.00498-0.0133
1 mg/kg-day
0.873
0.00783
1.27
0.00498-0.0133
10 mg/kg-day
0.861
0.00785
1.27
0.00499-0.0133
100 mg/kg-day
0.608
0.00795
1.27
0.00506-0.0134
1,000 mg/kg-day
0.078
0.00838
1.27
0.00535-0.0141
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-42 DRAFT-DO NOT CITE OR QUOTE
-------
Table 3-4. Predictions for fraction of tetrachloroethylene in conjugated
with glutathione (GSH) (mg/kg-day conjugated per mg/kg-day intake) using
posterior mode parameters (continued)
Species
continuous exposure
Baseline
Posterior
mode
GSD of
posterior
modes across
chains
Range of posterior
modes across chains
Human
0.01 ppm
0.000544
0.0936
17.5
3.16e-05-0.1
0.1 ppm
0.000543
0.0936
17.5
3.16e-05-0.1
1 ppm
0.000543
0.0936
17.5
3.16e-05-0.1
10 ppm
0.000535
0.0936
17.5
3.16e-05-0.1
100 ppm
0.000468
0.0935
17.5
3.16e-05-0.1
1,000 ppm
0.000207
0.0926
17.4
3.16e-05-0.0991
0.01 mg/kg-day
0.000972
0.177
17.1
6.47e-05-0.188
0.1 mg/kg-day
0.000972
0.177
17.1
6.47e-05-0.188
1 mg/kg-day
0.00097
0.177
17.1
6.47e-05-0.188
10 mg/kg-day
0.00095
0.177
17.1
6.47e-05-0.188
100 mg/kg-day
0.000788
0.177
17.1
6.47e-05-0.187
1,000 mg/kg-day
0.000289
0.175
17
6.47e-05-0.185
GSD = geometric standard deviation.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-43 DRAFT-DO NOT CITE OR QUOTE
-------
Table 3-5. Predictions for Trichloroacetic acid (TCA) produced systemically
(mg/kg-day systemic TCA per ppm in air or mg/kg-day systemic TCA per
mg/kg-day oral intake) using posterior mode parameters
Species
continuous exposure
Baseline
Posterior
mode
GSD of
posterior
modes across
chains
Range of posterior
modes across chains
Mouse
0.01 ppm
0.0361
3.74
1.08
2.63-3.94
0.1 ppm
0.0363
3.71
1.08
2.62-3.9
1 ppm
0.0384
3.45
1.07
2.53-3.59
10 ppm
0.0886
2.34
1.04
2.05-2.47
100 ppm
0.202
1.46
1.03
1.36-1.55
1,000 ppm
0.103
1.32
1.04
1.18-1.43
0.01 mg/kg-day
0.00325
0.497
1.08
0.35-0.509
0.1 mg/kg-day
0.00326
0.496
1.08
0.35-0.508
1 mg/kg-day
0.00331
0.493
1.07
0.349-0.505
10 mg/kg-day
0.00394
0.464
1.07
0.337-0.473
100 mg/kg-day
0.0224
0.31
1.04
0.273-0.325
1,000 mg/kg-day
0.032
0.212
1.03
0.191-0.222
Rat
0.01 ppm
0.00352
0.182
1.02
0.173-0.189
0.1 ppm
0.00353
0.182
1.02
0.173-0.189
1 ppm
0.00361
0.181
1.02
0.173-0.189
10 ppm
0.00465
0.177
1.02
0.169-0.183
100 ppm
0.0176
0.144
1.07
0.117-0.158
1,000 ppm
0.0108
0.0476
1.26
0.0261-0.0798
0.01 mg/kg-day
0.00127
0.0941
1.02
0.0875-0.0952
0.1 mg/kg-day
0.00127
0.0941
1.02
0.0875-0.0951
1 mg/kg-day
0.00128
0.094
1.02
0.0874-0.095
10 mg/kg-day
0.0014
0.0929
1.02
0.0866-0.0934
100 mg/kg-day
0.00382
0.0828
1.04
0.0724-0.0851
1,000 mg/kg-day
0.00607
0.0385
1.2
0.023-0.0559
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-44 DRAFT-DO NOT CITE OR QUOTE
-------
Table 3-5. Predictions for Trichloroacetic acid (TCA) produced systemically
(mg/kg-day systemic TCA per ppm in air or mg/kg-day systemic TCA per
mg/kg-day oral intake) using posterior mode parameters (continued)
Species
continuous exposure
Baseline
Posterior
mode
GSD of
posterior
modes across
chains
Range of posterior
modes across chains
Human
0.01 ppm
0.0106
0.0125
1.02
0.0117-0.0128
0.1 ppm
0.0106
0.0125
1.02
0.0117-0.0128
1 ppm
0.0106
0.0125
1.02
0.0117-0.0128
10 ppm
0.0104
0.0125
1.02
0.0117-0.0128
100 ppm
0.00906
0.0125
1.02
0.0117-0.0128
1,000 ppm
0.00388
0.0125
1.02
0.0117-0.0128
0.01 mg/kg-day
0.0153
0.0145
1.09
0.0111-0.0152
0.1 mg/kg-day
0.0153
0.0145
1.09
0.0111-0.0152
1 mg/kg-day
0.0153
0.0145
1.09
0.0111-0.0152
10 mg/kg-day
0.015
0.0145
1.09
0.0111-0.0152
100 mg/kg-day
0.0123
0.0145
1.09
0.0111-0.0152
1,000 mg/kg-day
0.00436
0.0145
1.09
0.011-0.0152
GSD = geometric standard deviation.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-45 DRAFT-DO NOT CITE OR QUOTE
-------
Table 3-6. Summary evaluation of the reliability of tetrachloroethylene dose
metrics
Calibration
Validation
Dose metric
species
error or
variability
(GSD)a
error or
variability
(GSD)a
Optimization
runs range"
Additional potential concerns
from sensitivity analysis
AUCCBld
Mouse
~2-fold
~2-fold
<10%
None
Rat
~2-fold
~2-fold
<10%
None
Human
~2-fold
~2-fold
<20%
None
FracOx
Mouse
~2-fold
~2-fold
<40%
Some sensitivity to lung
metabolism
Rat
~2-fold
~2-fold
<20%
None
Human
~2-fold
~3-fold
<1.5-fold
Some sensitivity to fraction of
oxidation to TCA
FracGSH
Mouse
NA
NA
~60-fold
None
Rat
~2-fold
NA
<30%
None
Human
~2-fold
NA
~3,000-fold
Calibration data cannot
distinguish between modes
TCASys
Mouse
~2-fold
~2-fold
<30%
Some sensitivity to fraction of
oxidation to TCA
Rat
~2-fold
~2-fold
<20%
Some sensitivity to fraction of
oxidation to TCA
Human
~2-fold
~3-fold
<40%
Some sensitivity to fraction of
oxidation to TCA
Evaluated in rodents at 10 ppm in air by inhalation and 100 mg/kg-day orally, and in humans at 0.01 ppm
in air by inhalation and 0.01 mg/kg-day orally.
GSD = geometric standard deviation.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-46 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
The next highest confidence is in the estimates of tetrachloroethylene oxidation and TCA
formation. Here, the estimates of tetrachloroethylene oxidation in mice and rats have similar
uncertainty to that for AUC of tetrachloroethylene in blood—predominantly twofold in the
residual difference between model predictions and the calibration and validation data. The range
in estimates of tetrachloroethylene oxidation in humans is largely dominated by interindividual
variability—i.e., the differences in urinary excretion of TCA across individuals. Thus, the
central tendencies for the population are well estimated—even if particular individuals may vary
to a fair degree. Thus, at the population level, these dose metrics should be considered reliable
for use in risk assessment with the acknowledgement of a residual error of about twofold or less.
In terms of predicted interspecies differences, the PBPK model generally predicts the
greatest oxidative metabolism in mice, followed by rats, and then humans. Humans would be
predicted to receive a smaller internal dose of oxidative metabolites for the same applied dose,
whether scaled by body weight or allometrically by body weight to the 3/4 power.
On the other hand, estimates of GSH conjugation appear more uncertain—especially for
humans. In rats, both the calibration data and the range of different optimization runs suggest
about a twofold uncertainty. In mice, there are no data on this pathway other than as a —lass
balance" from total metabolism (e.g., closed-chamber studies). Nonetheless, the range of
estimates based on the different optimization runs is about 60-fold. It is in the human predictions
that the range of estimates becomes extraordinarily large. In particular, there are evidently
two local maxima, each of which gives similar model fits, but for which model predictions differ
by 3,000-fold. This is a reflection not of the calibration data, which are fit quite well regardless,
but of the results of different optimization runs. Therefore, overall, the predictions for rat GSH
conjugation are considered reliable to about twofold, those for the mouse to about 60-fold, and
those for humans vary by about 3,000-fold. At this point, it is not possible to disentangle the
contributions of uncertainty and variability to the very large range of estimates of perc GSH
conjugation in humans.
Interestingly, the predictions appear to support the default assumption of equivalent
concentrations in air leading to equivalent internal doses, as the estimates of AUC of
tetrachloroethylene in blood are within twofold of each other across species. In addition, at the
higher oral doses (e.g., 100 mg/kg-day), rescaling the AUC in blood by body weight to the
3/4 power leads to estimates across species within threefold of each other. These can be explained
by the sensitivity analysis, which showed AUC in blood to be most sensitive to cardiac output,
alveolar ventilation, and the partition coefficient, all of which either are similar across species or
scale approximately allometrically by body weight to the 3/4 power across species.
The implications of these results are quite substantial—particularly for interspecies
extrapolation between rats and humans. In rats, all the evidence appears to support a low amount
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-47 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
(<1% of dose) of GSH metabolism. At environmental exposures, the overall posterior mode
predicts about 15- to 30-fold more GSH conjugation as a fraction of dose in humans relative to
rats, but the uncertainty range in humans overlaps with the rat estimates, so the data are also
consistent with humans having either equal or greater GSH conjugation..
The analysis in Chiu and Ginsberg (In Press) appears to have resolved a conflict between
PBPK model-based analyses that predicted high versus low amounts of tetrachloroethylene
metabolized in humans in two key aspects. This makes it particularly suited for use in this
assessment. First, there is now fairly high confidence in the predictions of oxidative metabolism
across species. Second, it has been made clear that the previously debated uncertainties in total
metabolism can be essentially attributed to uncertainty in GSH conjugation, which is substantial.
Those analyses that concluded low total tetrachloroethylene metabolism all restricted the fraction
of total (not oxidative) metabolism that was TCA to a fairly significant percentage—30-100%
(e.g., Chen and Blancato. 1989; Clewell et al.. 2005; Covington et al.. 2007; Qui et al.. 2009).
Thus, as was noted by the NRC (2010). total metabolism was essentially only measuring
oxidative metabolism. On the other hand, those analyses that concluded high total
tetrachloroethylene metabolism essentially lumped oxidative and GSH conjugation metabolism
together without restrictions as to the fraction producing TCA and/or made inferences based on
disappearance of the parent compound (e.g., Bois et al.. 1996; Bois et al.. 1990; Chiu and Bois.
2006; Reitz et al.. 1996; Ward et al.. 1988). Thus, the analysis in Chiu and Ginsberg (In Press)
essentially reconciles the disparate conclusions as to human tetrachloroethylene metabolism from
previously published PBPK models. First, the conclusion of —dw metabolism" is certainty true
for oxidation. Second, the conclusion of —high metabolism may be true for GSH conjugation
but is highly uncertain. In essence, both conclusions are consistent with the data if augmented by
some additional qualifications: oxidative metabolism is low in humans, while GSH conjugation
metabolism may be high or low in humans, with high uncertainty and/or variability.
Results obtained by applying the Chiu and Ginsberg (In Press) model for the
dose-response modeling in this assessment are presented in Section 5.
3.5.2. Age and Gender-Related Differences in Tetrachloroethylene Pharmacokinetics
Age and gender-specific differences in pharmacokinetics can have a substantial impact
on tissue dosimetry. The immaturity of metabolic enzyme systems in the perinatal period may
lead to decreased clearance of toxic chemicals as well as decreased production of reactive
metabolites. Clewell et al. (2004) examined these differences for various stages in life using
PBPK modeling for tetrachloroethylene and five other chemicals that differed considerably in
their physicochemical (lipophilicity, solubility, and volatility) and metabolic characteristics.
Parameters describing growth of various tissues were taken from the literature, and blood flow
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-48 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
changes with age were assumed to change proportionally with tissue volumes. For
tetrachloroethylene, only oxidative metabolism—specifically the production of TCA—was
considered. Data on age-dependent development of CYP2E1 were used for this purpose (Vieira
et al.. 1996). The parameters for tetrachloroethylene were taken from the Gearhart et al. (1993)
model, and the age dependence of metabolism was based on the CYP2E1 data. The Gearhart
et al. (1993) model describes the amount of TCA produced as 60% of the total metabolized
tetrachloroethylene; this was fixed in the life-stage model.
The dose metrics examined were blood concentrations of the parent compound and the
metabolite TCA. Continuous lifetime oral exposure was simulated at a daily dose rate of
1 |ig/kg-day. Table 3-7 provides the average daily dose during different life-stages of a male
expressed relative to that of a 25-year-old adult male. The gender and age differences in
tetrachloroethylene and TCA blood concentrations are detailed further in Figure 3-7.
Considerable gender differences in blood concentrations of TCA and tetrachloroethylene
were seen in these predictions. Internal dose during infancy differed most from the
corresponding dose in a 25-year-old. Tetrachloroethylene and TCA blood concentrations
increased with age, which the authors attributed to the lower metabolic and pulmonary clearance
of tetrachloroethylene when compared with other volatiles as well as its higher lipophilicity, both
resulting in storage of the compound in fat and other tissues. These age and gender differences
in pharmacokinetic sensitivity are significant, but they need to be considered together with
pharmacodynamic considerations in determining the contribution of exposure at a life-stage to
lifetime risk.
The same group of authors (i.e.. Gentry et al.. 2003) developed a PBPK model for
tetrachloroethylene that compared maternal and fetal/neonatal blood and tissue dose metrics
during pregnancy and lactation. The manuscript contains the details on the structure of the
model. Oxidative metabolism (TCA) in the mother and nursing infant was modeled using data
for CYP2E1 (Vieira et al.. 1996); metabolism in the fetus was not included due to lack of
information pertaining to the development of this pathway during gestation. The dose metrics
were the fetal and infant blood concentrations of tetrachloroethylene and TCA. Changes in fetal
blood concentrations were not pronounced because changes in tissue composition occurred in
both the mother and the fetus during pregnancy (Gentry et al.. 2003). A decrease of nearly
three orders of magnitude of blood concentrations in the lactating infant when compared with
that of the fetus was calculated. This decrease was attributed to the lower exposure rate during
lactation as compared with placental exposure. Concentrations in the lactating infant were
considerably lower, by more than two orders of magnitude, than in the mother. The largest
variation in blood concentration occurred in the early postnatal period.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-49 DRAFT-DO NOT CITE OR QUOTE
-------
As the authors indicated, validation of the results in the Clewell et al. (2004) and Gentry
et al. (2003) work and further refinement of the parameters in the models are necessary. It
would, therefore, be premature to consider the results of such analyses for use in risk assessment.
Further investigation of variability in the parameters used in the Clewell et al. (2004) analysis is
needed before the results from Table 3-3 can be used to weigh upon considerations of a
pharmacokinetic uncertainty factor for age and gender variability. Nonetheless, these models
Table 3-7. Ratio of average daily dose at various life-stages to the average
daily dose for a 25-year-old adult: physiologically based pharmacokinetic
(PBPK) simulations
Dose metric
Life-stage
0-6 months
0.5-5 years
5-25 years
25-75 years
PCE blood concentration
0.33
0.42
0.76
1.2
TCA blood concentration
0.057
0.16
0.59
1.4
Source: Clewell et al. (2004).
-r 5.0E-8
--4.0E-8 Q
u>
£
-- 3.0E-8 5
o
--2.0E-8 g
O
o
10E-8 £
J 0.0E+0
0 10 20 30 40 50 60 70 80
Age (years)
Figure 3-7. Physiologically based pharmacokinetic (PBPK) simulations of
variations with age and gender in blood concentrations of
tetrachloroethylene and its main metabolite trichloroacetic acid (TCA).
Simulations are for continuous lifetime oral exposure at a constant daily intake of
1 |ig/kg-day.
1.5E-4
O 1.0E-4
en
cj 5.0E-5
PERC - males
PERC - females
TCA - males
TCA - females
0.0E+0
h+h-h+h-h+h-h-H-h-H
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/2011 3-50 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Source: Clewell et al. (2004).
will enable future studies to focus on the key factors that are likely to influence pharmacokinetic
susceptibility.
3.5.3. Metabolic Interactions with Other Chemicals
Fisher et al. (2004) used PBPK modeling and complementary studies in mice to
investigate the effect of coexposures of orally administered carbon tetrachloride (CT) and
tetrachloroethylene on metabolic interactions between the two chemicals. CT is known to inhibit
its own metabolism (referred to as suicide inhibition). TCA was used as a biomarker to assess
the inhibition of the cytochrome P450 system by CT. Oral bolus intubation in the dose range of
1-100 mg/kg of CT was followed by a dose of 100 mg/kg of tetrachloroethylene an hour later. It
was concluded that dose additivity could not be used to predict interactions between the
compounds in this dose range because the metabolic interactions were found to be highly
nonlinear. The inhibition in metabolic capacity of tetrachloroethylene 2 hours after
administration of CT and 1 hour after single dose administration of tetrachloroethylene was
found to be 5, 52, and 90% at CT doses of 1.5, 10, and 19 mg/kg, respectively.
Dobrev et al. (2002) performed gas uptake studies in F344 rats and developed a mixture
PBPK model for humans to study interaction effects during coexposure to mixtures of TCE,
tetrachloroethylene, and methyl chloroform. Corresponding to a 10% increase in TCE blood
concentration, the production rates of toxic conjugative metabolites exceeded 17%, pointing to a
nonlinear interaction effect due to coexposure to TCE.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 3-51 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
4. HAZARD IDENTIFICATION
This section discusses tetrachloroethylene toxicity on an organ-specific basis. For each
of the major organ systems, human effects are presented first, followed by effects in animals and
in in vitro systems. Cancer and noncancer toxicity and mode of action (MOA) are also included
in the discussions. The order of presentation is as follows: neurotoxicity (see Section 4.1);
kidney and bladder toxicity and cancer (see Section 4.2); liver toxicity and cancer (see
Section 4.3); esophageal cancer (see Section 4.4); lung and respiratory cancer (see Section 4.5);
immunotoxicity, hematologic toxicity, and cancers of the immune system (see Section 4.6);
developmental and reproductive toxicity, and reproductive cancers (see Section 4.7);
genotoxicity (see Section 4.8); and susceptible populations (see Section 4.9). Section 4.10
provides a summary of the hazard identification.
4 1 NEUROTOXICITY
4.1.1. Human Studies
A wide range of effects on neurologic function have been observed for both acute and
chronic-duration exposure to tetrachloroethylene in humans, as summarized below. Most of the
reports evaluating neurological function in humans were inhalation chamber or chronic exposure
studies. Study designs, exposure-assessment methods, and results of individual studies are
presented with a discussion of chamber studies in Section 4.1.1.1 and chronic exposure studies in
Section 4.1.1.2. Within the latter section, the studies are further divided by type of exposure
setting (occupational; residential). In residential settings, exposure is more likely to be
continuous and of lower concentrations compared with the more intermittent, higher
concentration, more variable exposure experienced in work settings. Section 4.1.1.3 presents a
summary of neuropsychological and neurobehavioral effects in low- and moderate-exposure
studies with observations across studies discussed by neurological domain, categorized by visual
function, cognitive function, motor function, and neurological and behavioral disorders.
Acute controlled inhalation exposures of 100 ppm and higher induced symptoms
consistent with depression of the central nervous system (CNS), such as dizziness and
drowsiness. Changes in electroencephalograms (EEGs) have also been noted with controlled
inhalation exposures at this level (Stewart et al.. 1977). Acute exposure to lower levels of
tetrachloroethylene (50 ppm for 4 hours/day for 4 days) induced alterations in neurobehavioral
function, with changes indicative of visual system dysfunction including delayed neuronal
processing time (Altmann et al.. 1990; Altmann et al.. 1992). A wide range in susceptibility to
neurological effects among the participants in these studies was observed.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-1 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Epidemiologic studies of workers or residents with chronic exposure to
tetrachloroethylene show that the nervous system is a target, with most of these studies reporting
decrements in one or more nervous system domains. The vision and cognitive domains are most
commonly affected (Altmann et al.. 1995; Cavalleri et al.. 1994; Echeverria et al.. 1994;
Echeverria et al.. 1995; Ferroni et al.. 1992; Gobba et al.. 1998; Lauwerys et al.. 1983;
McDermott et al.. 2005; Nakatsuka et al.. 1992; NYSDOH. 2005a. b, 2010; Schreiber et al..
2002; Seeber. 1989; Sharanieet-Kaur et al.. 2004; Spinatonda et al.. 1997). Other reports (Laslo-
Baker et al.. 2004; Till et al.. 2001a; Till et al.. 2005; Till et al.. 2001b) suggest a vulnerability of
the fetus to organic solvent exposures, including tetrachloroethylene exposure. Deficits in
neurobehavioral parameters and in visual system functioning in young children of mothers
exposed during pregnancy compared with children of unexposed mothers were observed (Till et
al.. 2001a; Till et al.. 2005; Till et al.. 2001b). These reports are not discussed further in this
section because they do not provide specific data pertaining to tetrachloroethylene exposure.
Few studies are available on neurologic diseases such as Parkinson's disease, amyotrophic lateral
sclerosis, and Alzheimer's disease and organic solvents (TOM. 2002). and none of these reports
uniquely assess tetrachloroethylene. The influence of tetrachloroethylene exposure on risk of
these neurological diseases is not addressed in this Toxicological Review.
4.1.1.1. Chamber Studies
Several controlled experiments were conducted in the 1970s examining neurological
effects from short-term exposures (5-7.5 hours per day for 4 or 5 consecutive days) to
tetrachloroethylene at levels up to 100 ppm. There is no description in the published reports of
the informed consent and other human subjects research ethics procedures undertaken in these
studies, but there is no evidence that the conduct of the research was fundamentally unethical or
significantly deficient relative to the ethical standards prevailing at the time the research was
conducted.
In a study by Stewart et al. (1970). 12 healthy adults were exposed to 100 ppm for
7 hours; eye and nose irritation was reported by 60% of the subjects, a slight frontal headache by
26%, mild light-headedness by 26%, drowsiness by 40%, and difficulty speaking by 25%. Of
five healthy men exposed to 100 ppm for 7 hours/day on 5 consecutive days, one reported a mild
frontal headache during each exposure, and two consistently reported mild eye and throat
irritation. Individual responses during exposures to 0 ppm were not assessed. Three tests of
equilibrium (a modified Romberg test, where an individual stands on one foot with eyes closed
and arms at side; a heel-to-toe test; and a finger-to-nose test) were performed every 60 minutes
during each day of exposure. After 6 hours, neurobehavioral tests of motor function (the
Crawford manual dexterity and Flanagan coordination tests), cognitive function (arithmetic test),
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-2 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
and motor/cognitive function (inspection test) were also performed. Three of the subjects
exhibited impairments to equilibrium within the first 3 hours of exposure but were able to
perform the test normally when given a second chance. Stewart et al. (1970) concluded that
there were CNS effects in some subjects exposed to 100 ppm and that there exists a large range
of individual susceptibility to tetrachloroethylene.
In the 6-week study by Hake and Stewart (1977), four healthy men were exposed
7.5 hours/day to 0 ppm (2 days in Week 1, 1 day in Week 3, and 2 days in Week 6), 21 ppm
(4 consecutive days in Week 3), 100 ppm (5 consecutive days in Week 2), and a time-weighted
average (TWA) of 100 ppm (5 consecutive days in Week 4) when exposure levels were more
than 53, 100, or 155 ppm (5 consecutive days in Week 5). In addition, four healthy women were
exposed to 100 ppm for 7.5 hours/day on 5 consecutive days and to 0 ppm on 2 days. The
subjects were told that they would be exposed to various concentration of tetrachloroethylene,
but they were not told their sequence of exposures (a single-blind protocol). Reports of
symptoms (e.g., headache) varied among individuals, but overall, complaints during exposures
were similar to those during control conditions, exposures to 0 ppm of tetrachloroethylene. The
evaluation of electroencephalogram (EEG) recordings made during exposure suggested altered
patterns indicative of cortical depression in three of four men and four of five women exposed to
100 ppm (constant or TWA). In five subjects, altered EEG recordings occurred during Hours 4
to 7 of exposure; another subject had altered recordings within 10 minutes of exposure, which
gradually returned to normal during continued exposure, and the seventh subject showed changes
between 30 minutes and 6-7 hours of exposure. Recordings of visual-evoked potentials in
response to bright flashes of light (i.e., neurophysiological measurements of the electrical signals
generated by the visual system in response to visual stimuli) and equilibrium tests (Romberg and
heel-to-toe) were normal in men and women. The performance of men on neurobehavioral tests
of cognitive function (arithmetic), motor function (alertness), motor/cognitive function
(inspection), and time estimation was not significantly affected by any exposure. The
performances of men on a second test of motor function (Flanagan coordination) were
significantly decreased (p < 0.05) on 1 of 3 days during each of 2 weeks of exposure to 100 ppm
and on 2 of 3 days during the week of exposure to 155 ppm, but Hake and Stewart (1977)
concluded that only the results at 155 ppm were related to tetrachloroethylene. In women,
alertness (the only neurobehavioral endpoint evaluated) was not affected by exposure to
tetrachloroethylene. Hake and Stewart (1977) concluded that (1) there is considerable
interindividual variation in response to tetrachloroethylene vapors, (2) EEG analysis indicates
preliminary signs of narcosis in most subjects exposed to 100 ppm for 7.5 hours, (3) impairment
of coordination may occur in subjects exposed to 155 ppm for 7.5 hours, and (4) the effects are
likely due to tetrachloroethylene itself, given its slow metabolism in humans. They also reported
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-3 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
that their data suggested that a threshold limit value of 100 ppm contains no margin of safety for
susceptible subjects—both subjectively and neurologically—to the vapors of tetrachloroethylene.
Altmann et al. (1990; 1992) examined neurological effects of tetrachloroethylene on
healthy adults exposed to 10 ppm or 50 ppm for 4 hours on 4 consecutive days. Visual acuity of
all subjects was normal or corrected to normal. The study was a single-blind study (subjects
were not told their level of exposure), and subjects were randomly assigned to either group.
Sixteen subjects were exposed to 10 ppm, and 12 subjects were exposed to 50 ppm. However,
neurophysiological measurements were made on only 22 subjects (12 at the low-exposure level
and 10 at the high-exposure level). Three neurophysiological measurements were taken on the
day before exposure started and on each of the four exposure days: (1) visual evoked potentials
in response to black-and-white checkerboard patterns; (2) a visual contrast sensitivity (VCS) test;
and (3) recordings of brainstem auditory-evoked potentials (neurophysiological measurements of
the electrical signals generated by the hearing system in response to auditory stimuli) to evaluate
peripheral hearing loss. All measurements were started 2 hours after a subject entered the
chamber and were completed within 1 hour. A German version of the Neurobehavioral
Evaluation System was used to assess motor, motor/cognitive, and cognitive function of subjects.
The battery included nine tests (finger tapping, eye-hand coordination, simple reaction time,
continuous performance, symbol digit, visual retention, pattern recognition, digit span, and
paired associates). A vocabulary test and a test of emotional state (moods) were also given.
Each subject was assessed with a complete battery of tests during the preexposure baseline
assessment and at the end of the study. Subsets of the battery covering motor function and mood
were given at the beginning and end of each 4-hour exposure period. Tetrachloroethylene was
not detected in blood samples collected before the start of the first exposure period. The
detection limit was less than 0.0005 mg/L. Mean tetrachloroethylene blood levels increased
slightly over the 4-day period. Among subjects exposed to 10 ppm, mean blood levels were
0.33, 0.36, 0.4, and 0.38 mg/L at the end of Days 1, 2, 3, and 4 of exposure, respectively.
Among subjects exposed to 50 ppm, mean blood levels were 1.1, 1.2, 1.4, and 1.5 mg/L at the
end of Days 1, 2, 3, and 4 of exposure, respectively.
The visual-evoked potential latencies of subjects during the 3rd hour of exposure to
50 ppm on Days 1, 2, 3, and 4 of exposure were significantly longer (p < 0.05) compared with
those measured on the control day, and the differences became progressively longer on
successive exposure days. One set of visual-evoked potential latencies on the day after the end
of the exposure period remained longer than the control day values (statistical significance not
reported). Visual-evoked potential latencies in subjects with exposure to 10 ppm were not
statistically significantly longer than those recorded on the control day. There were significant
differences (p < 0.05) between the visual-evoked potential latencies of subjects exposed to
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-4 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
10 ppm and those exposed to 50 ppm. Data on contrast sensitivity indicated greater effects at
50 ppm than at 10 ppm; effects were most pronounced on the last day of exposure. However,
statistical analysis was not reported. There were no indications of peripheral hearing loss at
either exposure level. Neurobehavioral tests results were reported only for those tests given
repeatedly on 4 consecutive days (finger tapping, eye-hand coordination test, simple reaction
time, continuous performance, and moods). There were postexposure performance deficits (p =
0.05) among subjects exposed to 50 ppm when compared with the group exposed to 10 ppm in
tests of motor/cognitive function (continuous performance test for vigilance) and motor function
(eye-hand coordination), and a near-significant difference (p = 0.09) on a test of motor function
(simple reaction time). In all cases, the degree of improvement shown by the subjects exposed to
50 ppm was less than that shown by the subjects exposed to 10 ppm. There were no exposure-
related effects on the finger-tapping or moods test. Altmann et al. (1990) concluded that visual
function in healthy, young, adult males is mildly affected by tetrachloroethylene exposures to 50
ppm maintained for 4 hours on each of 4 days and stated that the impaired performance on tests
of motor/cognitive and motor function suggests that 50 ppm cannot be considered a NOAEL for
neurobehavioral endpoints indicative of CNS depression (Altmann et al.. 1992).
4.1.1.2. Chronic Exposure Studies
Table 4-1 summarizes details of the chronic-duration tetrachloroethylene exposure
studies evaluating neurological function using tests of specific neurological domains in humans.
Most of these are studies of dry-cleaning and laundry workers, but some studies examined
neurobehavioral or visual system effects among residents living in close proximity to a dry-
cleaning establishment (Altmann et al.. 1995; NYSDOH. 2005a. b; Schreiber et al.. 2002) or in
other workers employed in the same building as a dry-cleaning business (Schreiber et al.. 2002).
Exposure levels were approximately an order of magnitude higher in occupational settings
compared with residential exposure. Tetrachloroethylene concentrations reported in the dry-
cleaning and laundry worker studies ranged from an 8-hour TWA mean of 7 ppm for dry-cleaner
workers in Cavalleri et al. (1994) to an 8-hour TWA of 41 ppm for operators of a wet-transfer
dry-cleaning machine in Echeverria et al. (1995). Mean tetrachloroethylene concentrations in
residences near a dry-cleaning business were 0.4 ppm and 0.7 ppm, respectively, in studies in
New York City (Schreiber et al.. 2002) and Germany (Altmann etal.. 1995). Two additional
studies examining color vision in solvent-exposed workers (Muttrav et al.. 1997) and peripheral
neuropathy among patients with solvent-induced encephalopathy (Albers et al.. 1999) were
identified but are not presented because they involved solvent mixtures.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-5 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-1. Summary of human neurotoxicity studies of occupational or
residential exposures to dry-cleaning facilities using tetrachloroethylene
Subjects, methods
Exposure levels
Results
Reference(s)
Occupational exposures: dry-cleaning settings
Belgium, 26 dry cleaners, 33
unexposed workers (controls),
B, EA, PA, U; not blinded to
exposure status
Mean TWA = 21 ppm, mean
duration = 6.4 yr
Statistically significant differences
for simple reaction time (before
work) and critical flicker fusion
(before and after work); better
scores in exposed workers.
Lauwerys et
al. (1983)
Germany, 101 dry cleaners
(both sexes), 84 unexposed
workers (controls). PA, AA;
blinded to exposure status
Low-exposure group (n = 57):
mean TWA = 12 ppm, mean
duration = 11.8 yr; high-
exposure group (n = 44): mean
TWA = 53 ppm, mean duration
= 10.6 yr
Decrease in information-processing
speed (perceptual threshold, choice
reaction time), visual scanning
(cancellation dZ test), visuospatial
memory (digit reproduction) in dry
cleaners compared with controls; no
difference between high- and low-
exposure groups. No fine motor
function deficits.
Seeber (1989)
China, 64 dry cleaners,
120 controls (clerical workers
in factories). PA; not blinded to
exposure status
Geometric mean TWA =15
ppm (males), 11 ppm
(females), duration not
reported
No effect on color vision loss (using
less sensitive Lanthony test).
Nakatsuka et
al. (1992)
Italy, 60 dry cleaners, 30
controls (hospital launderers, no
solvent use). B, A; blinded to
exposure level but not status
Mean TWA =15 ppm, mean
duration = 10.1 yr
Impaired performance on simple
reaction time, vigilance, stress. No
fine motor function deficit. No
effects on digit symbol test. No
dose-response patterns seen.
Ferroni et al.
(1992)
Italy, 22 dry cleaners and 13
ironers, 35 controls. PA, EA;
blinded to exposure level
Mean TWA = 6 ppm (7.3 ppm,
dry-cleaning workers; 4.8 ppm,
ironers), mean duration = 8.8
yr
Color confusion index elevated
among dry cleaners (p = 0.007);
statistically significant exposure
(TWA)-response relationship. No
effect seen in ironers.
Cavalleri et al.
(1994)
Italy, 33 dry cleaners and
ironers. self controls (baseline
measurements in Cavalleri et
al.. 1994). PA; not clear if
blinded
Geometric mean TWA
ppm: Group A Group B
(n = 19) (n = 14)
Baseline 1.67 2.95
Follow-up 4.35 0.66
Increased CCI in Group A
(p < 0.01); no change in Group B.
CCI correlated with exposure levels
(r = 0.38,/? < 0.05).
Gobba et al.
(1998) (follow
ud of Cavalleri
et al.. 1994s)
Michigan, 65 dry cleaners,
pressers, clerks; no unexposed
group, PA; blinded to exposure
level
Chronic exposure score based
on work history: low (/? = 24:
2.1 yr), moderate (n = 18; 3.9
yr), high (« = 23; 14.6 yr)
Statistically significant decrease in
high compared with low exposure
on three tests of visuospatial
memory. No effect on digit span
Echeverria et
al. (1995)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-6 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-1. Summary of human neurotoxicity studies of occupational or
residential exposures to dry-cleaning facilities using tetrachloroethylene
(continued)
Subjects, methods"
Exposure levels
Results
Reference(s)
Washington, 45 dry cleaners
matched to 69 laundry workers,
59 pressers or counter clerks
from the same shop as the dry
cleaner operator. PA; blinded
to exposure level
Chronic exposure score groups
based on detailed work history
and estimated measures: mean
= 0, 68, and 1,150 with
corresponding 8-h TWAs of
<0.2, 3, and 9 ppm. Mean
duration = 2.6 to 11 yr for low-
and high-exposure groups,
respectively
Evidence of associations between
chronic exposure and reduced test
performance on three tests of
visuospatial memory: switching (p =
0.10), pattern memory (p = 0.03),
and pattern recognition (p = 0.09)
Echeverria et
al. (1994)
Italy, 35 dry cleaners, 39 age-
and education-matched
controls. AA; not blinded to
exposure status
Median = 8 ppm, grab sample.
Mean duration of employment
= 10.6 yr (from Figure 2)
Increase in vocal reaction time to
visual stimuli (reading task); dose-
response relationship
Spinatonda et
al. (1997)
Malaysia, 14 dry cleaners, 29
controls (support staff of
Universiti Kebangsaan
Malaysia, control Group 2); not
blinded to exposure status
No exposure information
presented in paper other than
PCE was used for dry cleaning
43 and 93% of dry cleaners
compared to 0 controls had errors on
the color vision D-15 test and FM
100 Hue test, respectively. Number
of errors on FM 100 Hue test also
increased in dry cleaners (p < 0.05)
Sharanjeet-
Kaur et al.
(2004)
Israel, 88,820 births,
1964-1976, identified in
Jerusalem Perinatal Study,
linked to national Psychiatric
Registry for hospitalization
with a schizophrenia-related
diagnosis through 1997
Occupation of mother and
father listed as dry cleaner on
birth certificate
Four cases were identified in 144
offspring of dry cleaners. RRof3.4
(95% CI: 1.3-9.2) for schizophrenia
in the offspring of dry cleaners
using proportional hazard modeling
Perrin et al.
(2007)
Occupational exposures: other settings
New York, 9 employees of day-
care center located in a building
with a dry-cleaning business, 9
age- and gender-matched
unexposed controls. PA, EA,
B, U; not blinded to exposure
status
Mean =0.32 ppm (monitoring
before closure of dry cleaners).
No information on duration of
employment
Decreased color discrimination
among exposed but not statistically
significant.
Lower (worse) scores on tests of
visual contrast sensitivity
Schreiber et al.
(2002)
New York. 4-yr follow-up of
13 children who had attended a
day care located in a building
with a dry-cleaning business, 13
children matched to exposed
children on age, gender, and
daycare experience; not blinded
to exposure status
Exposure had ceased 4 yr
earlier
No difference in visual function
(VCS, color vision) or
neurobehavioral function between
exposed children and controls
NYSDOH,
(2005b)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-7 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Table 4-1. Summary of human neurotoxicity studies of occupational or
residential exposures to dry-cleaning facilities using tetrachloroethylene
(continued)
Subjects, methods"
Exposure levels
Results
Reference(s)
Residential exposures
Germany, residents near dry-
cleaning business, 14 exposed
and 23 age- and gender-
matched nonexposed controls.
AA, B; not clear if blinded to
exposure status
Mean = 7 d monitoring period,
0.7 ppm, mean duration = 10.6
yr
Statistically significant increase in
simple reaction time and decrease in
continuous performance and
visuospatial function. No fine
motor function deficits
Altmann et al.
(1995)
New York, 17 exposed
(apartment residents living
above dry-cleaning business)
and 17 age- and gender-
matched controls. AA, PA, EA,
B, U; not blinded to exposure
status
Mean = 0.4 ppm (monitoring
before closure of dry cleaners).
Mean duration of residence = 6
yr
Decreased color discrimination
among exposed, but not statistically
significant.
Lower (worse) scores on tests of
visual contrast sensitivity
Schreiber et al.
(2002)
New York, 65 households (67
adults and 68 children) in
residential buildings with
colocated dry cleaners, 61
households (61 adults and 71
children) in residential
buildings without dry cleaners.
AA; not blinded to exposure
status
Geometric mean = 5 ppb
(0.005 ppm). Mean duration of
residence = 10 yr
Association (p < 0.05) between PCE
(indoor air and blood) and
performance on test of visual
contrast sensitivity in children. No
association observed in adults.
Color vision impairment (p < 0.05)
among children but not adult
exposed subjects as compared with
controls
NYSDOH,
(2005a);
McDermott et
al. (2005)
A = air sample, not specified area or personal sample, AA = area air samples, B = biological monitoring of blood,
CI = confidence interval, EA = exhaled air samples, PA = Personal air samples, RR = relative risk, U = biological
monitoring of urine for trichloroacetic acid, VCS = visual contrast sensitivity.
Vision testing in the four studies included tests of acuity, tests of spatial vision based on
contrast sensitivity, and tests of color vision. The visual acuity test measured the ability to
discriminate high-frequency (i.e., small) images at high contrast; e.g., reading successively
smaller black-on-white letters as part of an examination for corrective lenses. This measure
typically is dependent on the optics of the eye (and corrective lenses when needed) and is
insensitive to subclinical deficits in neurologic function. Contrast sensitivity measures the least
amount of luminance difference between dark and light bars needed to detect a given pattern
(e.g., a bar pattern). Impairments in color vision, beginning as blue-yellow confusion errors,
have been reported in populations exposed to organic solvents (Campagna et al.. 1996;
Campagna et al.. 1995; Mergler. 1987; Mergler et al.. 1988a; Mergler and Blain. 1987; Mergler
et al.. 1988b; Mergler et al.. 1991). The tetrachloroethylene exposure studies that assessed color
vision relied on various versions of the Lanthony color vision test. This type of test consists of a
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-8 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
series of small round -eaps" that the subject is asked to arrange in order by color. The types of
errors made can distinguish specific types of color vision deficiency; e.g., red-green color
confusion errors (blindness) is a common condition in males, mostly but not entirely of
congenital origin, whereas blue-yellow color confusion errors are very rarely due to congenital
conditions and, therefore, are considered as a hallmark of an acquired condition. Test scores are
based on the subject's ability to arrange a set of 15 caps according to a definite chromatic
sequence, with each mistake increasing the score above a perfect score of 1.00. A formula (the
Color Confusion Index [CCI]), based on Total Color Distance Scores can be used for scoring
(Bowman. 1982; Geller. 2001). The Lanthony D-15 desaturated test is more sensitive to mild
and moderate changes in color vision compared with other versions of the test that use more
contrasting hues (Lanthony. 1978). The vision tests are not recommended for epidemiological
studies of children under 5 years of age.
Other types of neurobehavioral effects were assessed in these studies using standardized
tests of cognitive or motor function, such as the digit symbol, digit span, Benton visual memory,
and simple reaction time tests. The standardized neurobehavioral battery has a high rate of
reliability and has been used to assess normal neurological function (Anger et al.. 2000).
As with most conditions, age is an important factor that needs to be considered in
interpreting measures of neurological function. Generally, the comparison group within these
studies was age-matched (individually or frequency-matched) to the exposed subjects. Measures
of cognitive function can also be influenced by education (or more broadly, socioeconomic status
variables), by other intelligence measures, and by alcohol use. Thus, these attributes would also
need to be considered in studies using cognitive tests such as visuospatial memory, vigilance,
and information processing. Alcohol use, smoking, certain medications, chronic neurological
conditions, and solvents other than tetrachloroethylene may affect visual contrast sensitivity and
color vision measures (Paramei et al.. 2004; Swinker and Burke. 2002). In contrast, color vision
and spatial vision have not been shown to be related to education or socioeconomic status, so
potential confounding by these factors is unlikely.
4.1.1.2.1. Occupational exposure studies: dry-cleaning settings
Lauwerys et al. (1983) studied 261 workers (24 women and 2 men) occupationally
exposed to tetrachloroethylene in six dry-cleaning shops in Belgium for a mean of 6.4 years
(range 0.1 to 25 years) and 33 controls (31 women and 2 men) working in a chocolate factory
(n = 20) or an occupational health service (n = 13) without occupational exposure to organic
solvents. No information is provided in the paper on the methods used to identify subjects or
1 Abstract of paper reports 22 subjects were exposed to tetrachloroethylene, but the full text of the paper includes
26 subjects.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-9 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
their reasons for participating in the study. The level of education was similar in the exposed and
control groups, but the prevalence of smokers was higher among dry-cleaning workers (50%)
compared with the controls (27%). Neurobehavioral tests of motor function (simple and choice
reaction time), sensory function (critical flicker fusion), and cognitive function (sustained
attention test) were given twice to each worker, once before work and once after work. Both
groups were tested in the middle of the workweek. Individuals also were questioned about
chronic neurological symptoms (e.g., fatigue, depression, sleep disturbances). Blood samples
were collected both before and after work. The mean tetrachloroethylene air concentration
(8-hour TWA) was 21 ppm, and the range of TWA values was 9 to 38 ppm, using results from
active sampling of personal air. The mean tetrachloroethylene blood level (30 minutes after the
end of work) was 1.2 mg/L (range of means from the shops was 0.6 to 2.4 mg/L).
Trichloroacetic acid, a metabolite of tetrachloroethylene, was not detected (level of detection
[LOD] not identified in published paper) in urine specimens from exposed subjects. An
evaluation of the subjects was performed at each worksite, so examiners were not blinded to
exposure status. The score of the critical flicker fusion test (a test of sensory function) was
significantly increased (better performance) in the exposed workers compared with controls
when given both before and after work. Decreased simple reaction time was seen among the
exposed workers in the tests performed before work (mean ± standard deviation [SD]: 0.374 ±
0.120 and 0.448 ±0.155 seconds in exposed and nonexposed workers, respectively) but not in
the tests performed after work (mean ± SD: 0.341 ±0.116 and 0.356 ± 0.128 seconds in exposed
and nonexposed workers, respectively). The dry-cleaning workers did not differ from controls
on the other three neurobehavioral tests. The prevalence of abnormal scores (those beyond the
5th or 95th percentile of the control group) did not vary significantly between the two groups.
Seeber (1989V evaluated the neurobehavioral effects of tetrachloroethylene in
101 German dry-cleaning workers (machine operators, ironers, touch-up workers, counter
attendants, and other employees) who were employed in coin-operated or while-you-wait shops,
all affiliated with one organization. The workers were separated into a low-exposure group
(50 women, 7 men) and a high-exposure group (39 women, 5 men) based on activities and room
air measurements. A third group of 84 sales personnel (64 women, 20 men) from several
department stores and receptionists from large hotels served as unexposed controls. No
information was provided on the methods used to identify subjects or their reasons for
participating in the study, although the authors reported that 29 service technicians were
excluded from the study because of either discontinuous exposure conditions with peak
1 Dr. Seeber provided additional information on this study in written correspondence to the New York State
Department of Health (NYSDOH) dated January 19 and May 20, 1996. This information appears in
NYSDOH (1997).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-10 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
concentrations or long periods of no exposure. Predominant characteristics of both groups
included primarily standing work, contact with customers, and moderate physical exercise.
Mean tetrachloroethylene concentrations (8-hour TWA) for the low- and high-exposure groups
were 12 (±8) ppm and 53 (±17) ppm, respectively, using results from active sampling of room air
and passive sampling of personal air. The mean durations of occupational exposure for the low-
and high-exposure groups were 11.8 and 10.6 years, respectively.
Several tests of neuropsychological functioning were administered, including
standardized personality tests, tests of sensorimotor function (including finger tapping and
aiming), the Mira and Santa Ana dexterity tests, and tests of information processing speed
(threshold of perceptual speed and choice reaction time) (Seeber. 1989). Some details of the
testing procedures were not provided, and one of the response variables, —delayedeactions,"
was not defined. The typical dependent variable measured in this task—response reaction
time—apparently was not measured; only the number of correct reactions was reported. Subtests
of the Wechsler Intelligence Test (digit span, digit symbol, and cancellations) were used, as was
recognition of words, faces, and digits. Intelligence was assessed using the logical thinking
subtest of the German Performance Test System. The neurobehavioral tests were given by two
specialized clinic staff members who did not question the subjects regarding exposure status.
The control group was younger than the dry-cleaning workers (mean ages: 38.2, 38.4,
and 31.8 years, respectively, in the low-exposure, high-exposure, and control groups,
respectively) and alcohol consumption also differed by group (mean: 8.2, 10.4, and 12.6 g/day in
the low-exposure, high-exposure, and control groups, respectively) (Seeber. 1989). Higher
scores on the intelligence test were observed among the control group (mean ± SD: 21.9 ± 5.8)
compared with the dry-cleaning workers (mean ± SD: 18.3 ± 5.0 and 19.2 ± 5.2 in the low- and
high-exposure groups, respectively). Age, gender, and intelligence scores were included in the
regression models analyzing the relation between exposure and neurobehavioral test scores;
additional control for group differences in alcohol consumption did not alter the observed results.
Performance of both the low-exposure and high-exposure groups differed significantly (p < 0.01)
from that of the unexposed control group on the threshold of perceptual speed and —delayed
responses" on a choice reaction time task (p = 0.08 and 0.03 for low-exposure and high-exposure
groups, respectively). Both exposed groups also had worse scores (p < 0.01) on two tests of
attention (digit reproduction and digit symbol) and on visual scanning (cancellations). There was
relatively little difference between the group mean scores comparing the low-exposure and high-
exposure groups on these tests (all ^-values >0.10). The low-exposure group also showed
significantly higher scores than did the control group on neurological signs (p < 0.01) and
emotional lability (p < 0.05). Scores of the high-exposure group for these measures were
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-11 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
intermediate between the control and the low-exposure group scores. There were no differences
between groups on the other tests.
Nakatsuka et al. (1992) evaluated the effects of tetrachloroethylene exposure on the color
vision of 64 dry-cleaning workers (34 women and 30 men) in China. Control workers
(72 women and 48 men) were recruited from the clerical sections of dry-cleaning shops and from
other factories (paint production plants or plants producing tetrachloroethylene from
trichloroethylene). No information is provided in the paper on the methods used to identify
subjects or their reasons for participating in the study. The mean ages of the dry-cleaning
workers (34.2 years for men, 35.3 years for women) were similar to those of the male controls
(34.0 years) but slightly higher than the female controls (32.6 years). The Lanthony new color
test, a test for screening color vision, and the Ishihara's color vision test, a test used for
confirmation of red-green vision loss, were carried out by ophthalmologists or occupational
health doctors in charge of the factories under one of two lighting conditions (natural sunlight or
a daylight fluorescent light). (This color vision test is not as sensitive as the Lanthony D-15 test
used in the other studies discussed in this section.) The geometric mean air concentrations of
tetrachloroethylene (averaging time not reported) were 15.3 and 10.7 ppm for the men and
women, respectively, using results from passive sampling of personal air. The overall geometric
mean was 13 ppm. The authors reported no significant difference in the performance of the dry-
cleaning workers (or other solvent-exposed groups included in the study) and unexposed controls
on the Lanthony new color vision test, with 60% of the male dry-cleaning workers and 63% of
male controls classified as "normal" color vision. Corresponding figures for females were 91
and 74%) in the dry-cleaning workers and controls, respectively. Results for the males were not
appreciably different when individuals with red-green vision loss were excluded.1 Nakatsuka et
al. (1992) concluded, overall, that they found no distinct color vision loss among the dry-
cleaning workers.
Ferroni et al. (1992)2 evaluated neurobehavioral effects and prolactin levels among
60 female dry cleaners and 30 unexposed female controls. Prolactin secretion by the pituitary is
controlled by hypothalamic dopamine; dopamine is also important to neurotransmitter systems,
and serum prolactin, as a biochemical signal and marker of nervous system function, is a
proposed alternative for assessment of nervous system toxicity (Manzo et al.. 1996). The
1 A statistical analysis of the dry cleaners data using a Fisher's exact test (for differences in proportions with at least
one sparse cell) indicated that tetrachloroethylene-exposed women were more likely to have normal color vision as
compared with unexposed women (p = 0.0423), but no difference was seen among the males (0.83, based on
Chi-squared test); reported in public comments of the Halogenated Solvents Industry Alliance to EPA (Halogenated
Solvents Industry Alliance. 2004) on the Neurotoxicity of Tetrachloroethylene Discussion Paper (U.S. EPA. 2003).
2 Dr. Mutti provided details on the selection process of exposed and control subjects and also clarified reported
results to Dr. Ken Bodgen, NYSDOH, in written correspondence dated July 29 and September 5, 1995 (see
NYSDOH. 1997).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-12 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
workers at every dry-cleaning shop in a small town outside of Parma, Italy, were invited to
participate in the study. There were no refusals. Controls were selected from the workers at a
hospital who cleaned clothes using a water-based process. Their jobs were essentially the same
as those of the dry cleaners, but they were not exposed to any organic solvents. Both groups
filled out a questionnaire on their health status, medication (including oral contraceptives),
lifestyle, and current and past jobs. Both groups met the following criteria: no history of
metabolic disorders, no history of psychiatric disorders, and low level of daily alcohol intake.
The dry cleaners and controls were comparable in age (mean ages: 39.7 and 37.6 years,
respectively), vocabulary level, height, weight, body mass index, smoking habits, and use of
medication. Workplace air samples were randomly collected throughout the workweek during
summer and winter to account for variability related to either the work cycle or seasonal
environmental fluctuations. Blood samples were collected during the workday during summer
and winter. The median tetrachloroethylene air concentration (4-hour TWA) was 15 ppm (range:
1 to 67 ppm). The subjects' range of tetrachloroethylene blood levels was 0.012 to 0.864 mg/L
[median = 0.145 mg/L; incorrectly expressed in Ferroni et al. (1992) as 12,864 and 145 mg/L,
NY State Department of Health (1997)1. The mean duration of occupational exposure was 10
years.
Workers and controls were given five neurobehavioral tests (part of the Swedish
Performance Evaluation System, —adjated" Italian version: finger tapping with both dominant
hand and nondominant hand, simple reaction time, digit symbol test, shape comparison-
vigilance, and shape comparison-response to stress) (Ferroni et al.. 1992). All subjects were
examined in the morning before their work shift in the same room by the same examiners, using
a standardized testing protocol (NYSDOH. 1997). Although the examiners were not blind to the
status of the subjects (dry cleaner or control), they were blind to the worker's exposure level
(NYSDOH. 1997). Serum prolactin levels were measured in all subjects using a blood sample
taken at the time of the neurobehavioral testing; analysis was limited to those samples obtained
during the proliferative (follicular) phase of the menstrual cycle (41 dry cleaners and
23 controls). Ferroni et al. (1992) did not describe the protocol for determining menstrual cycle
phase, however. Serum samples from dry cleaners and controls were alternated and analyzed in
the same experimental runs (NYSDOH. 1997).
The dry cleaners showed significantly reduced performance when compared with the
unexposed matched controls in three tests (simple reaction time,/? < 0.0001; vigilance,
p < 0.005; and stress, p < 0.005) (Ferroni et al.. 1992). Performance on the finger-tapping test
(both hands) and digit symbol test was not affected (NYSDOH. 1997). Additionally, the mean
serum level of prolactin was significantly higher in the workers than in the matched controls
(mean: 12.1 compared with 7.4 |j,g/L,/> < 0.001). Among the dry cleaners, none of the three
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-13 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
measures of exposure (duration of exposure and air or blood concentration of
tetrachloroethylene) was significantly associated with decreased test scores or increased serum
prolactin levels. Ferroni et al. (1992) concluded that tetrachloroethylene exposure in dry-
cleaning shops may impair performance.
Cavalleri et al. (1994) evaluated the effects of tetrachloroethylene exposure on the color
vision of dry cleaners and a comparison group of matched controls. The investigators compiled
a list of all the dry-cleaning shops in the municipality of Modena, Italy (110 shops employing
189 workers) and randomly selected 60 dry cleaners from 28 premises for recruitment into the
study (Aggazzotti etal.. 1994a). Only full-time workers (n = 52) were asked to participate, and
two declined. All 50 workers provided, via questionnaires, information on work history, health
status, occupational and hobby use of solvents, drinking and smoking habits, and drug use.
Thirty-five of the 50 dry cleaners (33 women, 2 men) met the inclusion criteria; others were
excluded for hypertension, smoking more than 30 cigarettes a day, alcohol consumption
exceeding 50 g of alcohol a day, oculo-visual pathology, or employed at a dry-cleaning facility
for less than 1 year. Another worker was excluded because a matched control could not be
found. The controls were factory workers who were not occupationally exposed to solvents or
other neurotoxic chemicals; they were selected and recruited into the study using the same
methods that were used for dry cleaners. The controls (n = 35) were from factories in the
Modena area and met the same inclusion criteria as the dry cleaners. They were matched to dry
cleaners by gender, age (±3 years), alcohol consumption (±10 g/day), and cigarette use
(±5 cigarettes a day). The mean age of both groups (35 years) and the percentages of each group
that were smokers (43%) or alcohol drinkers (71%) were comparable. All subjects appeared
healthy and met minimal status of visual acuity. None of the subjects reported hobby exposure
to solvents or other substances toxic to the eye. There were no known systematic differences
between exposed and control groups or between machine operators and ironers. Color vision
was assessed using the Lanthony D-15 desaturated panel test. Exposed and control subjects were
tested in random order (NYSDOH. 1997). All subjects were tested at the same time of day (in
the morning, before work) under the same lighting conditions by the same investigator. With
respect to exposed subjects, the investigator was unaware of both the exposure levels and the job
(operator or ironer) of each dry cleaner.
For all dry cleaners, the mean tetrachloroethylene air concentration (8-hour TWA) was
6 ppm, and the range of TWA values was 0.4-31 ppm, using results from passive sampling of
personal air (Cavalleri etal.. 1994). For operators (n = 22), the mean air concentration 8-hour
TWA was 7.3 ppm (range 0.4-31 ppm). For ironers (n = 13), mean air concentration (8-hour
TWA) was 4.8 ppm (range 0.5-11 ppm). The mean duration of occupational exposure was
8.8 years. Tetrachloroethylene concentrations were also measured in alveolar air for a subset of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-14 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
these dry cleaners, with a high correlation observed between tetrachloroethylene concentration in
alveolar air and 8-hour TWA levels in ambient air [r = 0.8 ,p< 0.001; Aggazzotti et al. (1994a)].
Only three dry-cleaning workers, as opposed to 13 controls, scored a perfect test score on
the color vision test {p < 0.01). Mistakes were made mainly in the blue-yellow range. Overall,
the workers showed poorer performance on the test as compared to controls, and they had a
significantly higher error rate (mean CCI score: 1.143 and 1.108 in workers and controls,
respectively, p = 0.03). The effect was seen in dry cleaners (mean 1.192 and 1.089 in dry
cleaners and their matched controls, respectively, p = 0.007) but not among the ironers (mean:
1.061 and 1.073 in ironers and their matched controls, respectively). There also was a
statistically significant positive correlation (p < 0.01) between TWA air concentrations and the
CCI (r = 0.52), which remained after multivariate analysis considered previous
tetrachloroethylene exposure, duration, age, number of cigarettes a day, and daily intake of
alcohol as covariates. The CCI values were not associated with two other measures of
tetrachloroethylene exposure (mean duration and an integrated index of exposure, yearly TWA
level). The study authors suggested that this may reflect the difficulty in controlling for the
interactive effects of age and exposure and accurately evaluating exposure. The effect on color
vision may not be rapidly reversible; preliminary data showed that the scores of some workers
did not improve when retested after 4 weeks of vacation (NYSDOH. 1997). Moreover, some of
these workers showed poorer performance on this test in the follow-up study by Gobba et al.
(1998). described below, suggesting color vision impairment is a chronic effect.
Gobba et al. (1998) reexamined color vision after a period of 2 years in 33 of the 35 dry
cleaners and ironers examined by Cavalleri et al. (1994). Two subjects had retired during the
2-year period between examinations. These investigators used the Lanthony D-15 test, the test
used by Cavalleri et al. (1994) to assess color vision, and performance was compared with the
subject's score from the initial survey. Tetrachloroethylene concentration in the occupational
setting was determined in the breathing zone using personal passive samplers. Monitoring was
carried out during the afternoon shift, as Cavalleri et al. (1994) did not show any differences
between morning and afternoon samples. Gobba et al. (1998) found that tetrachloroethylene
concentration had increased during the 2-year period for 19 subjects, identified as Group A
(geometric mean, from 1.67 ppm at the first survey to 4.35 ppm at the second survey), and had
decreased for 14 subjects, identified as Group B (geometric mean, from 2.95 ppm to 0.66 ppm).
The decrease in exposures was due to new equipment or other changes to the working
conditions. As found in the first survey, color vision was impaired primarily in the blue-yellow
range of color, with few subjects presenting a red-green errors. Color vision performance for the
entire group was related significantly to age (r = 0.45) and tetrachloroethylene concentration
(r = 0.39; p < 0.05). The mean CCI score for Group A subjects showed a statistically significant
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-15 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
difference between the two surveys (arithmetic mean: 1.16 and 1.26 in the first and second
surveys, respectively, p < 0.01). For Group B subjects, who experienced lower exposure
concentrations by the second survey, the CCI score did not change from that of the initial survey
(arithmetic mean: 1.15 and 1.15 in the first and second surveys, respectively). The findings in
Groups A and B were also supported using analysis of variance methods to examine the relation
between CCI score and exposure level (log TWA), adjusting for age, alcohol consumption, or
cigarette smoking between the subgroups.
Echeverria et al. (1995) assessed the performance of 65 dry-cleaning workers on
neurobehavioral tests. The testing was conducted in 1986. The owners of 125 shops in Detroit,
Michigan, were contacted, and 23 agreed to allow their workers to participate in the study.
Within each shop, operators were matched on education and age (±5 years) with a lower-
exposure subject. The subjects (35 men and 30 women) were grouped into three categories of
chronic tetrachloroethylene exposure (low, moderate, and high), based on type of shop (wet-
transfer or dry-to-dry), job title (counter clerk, presser, or operator), and years of employment.
All the operators were placed in the high-exposure category. There was no unexposed control
group. Dry-cleaning workers placed in the chronic exposure categories of low, moderate, and
high had been employed at their main job for 2.1, 3.9, and 14.6 years, respectively. Their mean
ages were 40.9, 40.6, and 43 years. The three groups were also characterized by estimates of
current exposure (low, medium, and high), which corresponded to mean tetrachloroethylene air
concentrations (8-hour TWA) of 11, 23, and 41 ppm, respectively, for counter clerks, pressers,
and operators in the more common wet-transfer shops (17 of 23 shops). Estimated air
concentrations for counter clerks, pressers, and operators in the dry-to-dry shops were 0.5, 10,
and 11 ppm, respectively. The estimates were based on a relationship between breath and air
concentrations derived from a larger independent study (Solet et al.. 1990). These estimates
were comparable to those found in other surveys of dry-cleaning facilities in the United States.
All subjects were tested in a minivan at the worksite in groups of two, in the afternoon
after work on the first or second day of their workweek (Echeverria et al.. 1995). Each subject
provided a breath sample and completed a medical, symptom, work history, and hobby
questionnaire. The subjects were administered six neurobehavioral tests, a test of verbal skills,
and questionnaires on emotional states (moods) and CNS symptoms. The neurobehavioral test
battery consisted of one test of motor/cognitive function (symbol digit) and five tests of cognitive
function (digit span, trailmaking A and B, visual reproduction, pattern memory, and pattern
recognition). Multivariate analysis was used to evaluate the relationship between a chronic index
of lifetime exposure and performance on neurobehavioral tests, accounting for the potential
confounding variables of current exposure, age, education, verbal skill, alcohol consumption,
hours of sleep, fatigue, mood, symptoms, medication, and secondary exposures to
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-16 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
neurotoxicants. After adjustment for factors affecting performance, the scores of the dry-
cleaning workers with high chronic exposure were reduced (compared with the low chronic
exposure group) by 4% for pattern recognition, 7% for pattern memory, and 14% for visual
reproduction (all ^-values <0.01). These impairments of visually mediated function were
consistent with the impairment of visuospatial functions observed in four patients who were
diagnosed with tetrachloroethylene encephalopathy who had been previously studied by
Echeverria et al. (1995). Other effects seen in the patients (mood changes and decreased
cognitive function in nonvisual tests) were not found in the dry-cleaning workers with high
lifetime exposures. Among complaints by the dry-cleaning workers, only the number of
complaints of dizziness from standing up rapidly and —alvent-induced dizziness" over the
previous 3 months was significantly elevated (p < 0.04) in the high-exposure group. Echeverria
et al. (1995) concluded that effects on visuospatial function were consistently found in subjects
employed as operators for an average of 14.6 years and exposed to an estimated
tetrachloroethylene 8-hour TWA air concentration of 41 ppm, suggesting a vulnerability of
visually mediated functions with tetrachloroethylene exposure. This conclusion was based on
the impaired performance of the high-exposure group when compared with a group of dry-
cleaning workers with low lifetime exposure, including workers who were probably clerks in
wet-transfer shops where the mean current exposure level was 11 ppm. This exposure level is
substantially above background ambient levels, and whether the performance of the low-
exposure group was impaired when compared with that of a group without occupational
exposure (i.e., an unexposed control group) is not known.
Echeverria et al. (1994) builds on the results of Echeverria et al. (1995),1 hypothesizing
degradation in behavior (particularly attention, executive function, visuospatial memory, short-
term memory, and mood) is an early indicator of neurotoxicity, leaving motor, language-based
skills, and long-term memory intact. The study was conducted in the Seattle/Tacoma,
Washington area from 1989 through 1993, when the area's dry-cleaning industry was switching
from wet-transfer to dry-to-dry machines. Initially, 320 dry-cleaning shops and laundries were
sent introductory letters requesting permission to allow their employees to participate in the
study. Of the 181 owners who responded, 39 agreed to participate. The most common reasons
for nonparticipation were disinterest, time constraints, lack of English proficiency, and concerns
about pending regulatory actions concerning tetrachloroethylene. Recruitment ended when a
total of 45 operators were enrolled. Each operator was matched with a less-exposed person from
the same shop. The subjects included laundry workers (n = 69), pressers or counter clerks
(n = 59), and operators or former operators (n = 45). The mean ages of the groups were 42.5,
1 Although published a year after this study (Echeverria et al.. 19951. the study by Echeverria et al. (19951. discussed
previously, was conducted in 1986, 3 years before this study.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-17 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
34.2, and 46.2 years, respectively. Women comprised 63% of the study population (109/173).
The subjects, who were paid volunteers, were eligible if they spoke English, had no history of
diabetes or CNS disorders, and had worked for more than 1 year in the trade. The final sample
excluded three subjects because of limited English and reading skills and six subjects who did
not wear glasses or were missing covariate information such as vocabulary test scores.
An index of chronic exposure and measures of subchronic and acute exposure were
developed for each subject. The chronic exposure index was based on a detailed work history,
including consideration of the type of dry-cleaning machine, job title, percentage of time at each
job title, estimated air levels associated with each job title, and employment duration. The
measures of subchronic and acute current exposure were based on mean 8-hour TWA air
concentrations measured on the day of neurobehavioral testing. Mean chronic indices were zero
for the never-exposed group of laundry workers, 68 for the dry-cleaning workers with low
exposure (pressers/clerks), and 1,150 for the dry-cleaning workers with high exposure
(operators). Mean exposures (8-hour TWA, using results from passive sampling of personal air)
for workers placed in these chronic exposure categories were <0.2 ppm (laundry workers), 3 ppm
(pressers/clerks), and 9 ppm (operators). Dry-cleaning workers placed in the chronic exposure
categories of low and high had been employed in their current job for 2.6 and 11 years,
respectively. The subjects also were placed in acute and subchronic exposure categories of
<1 ppm (laundry workers and some dry-cleaning workers, e.g., clerks), low (mainly pressers),
and high (operators), with corresponding current tetrachloroethylene 8-hour mean concentrations
of 0.5, 3, and 20 ppm, respectively. Dry-cleaning workers placed in the acute and subchronic
low exposure and high exposure categories had been employed in their current job for 5 and 9
years, respectively. Because of the changes in dry-cleaning practices over the course of the
study, many subjects in the high chronic-exposure category could be found in the low acute- and
low subchronic-exposure categories because these latter two indices were based on air
concentrations on the day of testing.
The test battery included tests of cognitive function, including visuospatial memory,
motor skills, mood, CNS symptoms, and basic verbal and arithmetic skills. The chronic and
subchronic assessment was based on tests given during the morning of each subject's day off and
on preshift scores. Each subject signed a consent form, provided a breath sample at each test
session, and completed a questionnaire covering transient factors that could affect performance
(e.g., headache). This was followed by questionnaires on medical history, medication, drug and
alcohol use, occupational and nonoccupational exposure to chemicals, symptoms, and mood.
Multivariate analysis was used to evaluate the relationship between exposure indices and
levels and performance on neurobehavioral tests after adjusting for the potential confounders of
age, gender, race, vocabulary level (as a surrogate for education and test-taking), and alcohol
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-18 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
consumption. Indications of associations between increased indices of chronic (lifetime)
exposure and reduced test performance were found in three tests of cognitive function: switching
(p = 0.1), pattern memory (p = 0.03), and pattern recognition (p = 0.09). The magnitude of
change attributable to tetrachloroethylene was a 3% loss in function for the latency of pattern
memory and an 11% loss in function for the correct number in visual reproductions. Subjective
measures of mood and symptoms were not significantly associated with exposure. Dry-cleaning
workers scored lower (but not significantly) on all but one of the remaining tests (the digit span
test). Analysis of the association between test scores and measures of subchronic exposure
(8-hour TWA tetrachloroethylene concentrations on the day of testing) confirmed the findings of
the chronic analysis: reduced scores on tests of switching (p = 0.1) and pattern recognition
(p = 0.04) as exposure increased. In summary, Echeverria et al. (1994) detected deficits in
visuospatial function (reduced performance in tests of pattern memory and pattern recognition)
in the dry-cleaning workers categorized as having high lifetime chronic exposure and whose
current exposure level was 9 ppm, 8-hour TWA. However, the exposure level of 9 ppm is not
representative of past chronic exposure levels because of changes occurring in the industry in the
study area (i.e., switching from wet-transfer to dry-to-dry machine). The investigators attributed
the reduced performance to exposures 3 to 5 years previously that were about two to four times
higher, and they hypothesized that a few years of reduced exposure may not be long enough to
eliminate the residual effects on visuospatial function caused by the exposures associated with
wet-transfer machines.
Spinatonda et al. (1997) assessed the effect of tetrachloroethylene exposure on vocal
reaction times among 35 dry cleaners and 39 unexposed controls. Controls were matched to
exposed individuals by age (mean age of 35 years for both groups) and education. The published
paper did not identify the population from which exposed subjects and controls were drawn or
the inclusion criteria for exposed subjects and controls. Exposure was assessed by a —gib
sample" collected at the time of the neurological testing and is not a TWA. Exposure monitoring
indicated a median concentration of tetrachloroethylene of 8 ppm (range: 2-136 ppm). An index
of cumulative exposure to tetrachloroethylene was also developed for each exposed subject by
multiplying the tetrachloroethylene concentration by the number of years worked. Latency to
and duration of vocal response to the stimulus (reading) were measured in each subject after the
presentation of a sequence of words on a computer screen. For each condition, subjects were
asked to say each word immediately or following delays of 0.1 or 0.5 seconds. The test was
performed using a random sequence of concrete or meaningless disyllabic words. These tests
were carried out at the place of employment for dry cleaners and in a clinical setting for controls,
indicating that the investigators were not blinded as to a subject's exposure status. Compared
with the control group, the exposed group had statistically significant longer mean reaction times
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-19 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
and/or vocalization durations under all response conditions (immediate or delayed response) with
either real or meaningless words. Furthermore, statistically significant positive correlations were
observed between cumulative tetrachloroethylene exposure and immediate reading and delayed
reading tasks (r = 0.69 and r = 0.73, respectively). No information on alcohol consumption or
other potential differences between exposed subjects and controls was reported, precluding an
analysis of how these factors may have affected the observed association between
tetrachloroethylene and reaction time.
Sharanjeet-Kaur et al. (2004) examined color vision in 14 workers, ages 24-53 years, in
three dry-cleaning facilities using tetrachloroethylene in Malaysia. This study was part of a
larger study assessing color vision in two other occupationally exposed populations (39 workers
in a factory producing polyethylene resin plastic storage containers and 40 workers
manufacturing polystyrene plastic bags). The paper does not report how facilities were identified
or recruitment methods for study subjects. Furthermore, the paper does not present any
information on tetrachloroethylene concentrations, tetrachloroethylene biomarkers, or exposure
levels in this type of work setting in Malaysia, making it difficult to judge the degree of
exposure. Controls (n = 29)1 were recruited from the support staff of the Universiti Kebangsaan
Malaysia and were age-matched to the age distribution of the dry-cleaning workers (mean age:
33 ± 8.5 years and 33 ± 3.9 years in dry cleaners and controls, respectively). However, dry-
cleaning workers differed from controls on several variables: work duration (mean: 6.7 and
12.6 years in dry cleaners and controls, respectively), hours worked per day (mean: 9.8 and
8.3 in dry cleaners and controls, respectively), cigarette smoking (36 and 7% in dry cleaners and
controls, respectively), and race (50 and 90% Malays in dry cleaners and controls, respectively);
no information is presented on possible differences between dry cleaners and controls in
socioeconomic status. Consent was obtained from all study participants. Visual testing was
carried out at the factory or dry cleaner, for exposed subjects, and at the Optometry Clinic in the
Universiti Kebangsaan Malaysia for control subjects. Thus, the investigators were not blinded to
exposure status during the testing procedure. Distance visual acuity was measured using the
Snellen chart, and near visual acuity was measured using a reading chart. Subjects with poor
visual acuity or with systemic, ocular, or neurological diseases were excluded; the number of
excluded subjects is not specified in the paper. Color vision was assessed binocularly using
Ishihara plates, the Lanthony D-15 test, and the Farnsworth Munsell (FM) 100 Hue test under a
light box at an illumination of 1,000 lux. None of the controls or dry cleaners had color vision
errors with the Ishihara plates. In contrast, errors on the Lanthony D-15 test and FM 100 Hue
test were reported for 6 dry cleaners (43%) and 13 dry cleaners (93%) compared to 0,
1 An additional control group, Control Group I, was included in the paper; this group was age-matched to the other
factory workers included in the study.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-20 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
respectively. Statistical testing of these differences was not presented. Total error scores for the
FM 100 Hue test differed between dry cleaners and controls (p < 0.05). It is difficult to interpret
these findings due to the lack of exposure information on potential tetrachloroethylene exposure
other than job title, and differences between dry cleaners and controls regarding test conditions
and smoking history.
Perrin et al. (2007) evaluated the risk of schizophrenia among a cohort of 88,829 births
born between 1964-1976 in the Jerusalem Perinatal Project, a population-based cohort. Births in
this cohort are linked to the database of Israel's Psychiatric Registry, with cases identified using a
broad definition of schizophrenia-related disorders as recorded as hospital discharge codes.
Diagnoses for individuals with psychosis were validated, and the date of onset was identified as
the date of first psychiatric admission. Of the 88,829 births, 136 offspring were born to parents
identified as having a job title of dry cleaner on the birth certificate; 120 offspring whose fathers
but not mothers were dry cleaners, 20 whose mothers but not fathers were dry cleaners; and 4
with both parents as dry cleaners; 4 of the 136 births had a later diagnosis of schizophrenia. The
relative risk (crude) between schizophrenia and parental employment in dry cleaning was 3.9
(95% confidence interval [CI]: 1.3-9.2) using proportional hazard methods. The investigators
noted risk estimates did not greatly change when fitting proportional hazard models that adjusted
for a number of potentially confounding variables; although adjusted relative risk (RR) estimates
are not reported in the paper. Variables considered as possible confounders were parents' age,
father's social class, duration of marriage, rural residence, religion, ethnic origin, parental
immigration status, offspring's birth order, sex, birth weight, and month of birth. Family history
of mental illness was not included as a covariate; rates of schizophrenia are higher among
relatives of patients than in the general population (Mueser and McGurk. 2004).
4.1.1.2.2. Occupational exposure studies: other settings
Schreiber et al. (2002) reported the findings from investigations using visual tests to
assess neurologic function in two populations: apartment residents1 and day-care employees who
had potential environmental tetrachloroethylene exposure due to close proximity to dry-cleaning
facilities. The study of day-care employees will be discussed in this section because their
exposure would have been of a similar pattern to others in an occupational setting. The day-care
facility, located near Albany, NY, was in a building that also housed a business that performed
dry cleaning. Atmospheric monitoring of the day-care facility before closure of the dry-cleaning
business showed airborne concentrations of tetrachloroethylene ranging from 0.27 to 0.35 ppm,
with median and mean concentrations of 0.32 ppm. Samples obtained at the time of visual
1 The results of the residential study are summarized in Residential Exposure Studies, following this section.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-21 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
testing, 5 weeks after removal of the dry-cleaning machines, approached background
concentrations (range: 0.0012-0.0081 ppm).
Objectives of the investigations were to characterize tetrachloroethylene exposure and to
screen for subclinical neurological effects using a battery of visual function tests (Schreiberet
al.. 2002). All participants signed consent forms. The study included all of the current staff
members of the day-care center (n = 9, all adult females). Controls were age- and gender-
matched acquaintances of the exposed participants, local retail shop employees, NYSDOH
employees, or staff from other local day-care centers with no known tetrachloroethylene
exposure. All subjects in the exposed and control groups were Caucasian (telephone
communication from K. Hudnell, EPA, to D. Rice, EPA, February 2003). Mean age was
27.7 years for control participants and 27.2 years for day-care workers; mean duration of
employment at the day-care center was 4 years. Sociodemographic data, lifestyle factors (e.g.,
personal and passive smoking, alcohol consumption, and exercise), medical history, and
neurotoxicant exposure were obtained by questionnaire. Reported alcohol consumption was
similar (low or moderate) in the exposed and control groups.
Visual function testing consisted of near visual acuity, near visual contrast sensitivity,
and color vision (Schreiber et al.. 2002). Examiners were not blinded as to a subject's exposure
status. In the contrast sensitivity test, luminance varied between the bars in sine-wave fashion,
and each test pattern represented one size of bars or spatial frequency. The bar patterns were
presented at five different spatial frequencies, thereby breaking spatial visual function into its
essential components. The least amount of luminance contrast needed to detect each bar size
was measured. A strength of this study is that the test of contrast sensitivity employed a forced-
choice procedure, providing better reliability and consistency than other approaches.
Multivariate analysis of variance was used to analyze the visual contrast sensitivity data. Color
vision was assessed using the Lanthony D-15 test, with calculation of color confusion index
(CCI) based on the accuracy of the chip placement. Group differences in the CCI were assessed
using two-tailed Student's ^-tests for matched-pair analyses.
The mean measure of visual acuity was 20:22.2 in the exposed day-care workers and
20:26.4 in controls (p = 0.16). There was a statistically significant lower group mean visual
contrast sensitivity score across all spatial frequencies when day-care employees were compared
with the control group (see Figure 4-1). The mean CCI scores were 1.22 and 1.18 in the exposed
day-care workers and controls, respectively (p = 0.39).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-22 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
_ 160
£ 140
w 120
tl 100
>r 80
I 60
E 40
5 20
u
£ 10
1.5 3 G 12 15
Spatial frequency (cycles/degtqq)
Figure 4-1. Visual contrast sensitivity functions for control and exposed
participants in a study of workers in a day-care center located in a building
with a dry-cleaning facility (Schreiber et al.. 2002). The X-axis represents the
frequency of the stimulus bars, with finer bars toward the right. The Y-axis
represents the inverse of the contrast at which the subject could no longer
distinguish the orientation of the bars (threshold). Blue circles (top line) =
controls; red circles (bottom line) = exposed. For any frequency, a higher contrast
sensitivity threshold represents better visual function. Visual contrast sensitivity
was significantly lower across all spatial frequencies in exposed workers at a day-
care center colocated with a dry-cleaning facility compared with their matched
controls.
Although it should be noted that the controls came from a different area (a rural area in
upstate New York) compared to the exposed subjects from New York City, there is little
evidence that degree of urbanity would be related to visual contrast sensitivity. Education has
not been found to be related to performance on the visual contrast sensitivity test (Frenette et al..
1991; Hudnell et al.. 2001; Mergler et al.. 1991; NYSDOH. 2005b; U.S. EPA. 2004).
Additionally, occupation is highly correlated with socioeconomic status (Deonandan et al.. 2000)
and is also not likely to confound the visual contrast sensitivity test.
The Pumpkin Patch Day Care Center Follow-up Evaluation (NYSDOH. 2005a. b, 2010)
examines the effect of tetrachloroethylene exposure on visual function in former students of the
day-care center collocated in a building with a dry-cleaning facility that was studied by Schreiber
et al. (2002). This study is discussed in this section because the children's exposure would have
been of a similar pattern to others in an occupational setting, although exposure ceased 4 years
prior to this study. Children eligible for testing in the current evaluation were enrolled in the
New York State Volatile Organic Chemical (VOC) Registry and had attended the day-care
center. Of the 115 who met this criterion, 27 children with the highest number of hours spent at
the day-care center were invited to participate; 17 children completed vision testing, and
O ControJ • r? = 91
¦ Exposed [ji= 6j
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/2011 4-23 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
13 children completed some or all of the neurobehavioral assessment. Referents (controls) were
children who attended other day-care centers, and were matched to the exposed children by day-
care experience, age, and gender. No information is provided on methods employed for referent
participation. Overall, 17 Pumpkin Patch Day Care Center and 13 comparison children
(13 matched pairs) completed vision testing, and 13 Pumpkin Patch Day Care Center and
13 comparison children (8 matched pairs) completed neurobehavioral testing, consisting of a
battery of tests that assess general intellectual function, attention/information processing speed,
visuospatial ability, reasoning and logical analysis, memory, motor functions, and sensory-
perceptual functions. A parent or guardian completed the Child Behavioral Checklist and a
background history questionnaire. Neurobehavioral function of the 13 Pumpkin Patch Day Care
Center children evaluated in this follow-up study did not differ from that of the 13 referent
children, and Pumpkin Patch Day Care Center children performed better than referent children
on several tests. Visual function testing consisted of visual acuity, far visual contrast sensitivity,
and color vision. Visual contrast sensitivity was determined using the Functional Acuity
Contrast Test distance chart placed 10 feet from the participant under light conditions specified
by the manufacturer. Scores for each eye were recorded on a graph showing a normal range
(90% CI) of visual contrast sensitivity at each spatial frequency. Color vision was assessed using
both the Farnsworth D15 and Lanthony's D-15 tests. Both color vision and contrast sensitivity
tests were performed monocularly. Examiners were not specifically blinded to exposure status,
but this information could have been revealed by the participant during the examination. Using
the Wilcoxon matched-pairs signed-ranks test, Pumpkin Patch Day Care Center children
performed better on the visual contrast sensitivity test compared to referent children. No
significant difference in the proportions of children with abnormal color vision or with children
making major errors, or with CCI scores were seen between Pumpkin Patch Day Care Center and
referent children. Similar results on the vision tests were seen when excluding two pairs who
were <6 years old.
4.1.1.2.3. Residential exposure studies
This section discusses studies of residential exposure scenarios. Residential exposure to
tetrachloroethylene can result in nearly continuous exposure (NYSDOH. 2005b) and is distinct
from the pattern of tetrachloroethylene exposure experienced by the occupational populations.
Altmann et al. (1995) examined neurological effects of long-term exposure to
tetrachloroethylene among residents of Mulheim, Germany, who lived near dry-cleaning shops.
A total of 19 exposed subjects were chosen from a population of 92 individuals living in
neighborhoods close to dry-cleaning facilities. Three criteria were used to select subjects: a
tetrachloroethylene blood level above 0.002 mg/L, a period of living above or next to a dry-
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-24 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
cleaning facility for at least 1 year, and no occupational exposure to organic solvents. The mean
age of the exposed subjects was 39.2 years (range: 27-58 years), and the mean duration of living
near a dry-cleaning facility was 10.6 years (range: 1-30 years). Thirty potential controls (mean
age: 37.2 years, range: 24-63 years) were recruited, mainly from the staff of a public health
office or an institute for environmental hygiene. One or two controls, matched for age (±1 year,
but ±3 years in one case and ±6 years in another case) and gender, were chosen for each exposed
subject. Consent was obtained from all subjects prior to the initiation of testing. Five exposed
(26%) and seven control subjects (23%) were excluded for various medical reasons, including
impaired vision, diseases with potential neuropathy, hypertension, and joint impairment. All
subjects met standards for visual acuity and vibration perception. The final exposed group
included 14 subjects (5 men, 9 women), and the control group included 23 subjects (9 men,
14 women). The two groups did not differ with regard to consumption of alcoholic beverages,
regular medication, smoking, or body mass index. Level of education was divided into three
categories, -k>w," "medium," or "high" (definitions of these categories were not provided). The
number of exposed subjects by education group (low, medium, and high) was 4, 8, and 2,
respectively; the number of controls in these respective groups was 1, 12, and 10, indicating a
considerable imbalance across these strata. The effect of tetrachloroethylene exposure on the
neurophysiological and neurobehavioral measurements was evaluated using linear regression,
adjusting for age, gender, and the three-level education variable.
Visual evoked potentials in response to black-and-white checkerboard patterns were
recorded for all individuals (Altmann et al.. 1995). Vibration perception using a tuning fork—a
crude measure of peripheral neuropathy—was assessed at the ankle. Five tests included in the
Neurobehavioral Evaluation System developed in the United States and adapted for testing on a
German population were used: (1) finger-tapping speed with the index finger of both the
dominant and the nondominant hand; (2) hand-eye coordination using a joystick to follow a sine
wave on a computer screen; (3) a continuous performance test for assessment of vigilance, which
requires a response to a specific stimulus appearing on the computer screen and failure to
respond to other stimuli; (4) simple reaction time, which requires the fastest possible response to
a simple visual stimulus (measured twice); and (5) visual memory on the Benton visual retention
test, which requires a match of a previously displayed stimulus out of several choices after a
short delay interval. All testing was completed in a single 3-hour session; testing times were
selected randomly for both exposed or control subjects.
Blood samples were taken in the exam room immediately before testing (all subjects)
and, if possible, once when the exposed subjects were at home (Altmann et al.. 1995). The mean
blood level for exposed subjects at the examination was 0.0178 mg/L (standard deviation:
0.469 mg/L). For seven of the nine exposed subjects, blood concentrations in samples collected
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-25 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
at home were higher than those in samples collected at the examination. None of the blood
concentrations in the control group exceeded the detection limit of 0.0005 mg/L. For the
exposed subjects (data from 13 apartments), indoor air sampling indicated that the mean (7-day
TWA) air concentration was 0.7 ppm (standard deviation: 1 ppm) and the median was 0.2 ppm.
For the control group, the mean and median values were 0.0005 ppm (standard deviation: 0.0005
ppm) and 0.0003 ppm, respectively. There was a good correlation between home indoor air
concentrations and blood levels of tetrachloroethylene in the exposed subjects (/' = 0.81). The
correlation was much lower when the examination room blood samples were used (r = 0.24).
Altmann et al. (1995) observed statistically significant differences between the adjusted
mean scores of exposed and control subjects on neurobehavioral tests of simple reaction time
(p < 0.05 for the first test andp < 0.01 for the second test), continuous performance (p < 0.05),
and visual memory as tested with the Benton visual retention test (p < 0.05). In all cases, the
exposed subjects had slower response times or more errors than did the unexposed controls. The
degree of change from control was approximately 15-20% for these tests. The potential for
residual confounding by education should be considered, however, as education level was
independently associated with these measures, and use of three categories for education in the
multivariate regression analyses may not fully account for all effects from this covariate,
particularly given the observed differences in education levels among the exposed and control
groups. No statistically significant differences were observed between the performance of the
exposed and control groups on the finger-tapping or hand-eye coordination tests, which are
measures of fine motor function; on visual evoked potentials, which may be less sensitive than
direct measurement of visual function; or on vibration perception at the ankle using a tuning
fork.
Schreiber et al. (2002) examined neurologic function as assessed by visual tests among
apartment residents who had potential environmental tetrachloroethylene exposure due to close
proximity to dry-cleaning facilities.1 The apartment residents lived in two separate buildings in
New York City that each contained a dry-cleaning business. The residential study served as a
pilot for a larger study that is investigating visual effects among tetrachloroethylene-exposed
residents. The exposed group consisted of 17 subjects (11 adults between the ages of 20 and 50,
2 adults over the age of 60, and 4 children, ages 6-18) from six families residing for a median of
6 years in two apartment buildings in New York City2 (Schreiber et al.. 2002). Preliminary
1 Another study by Shreiber et al. (20021 of day-care staff from a center colocated with a dry-cleaning facility, using
a similar testing protocol, was described in the Occupational Exposure Studies—Other Settings section.
2 Study subjects were identified through several methods: (1) both families in the first building (Building A) had
been referred to the NYSDOH for information about participating in the study by Consumer Union/Hunter College
researchers, (2) one family in the second building (Building B) had previously contacted NYSDOH about exposure
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-26 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
monitoring of these buildings indicated tetrachloroethylene concentrations were elevated
compared to eight other buildings also monitored by the NYSDOH. Exposed residents were
from an affluent, English-speaking, Caucasian population living near New York City's Central
Park (telephone communication from K. Hudnell, EPA, to D. Rice, EPA, February 2003).
Exposed participants were generally unaware of the tetrachloroethylene exposure, although some
study participants noted tetrachloroethylene-like odors prior to the study. Controls were
recruited from among NYSDOH Albany, New York employees and their families. All controls
were Caucasian, except for one Asian individual, and were age- and sex-matched to exposed
apartment residents. In some cases, more than one control participant was matched to an
exposed subject. Mean age was 34.5 years for exposed apartment residents and 33.2 years for
control subjects.
The assessment of tetrachloroethylene exposure of residents consisted of concentrations
in indoor air and personal air samples, exhaled breath, and blood, which were collected at the
time of visual testing. Testing was performed during a period of active dry cleaning for four of
the families and 1 month after closure of the facility for the remaining two families in the
residential study. Adult residents also provided urine samples, which were analyzed for
tetrachloroethylene as well as for three products of its metabolism: TCA, trichloroethanol, and
the urinary acetyl metabolite. Ambient concentrations of tetrachloroethylene from 1 to 3 months
before the date of visual testing, when active dry cleaning was occurring in both apartment
buildings, were available for all subjects. Median concentrations in these samples were
0.21 ppm (mean: 0.36 ppm; range: 0.1-0.9 ppm). Airborne tetrachloroethylene concentrations
had decreased in samples collected at the time of visual testing; median tetrachloroethylene
concentration was 0.09 ppm (mean: 0.18 ppm; range: 0.01-0.78 ppm). Tetrachloroethylene
levels in blood correlated well with levels in room air, personal air, and breath.
All participants, or their guardians in the case of children, signed consent forms prior to
study commencement. Information on sociodemographics; lifestyle factors such as exposure to
direct or passive smoke, alcohol consumption, and exercise; medical history; and neurotoxicant
exposure in addition to the visual tests was obtained by questionnaire from both study
populations and their controls. Exposed participants had no known exposure to other
neurotoxicants, ongoing illness, current use of neuroactive drugs, or a medical history indicative
of neurologic dysfunction. Reported alcohol consumption (low to moderate) was similar in the
adult exposed and control groups, and the Profile of Moods test scores of all residential exposed
subjects were within normal limits. However, two of the four children had medically verified
diagnoses of learning disabilities or developmental delays (NYSDOH. 2004).
concerns and desired to participate in a study, and (3) three other families in Building B were recruited by a
participating family (NYS OAG. 20041.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-27 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
As described in the previous discussion of Schreiber et al. (2002) (see Occupational
Exposure Studies—Other Settings section), visual function testing consisted of near visual
acuity, near visual contrast sensitivity, and color vision, and the investigators were not blinded as
to a subject's status as either exposed or nonexposed. The mean measure of visual acuity was
20:27.7 in exposed residents and 20:22.8 in controls (p = 0.12). Group mean scores for visual
contrast sensitivity across spatial frequencies were statistically significantly lower in exposed
residents than in controls, indicating poorer visual function in the exposed groups (see
Figure 4-2). An exposure-response analysis did not show an association between poorer
performance and increasing tetrachloroethylene concentration. CCI scores (a measure of color
vision) of the exposed group were lower than those of controls, but the difference was not
statistically significant (mean: 1.33 and 1.20 in exposed and control groups, respectively,
p = 0.26).
I GO
140
120
100
00
GO
40
20
10
IE 3 6 12 18
Spatial frequency (cycles/degree)
tl
>•
"5
:=
4+
s
O
Figure 4-2. Visual contrast sensitivity functions for control and exposed
participants in residential exposure study (Schreiber et al.. 2002). The X-axis
represents the frequency of the stimulus bars, with finer bars toward the right.
The Y-axis represents the inverse of the contrast at which the subject could no
longer distinguish the orientation of the bars (threshold). Blue circles (top line) =
controls; red circles (bottom line) = exposed. For any frequency, a higher contrast
sensitivity threshold represents better visual function. Visual contrast sensitivity
was significantly lower across all spatial frequencies in exposed residents of
apartments in building with dry-cleaning facilities compared with their matched
controls.
A larger study of the effect of tetrachloroethylene exposure on visual function in
residents living in buildings colocated with a dry-cleaning establishment was conducted by the
(NYSDOH. 2005a. b, 2010). This study, the New York City Perc Project, did not include the
subjects in Schreiber et al. (2002) and employed different methods for testing visual contrast
o Contrd i>= 17|
4 Exposed li? = 1Ti
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/2011 4-28 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
sensitivity and color vision. Study design and protocols were approved by Institutional Review
Boards at the NYS DOH and other collaborating institutes (Mt. Sinai Medical Center and CDC).
Sixty-five households in 24 residential buildings with dry cleaners using tetrachloroethylene on-
site, and 61 households in 36 buildings without dry cleaners were recruited. Health outcome and
tetrachloroethylene concentrations as measured from indoor air monitoring and in exposed
subject's breath and blood were obtained over the period from 2001-2003. McDermott et al.
(2005) presents exposure monitoring findings from the dry-cleaner households.
Subjects were identified in buildings from eleven contiguous zip code areas surrounding
Central Park, New York City. Household eligibility criteria included the presence of at least one
adult (20-55 years old) and one child (5-14 years old), so as to assess whether residential
tetrachloroethylene exposure would disproportionately affect children. Initial monitoring
indicated few residences in dry-cleaner buildings with elevated indoor air concentrations of
tetrachloroethylene above the current NYS DOH residential air guideline of 0.015 ppm
(0.1 mg/m3). The study area was broadened to include buildings that had been the subject of a
resident complaint and to include buildings in additional zip codes, primarily characterized by
lower socioeconomic status or higher percentage of minority residents. Of the 1,261 dry-cleaner
and 1,252 reference households contacted, 132 dry-cleaner households and 175 reference
households included age-eligible adult-child pairs. A total of 65 dry cleaner (67 adults,
68 children) and 61 referent households (61 adults, 71 children) participated in the study. The
socioeconomic status characteristics, residence duration, education level, age, and smoking and
alcohol use were similar in the adult residents of reference buildings and the residents of
buildings with dry cleaners. Differences between child residents in gender or residence duration
are not apparent, but the highest exposure group is about a year younger and has about one less
year of education than children in the other exposure groups. All participants or their guardians
signed voluntary consent forms prior to study commencement.
NYSDOH staff visited participants in their residences to collect 24-hour indoor air
samples and breath samples, and to give adult participants a questionnaire seeking information
on residential, occupational, and medical history for themselves and their children. Indoor air
tetrachloroethylene concentrations had decreased since 1997, the period of the pilot study
(Schreiber et al.. 2002). and ranged up to around 0.77 ppm (5 mg/m3) with a geometric mean of
0.005 ppm (0.035 mg/m3) in apartment buildings colocated with a dry cleaner. Monitoring was
carried out using passive monitoring badges. In comparison, tetrachloroethylene concentrations
in buildings without dry cleaners ranged up to 0.014 ppm (0.09 mg/m3) with a geometric mean of
0.0004 ppm (0.003 mg/m3). Both breath and blood tetrachloroethylene levels were significantly
(p < 0.05) correlated with indoor air concentrations for adult and for child subjects of dry-
cleaning buildings. LODs were 5 |ig/m3 air and 0.048 mg/mL blood. Air, breath, and blood
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-29 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
tetrachloroethylene concentrations were inversely correlated with income and were higher
among minority compared to nonminority subjects. Participants received financial compensation
after completing the home visit ($50.00) and ophthalmology clinic visit ($50.00).
Ophthalmologic examinations and visual function tests were given to study participants
at the Mt. Sinai Medical School of Medicine Department of Ophthalmology research clinic. The
final report does not describe whether examiners were or were not blinded as to a subject's
exposure status (NYSDOH. 2005a). The examination included determination of past ocular and
medical history; measurement of visual acuity, pupil size, extraocular motility, and intraocular
pressure; and anterior and posterior segment exams. Subjects with abnormalities or taking
medications that could influence visual contrast sensitivity and/or color vision were excluded
from further testing. Furthermore, visual functional tests for some children were excluded from
the statistical analysis because of their young age or because they were identified by their parents
as learning disabled or having attention deficit hyperactivity disorder. Visual contrast sensitivity
was determined using the Functional Acuity Contrast Test (FACT) distance chart placed 10 feet
from the participant under light conditions of 68-240 cd/m2. These testing conditions differ
from those employed by Schreiber et al. (2002) in their residential study where visual testing was
carried out, assessing near-contrast sensitivity.
Adults and children demonstrated a ceiling effect with visual contrast sensitivity
performance, i.e., a maximum score at 1.5, 3, 6, 12, and 18 cycles per degree (cpd) is achieved
by some study participants. Visual contrast sensitivity scores among adults were not correlated
with any socioeconomic status factor or personal characteristics (smoking, alcohol use, education
level, duration of residence). Among all children, poorer visual contrast sensitivity at 1.5, 3, and
6 cpd was significantly correlated with speaking primarily Spanish at home.
Analyses examining relationships between tetrachloroethylene and visual function were
conducted using three categories of exposure: the referent exposure group (background exposure,
living in a building without a dry cleaner, geometric mean: 2.9 [j,g/m3 [0.0004 ppm], range:
1.5-4.2 |ig/m3 [0.0002-0.0006 ppm]); <100 |ig/m3 [geometric mean: 11.6 |ig/m3 {0.002 ppm},
range: 4.2-42.0 |ig/m3 {0.0002-0.006 ppm}]; and >100 |ig/m3 [geometric mean: 477.9 |ig/m3
{0.07 ppm}, range: 268.9-735.3 |ig/m3 {0.04-0.11 ppm}].l A decreasing trend (p < 0.05) was
observed across these three exposure groups and the proportion of adults achieving the
maximum contrast sensitivity score at 6 cpd (28.3, 14.3, and 8.3% in the referent, <100 and
>100 |ig/m3 groups, respectively). This pattern was also seen in analyses stratified by race or
ethnicity, or by income, although the smaller sample sizes resulted in larger p-values (from 0.09
to 0.30) for each of the individual strata. In children, decreasing scores were seen at 6 cpd (43.4,
1 100 |ig/m3 = 0.015 ppm.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-30 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
33.3, and 18.2% in the referent, <100, and >100 |ig/m3 groups, respectively, trend: p = 0.05) and
12 cpd (37.7, 33.3, and 0.0% in the referent, <100, and >100 |ig/m3 groups, respectively, trend:
p = 0.02). These effects were limited to minority and low income children in the ethnicity and
income-stratified analyses.
Results from logistic regression analyses further support susceptibility of children but not
adults to an adverse effect of tetrachloroethylene exposure on visual contrast sensitivity.
Whereas adult visual contrast sensitivity in the worse eye at 6 or 12 cpd was not significantly
influenced by any measure of tetrachloroethylene exposure, visual contrast sensitivity
performance at 12 cpd among children was significantly influenced (p < 0.05) by
tetrachloroethylene concentrations in either indoor air or in blood; i.e., a lower percentage of
children achieved a maximum visual contrast sensitivity score with higher tetrachloroethylene
exposure. Odds ratio estimates were 2.64 (95% CI: 1.41, 5.52), 3.37 (95% CI: 1.44, 9.29), and
3.54 (95%) CI: 0.94, 17.79) for the association between visual contrast sensitivity performace in
the worse eye at 12 cpd and indoor tetrachloroethylene, exhaled breath tetrachloroethylene at
home, and blood tetrachloroethylene, respectively. The logistic regression models examining
visual contrast sensitivity findings were adjusted for ethnicity or race and age, and, in adults,
smoking and alcohol use.
Color vision was assessed biocularly using both the Farnsworth D-15 test (differentiates
between strong/moderate and mild/normal CCI) and Lanthony's D-15 test (differentiates
between normal and mild CCI). Both tests were administered under light conditions specified by
the manufacturer. Analyses were carried out using the proportion of subjects with no errors,
comparing quantitative differences in CCI, and logistic regression modeling to assess
associations between tetrachloroethylene exposure measures and occurrence of any major errors.
A high proportion of adult and child participants scored perfectly on both the Farnsworth and
Lanthony color vision tests. Lower annual household income, being a member of a minority
group, speaking primarily Spanish at home, and fewer years of education were all significantly
associated with increased CCI on both color vision tests. Tetrachloroethylene measures of
exposure were unrelated to color vision performance among adults; however, similar to visual
contrast sensitivity performance, children appear to be a more susceptible population. There
were no differences between exposure groups among adults or children in the percentage of
subjects with major errors on both color vision tests. A comparison of mean CCI between
exposure groups showed that children in the high-exposure category performed worse (mean:
CCI of 1.3, range: 1.0-1.9) compared with children in the low-exposure category (mean: CCI of
1.1, range: 1.0-1.7) and compared with referent children (mean: CCI of 1.2, range: 1.0-2.0) on
the Lanthony test; the test for trend for the three exposure groups was statistically significant
(p < 0.05). Performance (mean CCI) on the less sensitive Farnsworth test was not associated
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-31 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
with tetrachloroethylene exposure in either adults or children. Moreover, for children,
tetrachloroethylene in breath was significantly associated (p < 0.05) with making one or more
major errors on the Lanthony color vision test in logistic regression analyses that adjusted for the
effects of age and gender. Logistic regression analyses examining color vision and other
tetrachloroethylene measures such as indoor tetrachloroethylene concentration or breath
concentration were not discussed in NYSDOH (2005a). The higher mean difference in CCI
between children and adults in the highest exposure category (>0.015 ppm or >100 |ig/m3)
compared with referents was statistically significant. Children in the high-exposure group were a
year younger than in other exposure groups; age was correlated with CCI and with
tetrachloroethylene exposure in this study. The highly correlated variables and the few numbers
of children in the high exposure group limit analysis of age effects on the association between
breath tetrachloroethylene concentration and CCI.
In summary, this study adopts a different approach than Schreiber et al. (2002) to assess
vision, using far vision methods as opposed to the near vision methods of Schreiber et al. (2002).
For both contrast vision and color vision, a number of analyses in (Kaufman et al.. 2009;
NYSDOH. 2005a. 2010) are suggestive of vulnerability among children. Exposure to
>0.015-ppm (> 100-|ig/m3) tetrachloroethylene was highly correlated with race and children's
age, and the sample sizes in the highest exposure group, especially in higher income,
nonminority groups, makes it difficult to fully examine possible effects of income, race, and age
on vision. However, association of tetrachloroethylene exposure >0.015 ppm (>100 |ig/m3) with
visual deficits suggests a susceptibility of the children studied.
4.1.1.2.4. Oral exposure studies
Risk of learning and behavioral disorders was evaluated in relation to prenatal and
postnatal exposure to tetrachloroethylene in Cape Cod towns with a contaminated water
distribution system during 1969-1983 (Janulewicz et al.. 2008). Mothers reported
developmental and educational histories and learning and behavioral disorders in self-
administered questionnaires returned during 2002-2003. Developmental risks were evaluated in
relation to the amount of tetrachloroethylene delivered to each subject's residence during the
prenatal period (from the month and year of the last menstrual period through the month and year
of the birth) and during the early postnatal period (from the month and year of the birth through
the month and year of the 5th birthday). Prenatal and postnatal exposures were evaluated
separately in generalized estimating equation regression models. After excluding 404 subjects
because they had an attribute with a known association with the outcomes under study, there
were 2,086 children in the final data set. Of these, 842 and 1,244 children had no and any
prenatal exposure, respectively, and 760 and 1,326 children had no and any postnatal exposure,
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-32 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
respectively. Exposed and unexposed children were similar with respect to demographic
characteristics and behaviors. Low- and high-exposure categories were developed for the
9-month prenatal period and 5-year postnatal period using the number of grams of
tetrachloroethylene that corresponded to an average drinking water concentration of 40 (J,g/L, the
action level used in 1980, as a cutpoint. The authors reported that no meaningful associations
were observed between prenatal exposure and receiving tutoring for reading or math, being
placed on an Individualized Education Plan, or repeating a school grade. Increased odds ratios
were noted among subjects with low exposure compared to no exposure for receiving a diagnosis
of attention deficit disorder or hyperactivity disorder, special class placement for academic or
behavioral problems, or lower educational attainment (high school graduate or less). However,
odds ratios were not markedly increased for subjects with high exposure (<1.1). For example, in
generalized estimating equations models adjusted for maternal age, race, and education, child's
sex, and prematurity and/or low birth weight, the odds ratio for attention deficit disorder was 1.4
(95% CI: 0.9-2.0) among subjects with low prenatal exposure and was 1.0 (95% CI: 0.7-1.6)
among subjects with high prenatal exposure. For postnatal exposure, no associations were
observed for receiving tutoring for reading or math, special class placement for academic or
behavioral problems, repeating a grade in school, or lower educational attainment. The same
pattern of risk with exposure level also was observed for low and high postnatal exposure
compared to no exposure. For example, the adjusted odds ratio for attention deficit disorder was
1.3 (95%) CI: 0.9-1.9) among subjects with low postnatal exposure and was 1.0 (95%>
CI: 0.6-1.7) among subjects with high postnatal exposure.
4.1.1.3. Summary of Neuropsychological Effects in Low- and Moderate-Exposure
Studies
A summary of neuropsychological effects seen in chronic occupational or residential
exposure studies of tetrachloroethylene is shown in Table 4-2 and discussed by domain below.
Several studies (Altmann et al.. 1995: Echeverria et al.. 1995: NYSDOH. 2005a. b, 2010:
Schreiber et al.. 2002: Storm et al.. In Press) employed multiple measures of exposure (indoor air
monitoring, personal monitoring, and in some cases, biological monitoring). Although some
variation is expected and was seen in individual studies (Altmann et al., 1995, for example), the
correlation between tetrachloroethylene concentration as assessed from indoor air monitoring or
personal monitoring
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-33 DRAFT-DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-2. Summary of effects of chronic tetrachloroethylene exposure in humans seen in studies of
neuropsychological function3
(Reference), n exposed,
mean or median
exposure(s)
Visual domain"
Cognitive domain (executive function, attention)3
Motor"
Spatial
vision
(VCS)
Color
visionb
YEP
Visuo-
spatial
memory0
Vigilance
Trail-
making
Digit
span,
symbol
Cancellation
Information
processing11
Simple
reaction
time
Fine
motor
function
Occupational exposures—dry-cleaning settings
Lauwervs etal. (19831
n = 26, 21 ppm
*
Seeber (1989), n= 101,
12 and 53 ppm
+
+
+
+
—
Naskatsuka et al. (1992).
n = 64, 13 ppm
—
Ferroni et al. (19921
n = 60, 15 ppm
+
—
—
+
—
Cavalleri et al. (19941
n = 35, 7 ppm
+
Cavalleri et al. (1994
follow-uo: 19981 n = 33.
4 ppm
+
Echeverriaetal. (19951
n = 65, 11, 23, 41 ppm
+
—
—
Echeverria etal. (19941
n = 173, <0.2, 3, 9 ppm
+
—
—
Smnatonda et al. (19971
n = 35, 8 ppm
+
Sharanjeet-Kaur et al.
(20041 n = 14. not
reported
+
a -r
>1
to 0\
2 ^
3
~n
H
O
O
-------
a
o
<
s?
cs
>!
o
^3
§•
rs
st
3
>!
sT
St
o
>1
>!
09
o
>!
Si
Si
s
O
Sfc
I
§•
K) On
2
Table 4-2. Summary of effects of chronic tetrachloroethylene exposure in humans seen in studies of
neuropsychological function (continued)
(Reference), n exposed,
mean or median
exposure(s)
Visual domain"
Cognitive domain (executive function, attention)3
Motor"
Spatial
vision
(VCS)
Color
visionb
YEP
Visuo-
spatial
memory0
Vigilance
Trail-
making
Digit
span,
symbol
Cancellation
Information
processing"1
Simple
reaction
time
Fine
motor
function
Occupational exposures—other settings
Schreiber et al. (2002).
Day-care workers n = 9,
0.32 ppm
+
Residential exposures
Altmann etal. (1995).
n = 19, 0.7 ppm
—
+
+
+
—
Schreiber et al. (2002).
n = 17 (13 adults and 4
children), 0.4 ppm
+
+
(trend)
McDermott et al. (2005);
NYSDOH (2010; Storm
etal.. In Press), n = 68
children (C), n = 67
adults (A), 0.005 ppm
+ (C),
-(A)
-(C),
-(A)
a + denotes effects seen (i.e., worse performance) in exposed group; — denotes no effect or better performance in exposed group; —* denotes better performance
in the exposed group (before shift measure); blank cell denotes test not performed.
-{^ b Based on Lanthony D-15 test, except for Nakatsuka et al. (1992). who used a less sensitive version of this test.
^ 0 Tests include digit reproduction (in Seeber. 1989): switching, pattern memory, and pattern recognition (in Echeverria et al.. 1994: Echeverria et al.. 1995). and
Benton (in Altmann et al.. 1995).
d Tests include choice reaction time, perceptual threshold, and vocal reproduction to reading stimuli.
O
~n
H
O
O
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
and biological metrics such as blood tetrachloroethylene concentration was quite strong,
suggesting indoor air concentration as a reasonable exposure metric. Many studies did not
include exposure monitoring of individual subjects, and the statistical analyses compare groups
using Mests or chi-square tests (Terroni et al.. 1992; Seeber. 1989; Spinatonda et al.. 1997).
Dose response and multiple logistic regression analyses are statistically more powerful, and five
studies observed correlation or association between various tetrachloroethylene exposure
measures and specific neurobehavioral tests (Altmann et al.. 1995; Cavalleri etal.. 1994;
Echeverria et al.. 1995; NYSDOH. 2010; Storm et al.. In Press).
4.1.1.3.1. Visual function domain
Color vision and visual contrast sensitivity are the visual domains that have been
observed to be affected by chronic exposure to tetrachloroethylene (see Table 4-2).
Only Schreiber et al. (2002) and NYSDOH (2005a) assessed spatial vision (VCS, visual
contrast sensitivity), an effect reported for exposure to other solvents (Bowler et al.. 1991;
Broadwell etal.. 1995; Campagna et al.. 1995; Donoghue et al.. 1995; Frenette et al.. 1991;
Hudnell etal.. 1996a; Hudnell et al.. 1996b; Mergler et al.. 1991; Schreiber et al.. 2002). In
Schreiber et al. (2002). visual contrast sensitivity deficits in subjects (mostly adults) with normal
visual acuity were observed at low-exposure concentrations in residential populations, and in
NYSDOH (2005a); evidence of these effects were seen in children but not in adults. Exposure
levels were lower in the latter study [mean: 0.4 ppm and geometric mean: 0.005 ppm in
Schreiber et al. (2002) and NYSDOH (2005a). respectively]. Potential bias and confounding
could have been introduced, however, from a lack of blinding of testers and, in the latter study,
the inability to control for socioeconomic and other factors that were highly correlated with
higher tetrachloroethylene exposures.
Deficits in blue-yellow color vision, a well established effect of solvents, were observed
in dry-cleaning workers in Italy in Cavalleri et al. (1994) and in a follow-up study (Gobba et al..
1998) of this population. Cavalleri et al. (1994) specifically noted that the color vision testing
was conducted by examiners who were blinded to exposure level of individual study participants,
and the study participants were well-matched in terms of age, smoking, and alcohol use. Mean
TWA exposure levels were approximately 7 ppm among the dry cleaners in Cavalleri et al.
(1994). There also was a statistically significant positive correlation (p < 0.01) between TWA
air concentrations and the CCI (r = 0.52), which remained after multivariate analysis considered
previous tetrachloroethylene exposure, duration, age, number of cigarettes a day, and daily intake
of alcohol as covariates. This type of color vision deficit was not seen in the dry cleaners study
by Nakatsuka et al. (1992). but the form of the color vision test used in the latter study, the
Lanthony 15, is less sensitive to mild and moderate changes in color vision compared with the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-36 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
desaturated version of the test (Lanthony D-15) used in the other studies (Lanthony. 1978).
Effects on color vision were also seen among 14 dry cleaners in the small study in Malaysia by
Sharanjeet-Kaur et al. (2004). but the lack of exposure information (other than job title), and
differences between dry cleaners and controls regarding test conditions and smoking habits
indicate that this study should provide little weight in the overall conclusions regarding color
vision. Two other small studies also reported lower scores on the Lanthony D-15 color vision
test in exposed groups compared with controls, but the differences were not statistically
significant: in a study of residents living above dry cleaners (mean tetrachloroethylene exposure
during active dry cleaning = 0.4 ppm), the mean CCI scores were 1.33 and 1.20 in 17 exposed
and 17 control groups, respectively (p = 0.26); in a study of workers in a day-care center located
in a building with a dry-cleaning business (mean tetrachloroethylene exposure: 0.32 ppm), the
mean CCI scores were 1.22 and 1.18 in the exposed day-care workers and controls, respectively
(p = 0.39) (Schreiber et al.. 2002). The follow-up study of NYSDOH (2005a) further suggests
tetrachloroethylene effects on color vision, particularly in children.
Peer-consultation comments on EPA's earlier draft Neurotoxicity of Tetrachloroethylene
(Perchloroethylene) Discussion Paper (U.S. EPA. 2003) noted that the deficit in contrast
sensitivity could reflect a sensitivity of the visual system to tetrachloroethylene, or it may be that
this test was relatively more sensitive than other vision tests or tests used for other domains (U.S.
EPA. 2004). Furthermore, the peer consultants also suggested that contrast sensitivity loss may
reflect impaired function throughout the brain, because contrast sensitivity is affected by retinal,
optic nerve, or central brain dysfunction (U.S. EPA. 2004). Nonetheless, drawing strong
conclusions from these studies is difficult, particularly in light of the paucity of data on this test
in occupational populations with higher exposure concentrations and in animal studies.
Although Altmann et al. (1990: 1992) reported alterations in visual evoked potentials
(p < 0.05) with 4-hour acute exposure at 10 ppm, they were not altered in residents exposed
chronically to a median of around 1-ppm tetrachloroethylene (Altmann et al.. 1995). Acute and
chronic exposures are of different patterns—short-term peak exposure versus longer-duration
exposure—and, therefore, may result in a different pattern of response.
4.1.1.3.2. Cognitive domain
Cognitive domains affected by tetrachloroethylene include visuospatial memory,
attention, vigilance (continuous performance), and speed of information processing (see
Table 4-2). Effects on visuospatial memory are of particular interest, given the similar results in
the studies that examined this type of effect in occupational (Echeverria et al.. 1994: Echeverria
et al.. 1995: Seeber. 1989) or residential (Altmann et al.. 1995) settings, and given similar reports
for other solvents (Daniell et al.. 1999: Morrow et al.. 1990). Echeverria et al. (1995) found
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-37 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
effects among 23 dry cleaners classified as having a high chronic exposure (based on type of
shop Job title, and years of employment) on tests of pattern memory, visual reproduction, and
pattern recognition in the absence of effects on attention (digit symbol and digit span) or
executive function (Trailmaking A and B). Further, Echeverria and colleagues (1994) confirmed
these findings in an independent sample of dry cleaners categorized as having high lifetime
chronic exposure and whose current exposure level was 9 ppm, 8-hour TWA; the exposure level
of 9 ppm is not representative of past chronic exposure levels because of changes occurring in
the industry (i.e., switching from wet-transfer to dry-to-dry machine). Seeber (1989) also
reported impaired visuospatial recognition in a low exposure (mean: TWA 12 ppm) and a high
exposure group (mean: TWA 53 ppm), and Altmann et al. (1995) observed deficits on a test of
visuospatial function in residents with much lower exposure concentrations (mean 0.7 ppm) than
those of the occupational studies. All of these studies except Altmann et al. (1995) reported that
investigators were blinded to knowledge of the exposure level of the subject. These studies
provide strong weight, given the numbers of subjects and their use of appropriate statistical
methods, including adjustment for potentially confounding factors that may be relevant for
measures of the cognitive domain. For example, Seeber (1989) adjusted for age, gender, and a
measure of intelligence (alcohol was examined but not shown by these investigators as
confounding the association between tetrachloroethylene and cognitive performance), and a
variety of potential confounders were evaluated by Echeverria et al. (1994: 1995). It should be
noted, however, that residual confounding from education level differences between exposed and
referent subjects may still be present in Altmann et al. (1995).
The results pertaining to cognitive measures other than visuospatial memory are
somewhat mixed. Altmann et al. (1995) and Ferroni et al. (1992) assessed vigilance using a
continuous performance procedure in which the subject faces a screen that presents one of
several different stimuli at random intervals. The subject must make a response to a specified
stimulus and not to the others. This test measures sustained attention and is correlated with
performance on tests of executive function. Both studies found deficits as a result of
tetrachloroethylene exposure on this task. Seeber (1989) found effects on two tests of attention
(cancellation d2 and digit symbol) that are subsets of the Weschler IQ tests and were designed to
be sensitive to performance within the normal range. These investigators also found positive
effects on a visual scanning test that is usually used to assess laterality of brain damage but has
also proved sensitive to toxicant (lead) exposure (Bellinger et al.. 1994). In contrast, Echeverria
et al. (1995) and Ferroni et al. (1992). as described in NYSDOH (1997) did not find effects on
digit span, which is given as a test of attention and memory, or digit symbol, despite higher
levels of exposure than in Seeber (1989). Speed of information processing was assessed in two
studies: Seeber (1989) and Spinatonda et al. (1997). Seeber used two tasks: recognition and
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-38 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
choice reaction time. Effects were observed in both groups on a task requiring recognition of
briefly presented stimuli. In a choice reaction time task, effects were borderline in the lower-
exposure group and negative in the higher-exposure group, with no exposure-response
relationship. Spinatonda et al. (1997) observed longer mean reaction times and/or vocalization
durations to vocal and visual stimuli.
Two studies—an occupational study with relatively higher exposure (Ferroni et al.. 1992)
and the Altmann et al. (1995) residential study—also assessed simple reaction time, a task that
uses a motor response and demands a relatively modest amount of attention. In both studies,
lower performance [ranging from an increase in reaction time from 24 (11%, 102 mg/m3)
(Ferroni et al.. 1992)1 to 50 ms [20%, 4.99 mg/m3 (Altmann et al.. 1995)1 was seen among the
exposed workers compared with referents. A third study, Lauwerys et al. (1983). reported better
performance on simple reaction time in exposed workers compared with referents when
measured before a work shift but not when measured after work.
4.1.1.3.3. Motor function domain
Tetrachloroethylene exposure has not been reported to affect fine motor tests. Seeber
(1989). Ferroni et al. (1992). and Altmann et al. (1995) each assessed fine motor control using
various instruments, and all three found no significant decrements in fine motor performance.
4.1.1.3.4. Other clinical tests and conditions
A clinical neurological examination that includes the Romberg test, neuroradiological
examination, neurophysiological tests such as EEGs, and nerve conduction tests or other tests for
peripheral neuropathy have seen limited use for assessing neurotoxicologic effects in
tetrachloroethylene-exposed populations. Mental disease and behavioral disorders of neurologic
origin have not been well studied with respect to environmental factors. Perrin et al. (2007). who
reports an association between schizophrenia and parental exposure in dry cleaning, is the only
such study. A fourfold increased risk of schizophrenia was seen among offspring. However, in
a small study, Janulewicz et al. (2007) did not observe an association between prenatal or early
postnatal drinking water exposure to tetrachloroethylene and disorders of learning, attention, and
behavior. Therefore, other studies are needed to understand the role of parental
tetrachloroethylene exposure in the development of mental disease and behavioral disorders in
children.
4.1.2. Animal Studies
Tetrachloroethylene exposure in experimental studies in animals results in general
CNS-depressant activity (decreased activity, anxiolytic behavior, lethargy), impairment in
balance and motor coordination, cognitive defects, sleep cycle changes, and changes in visual
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-39 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
function and nerve conduction velocity. These changes have been observed following either an
inhalation or oral/intraperitoneal (i.p.) exposure. In addition to these effects, several effects on
brain pathology including DNA and RNA level changes, changes in neurotransmitter levels such
as acetylcholine and glutamate, and changes in brain fatty acid composition, have been observed.
Some studies also document potential developmental neurotoxicity consequences following
exposure to tetrachloroethylene during the gestation period.
4.1.2.1. Inhalation Studies
The animal inhalation neurotoxicity studies are summarized in Table 4-3 and described in
more detail below. Neurobehavioral, neurophysiological, and developmental neurotoxicity
effects have been reported following tetrachloroethylene exposure. Two neurobehavioral studies
observed that there was an increase in motor activity following a 1-hour exposure in NMRI mice
at 90 ppm and higher (Ki ell strand et al.. 1985). and there was a decrease in immobility in Swiss
OF1 mice at 649 ppm and higher during 4 hours of exposure (De Ceaurriz et al.. 1983). A more
recent neurobehavioral study examined effects of Long-Evans rats in a signal detection test and
reported decreased sustained attention as a measurement of decreased trial completions and
increased reaction time during an hour exposure to 500 ppm or higher (Oshiro et al.. 2008). In
F344 rats, significant changes in FEP latency and amplitude following a 12-week repeated
exposure to 800 ppm or higher were reported by Mattsson et al. (1998). and in Long-Evans rats,
changes in visual evoked potential amplitudes during an acute (60-120 minutes) exposure to
250 ppm or higher were reported by Boyes et al. (2009). Developmental neurotoxic effects were
noted in three studies (Nelson et al.. 1980; Szakmary et al.. 1997; Tinston. 1994) where changes
such as decreases in muscular strength and exploratory behavior as well as other behavioral
habits were significantly different from nonexposed litters. Finally, there were many changes in
brain pathology as noted by decreased brain weight, brain DNA levels, and changes in
neurotransmitter levels (Briving et al.. 1986; Honma et al.. 1980a; Honmaetal.. 1980b; Karlsson
et al.. 1987; Kiellstrand et al.. 1984; Kyrklund et al.. 1984; Kyrklund and Haglid. 1991;
Kyrklund et al.. 1987. 1988. 1990; Rosengren et al.. 1986; Savolainen etal.. 1977a; Savolainen
et al.. 1977b; Wang et al.. 1993).
4.1.2.1.1. Neurobehavior
De Ceaurriz et al. (1983) exposed male Swiss OF1 mice (n = 10 per exposure group) to
596-, 649-, 684-, or 820-ppm tetrachloroethylene for 4 hours. Immediately following exposure,
the mice were immersed in a cylinder filled with water, and the duration of immobility was
observed for 3 minutes. The term —behaviral despair" has been coined for this initial
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-40 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-3. Summary of animal inhalation neurotoxicology studies
Subjects
Effect
NOAET /T OAET a fnnmt
Reference
Neurobehavioral studies
Swiss OF1 mice, males
10/dose
Decreased duration of immobility
596. 649. 684, 820; 4 h
De Ceaurriz et
al. C1983)
NMRI mice, males
(n = 27 for 90, 320, 400,
600; n = 14 for 800,
1,200, 1,800, 3,600 ppm)
Increased motor activity
2Q, 3,600; 1 h
Kjellstrand et al.
(1985)
Long-Evans rats, males
(n = 12 total; animals
served as own controls)
Increased number of false alarms,
increased reaction time, and
decreased trial completions in a
signal detection task measuring
sustained attention
0,5QQ, 1,000, 1,500; 60 min
Oshiro et al.
(2008)
Neurophysiological studies
F344 rats
Pilot study: male
10/dose
Follow-up study: males
and females
12/sex dose
Changes inFEP, SEP, EEG
Increased amplitude and latency
in late component of FEP
0, SQQ; 4 d, 6 h/d
50, 200, 2QQ;
13 wk, 6 h/d, 5 d/wk
Mattsson et al.
(1998)
Long-Evans rats, males
(n = 9-10/exposure)
Decreased F2 amplitude in the
steady state YEP
0,250,500, 1,000 for 1.5 h
Boyes et al.
(2009)
Developmental neurotoxicity studies
S-D rats
pregnant females
13-21 litters/dose;
males and female
offspring assessed
Decreased weight gain
Behavioral changes, more
extensive for late pregnancy
exposure
Decreased brain acetylcholine
0, 100, 200 on
GDs 7-13 or on GDs 14-20, 7 h/d
Nelson et al.
(1980)
CFY rats
pregnant females
15 litters/dose;
male and female
offspring assessed
Transient decreases in muscular
strength and exploratory behavior.
Latent increases in motor activity
in females at 100 d postnatally
0. 1.500 or 4.500 me/m3
GDs 1-20 for 8 h/d
Szakmary et al.
(1997)
S-D rats, multigeneration
study
28 litters/dose
CNS depression in first 2 wk of
F1 and F2 generations, which
ceased 2 h after daily exposures
0, 100, 300, 1,000;
6 h/d, 5 d/wk, except during
mating, 6 h/d-7 d/wk
Tinston (1994)
Brain pathology
S-D rats, males
8/dose
Decreased brain weight, DNA,
protein
300. 600:
4 or 12 wk continuous (24 h/d)
Wang et al.
(1993)
S-D rats, males
10/dose
Decreased brain RNA, increased
brain cholinesterase and increased
motor activity
200: 4 d
Savolainen et al.
(1977a: 1977b)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-41 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-3. Summary of animal inhalation neurotoxicology studies
(continued)
Subjects
Effect
NOAFT /T OAFT a fnnml
Reference
S-D rats, males
5-6/dose
Change in fatty acid composition
of cerebral cortex
320: 12 wk
continuous (24 h/d), 30-d washout
period;
320: 4 wk continuous (24 h/d)
Kyrklund et al.
(1988. 1990)
S-D rats, males
5-6/dose
Neurotransmitter changes, brain
regions
200, 400. 800: 4 wk continuous
(24 h/d)
Honma et al.
(1980a: 1980b)
Mongolian gerbils
males and females
6/sex/dose
Decrease in DNA, frontal cortex
Decrease in brain weight
60. 300; 12 wk, continuous
(24 h/d); 16-wk washout period
Rosengren et al.
(1986)
Mongolian gerbils
males and females
4/sex/dose
Decrease in DNA, frontal cortex
Decrease in brain weight
60:12 wk, continuous (24 h/d)
Karlsson et al.
(1987)
Mongolian gerbils
males and females
8/sex/dose
Taurine, glutamine changes in
brain regions
120:12 mo continuous (24 h/d)
Briving et al.
(1986)
Mongolian gerbils
gender unspecified
6/dose
Decrease in brain weight, change
in fatty acids
320: 12 wk continuous (24 h/d)
Kyrklund et al.
(1987)
Mongolian gerbils
males 6/dose
Decreased brain long-chain fatty
acids
120:52 wk continuous (24 h/d)
Kyrklund et al.
(1984)
Guinea pigs
pregnant females
3/litters/dose
males and female;
offspring assessed
Decrease in brain stearic acid in
offspring after in utero exposure13
Maximum exposure 160: GDs 33
to 65 continuous (24 h/d)
Kyrklund and
Hagid (1991)
NMRI mice,
males and females
3-8/sex/dose
Increase in butyl cholinesterase
9°, 37, 75, 150; 4 wk continuous
(24 h/d)
Kjellstrand et al.
(1984)
Males and females
10/sex/dose
Increased motor activity
150: 4 wk intermittent-
(1, 2, 4, 8, or 16 h/d)
Kjellstrand et al.
(1984)
1
2
3
4
5
6
7
Experimental/observational NOAEL is underlined, LOAEL is double-underlined.
b Questionable findings because litter was not used as the unit of measure in analysis.
0 LOAEL for changes in liver weight.
FEP = Flash-evoked potential; GD = Gestational day; S-D = Sprague-Dawley; SEP = Somatosensory-evoked
potential; VEP = Visual Evoked Potential
immobility, and the length of immobility is shortened by antidepressant administration.
Tetrachloroethylene exposure also shortened the period of immobility, with a no-observed-effect
level (NOEL) of 596 ppm.
The effects of exposure to 90-3,600-ppm tetrachloroethylene for 1 hour on motor activity
were examined in male MRI mice (n = 14-27 per exposure group ) (Ki ell strand et al.. 1985). A
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-42 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
strong odor (cologne) was used as the control condition. Total activity was monitored during the
dark period during exposure and for several hours thereafter. All doses produced increased
activity during exposure; activity decreased over several hours after cessation of exposure.
Although apparently no statistical analyses were performed, it is clear from the figures that the
lowest dose produced an average performance that was well outside the boundary of the 95% CIs
of the cologne-treated controls, and performance was dose-dependent.
Male Long-Evans rats (n = 12) previously trained to perform a visual signal detection
task were exposed to 0-, 500-, 1,000-, and 1,500-ppm tetrachloroethylene for 60 minutes (Boves
et al.. 2009). In this learned task, rats are trained to respond to a light stimulus by pressing the
stimulus lever and to press the blank lever when there is no stimulus. Food pellets are provided
to the rat for each correct lever response. The parameters evaluated included measures of (1)
correct responses (pressing stimulus lever with stimulus), (2) correct rejections (pressing blank
lever when stimulus is not presented), (3) false alarms (pressing stimulus lever without stimulus),
and (4) misses (pressing the blank lever when the stimulus is presented). Other endpoints
measured included reaction times from presentation of stimulus to pressing of the lever and if the
rat completed the signal detection task within the allotted period of time (2 minutes).
Tetrachloroethylene (500-1,500 ppm) exposure significantly increased the number of false
alarms, indicative of a decrement in sustained attention. Additionally, the authors reported that
there was a dose-dependent increase in reaction time and decreased trial completions. Rats were
also tested with different signal intensities to evaluate if the changes were partially due to visual
deficits. The number of hits did not significantly change with the signal intensity of the stimulus,
which strongly suggests that the observed effects of tetrachloroethylene in this study are due to
cognitive changes rather than visual effects. The study authors reported a LOAEL of 500 ppm
(60-minute exposure) for effects related to decrements in sustained attention.
4.1.2.1.2. Neurophysiology
Mattsson et al. (1998) studied the effects of acute exposure to tetrachloroethylene for
13 weeks observing flash-evoked potentials (FEPs), somatosensory-evoked potentials (SEPs),
EEGs, and rectal temperature in F344 rats. During the acute (pilot) study, male rats were
exposed to 0- or 800-ppm tetrachloroethylene for 6 hours/day for 4 days and tested before and
after exposure on the 4th day. Changes in FEP, SEP, and EEG components were observed after
acute exposure. In the subchronic study, the above evoked potentials and caudal nerve
conduction velocity were determined in male and female rats exposed to 0, 50, 200, or 800 ppm
for 6 hours/day for 13 weeks. Testing was performed during the week following cessation of
exposure. A significant increase in the amplitude and in latency (-3.0 ms) for the mid
component peak of the FEP was observed at the highest dose (800 ppm). Several measures of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-43 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
the evoked potential were affected at 50 ppm but not at higher doses. Other measures were not
affected, and no dose-response was observed.
Male Long-Evans rats (n = 9-10/group) were exposed to concentrations of
tetrachloroethylene ranging from 0-4,000 ppm in two separate experiments measuring pattern-
elicited steady state visual evoked potentials (Boves et al.. 2009). In the first experiment, rats
were exposed to (mean ± SEM in parentheses) 0, 1,000 (1,006 ± 7.4), 2000 (1993 ± 8.3), 3,000
(3,018 ± 6.9), or 4,000 (4,016 ± 19) ppm for 2 hours (0, 1,000, 2,000 ppm), 1.3 hours
(3,000 ppm), or 1 hour (4,000 ppm). In the second experiment, rats were exposed to 0, 250
(249 ± 1.1), 500 (488 ± 2.9), or 1,000 (1,053 ± 9.6) ppm for 1.5 hours. In both experiments, the
visual evoked potentials were measured while the animal was exposed to tetrachloroethylene.
The steady state visual evoked potential responses measured from the animals are sinusoidal in
nature, and the potentials were transformed so that amplitudes were tabulated at the frequency of
pattern presentation (Fl) and at double the frequency of pattern presentation (F2). At all test
conditions, tetrachloroethylene significantly decreased the F2 amplitude of the steady state visual
evoked potential. The LOAEL for steady state visual evoked potentials for this study is 250-ppm
tetrachloroethylene for 1.5 hours.
4.1.2.1.3. Developmental neurotoxicity
Developmental neurotoxicity is also discussed in Section 4.7.1.2. Nelson et al. (1980)
investigated developmental neurotoxicity in Sprague-Dawley (S-D) rats by exposing pregnant
dams to tetrachloroethylene at concentrations of 100 or 900 ppm during both early pregnancy
(gestation days [GDs] 7 to 13) or late pregnancy (GDs 14 to 20). The investigators made
morphological examinations of the fetuses and performed behavioral testing and neurochemical
analysis of the offspring. There were no alterations in any of the measured parameters in the
100-ppm groups. At 900 ppm, there were no skeletal abnormalities, but the weight gain of the
offspring as compared with controls was depressed about 20% at Weeks 3-5. Developmental
delay was observed in both the early and late pregnancy groups. Offspring of the early
pregnancy-exposed group performed poorly on an ascent test and on a rotorod test (evaluation of
neuromuscular function), whereas those in the late pregnancy group underperformed on the
ascent test only at postnatal day (PND) 14. However, later in development (PNDs 21 and 25),
their performance was higher than that of the controls on the rotorod test. These pups were
markedly more active in the open field test at PNDs 31 and 32.
There were no effects on running in an activity wheel on PNDs 32 or 33 or avoidance
conditioning on PND 34 and operant conditioning on PNDs 40 to 46. Neurochemical analyses
of whole brain (minus cerebellum) tissue in 21-day-old offspring revealed significant reductions
in acetylcholine levels at both exposure periods, whereas dopamine levels were reduced among
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-44 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
those exposed on GDs 7-13. However, none of the statistics for the 100-ppm treatments were
presented. The authors observed that more behavioral changes occurred in offspring exposed
during late pregnancy than in those exposed during early pregnancy.
Szakmary et al. (1997) exposed CFY rats to tetrachloroethylene via inhalation throughout
gestation (i.e., GDs 1-20) for 8 hours/day at concentrations of 0-, 1,500-, or 4,500-mg/m3
tetrachloroethylene. The primary focus of the study was prenatal developmental evaluations (see
Section 4.7.2). However, a cohort of rats (15 litters/group) was allowed to deliver, and the
offspring (standardized to 8 pups/litter) were maintained on study until PND 100 and evaluated
for growth, development, and neurotoxic effects. The report did not specify whether the animals
were exposed to tetrachloroethylene after birth. Preweaning observations included weekly body
weights, developmental landmarks (pinna detachment, incisor eruption, and eye opening), and
functional assessments (forward movement, surface righting reflex, grasping ability, swimming
ontogeny, rotating activity, auditory startle reflex, and examination of stereoscopic vision). After
weaning, exploratory activity in an open field, motor activity in an activity wheel, and
development of muscle strength were assessed. The study authors reported that adverse findings
included a decreased survival index (details were not provided), a minimal decrease of
exploratory activity and muscular strength in treated offspring (presumably at both exposure
levels) that normalized by PND 51, and significantly increased motor activity on PND 100 of
females exposed to 4,500 mg/m3. Litter was evaluated as the statistical unit of measure for all
outcomes. There is no clear indication of group means for postnatal measures reported. The
lack of experimental detail in the postnatal evaluation part of this study reduces the overall
confidence in the findings. There was no evaluation of postnatal histopathology of the nervous
system reported or cognitive testing during the postweaning period or during adulthood.
Tinston (1994) performed a multigeneration study of the effects on rats exposed to
airborne concentrations of tetrachloroethylene. The details of the study are discussed in
Section 4.7.2. The investigators observed several developmental effects. Of interest here were
the signs of CNS depression (decreased activity and reduced response to sound) observed for the
first 2 weeks in both adult generations and when the exposure was resumed on Day 6 postpartum
in the F1 generation (adults and pups). These effects disappeared about 2 hours after cessation
of the daily exposure. Other overt signs of tetrachloroethylene poisoning among the adults
included irregular breathing and piloerection at both 300 and 1,000 ppm. These changes stopped
concurrently with cessation of exposure or shortly thereafter.
4.1.2.1.4. Brain pathology changes
Wang et al. (1993) exposed male S-D rats to 300-ppm tetrachloroethylene continuously
for 4 weeks or 600 ppm for 4 or 12 weeks. Exposure to 600 ppm at either duration resulted in
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-45 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
reduced brain weight gain, decreased regional brain weight, and decreased DNA in the frontal
cortex and the brain stem but not the hippocampus. Four specific proteins [S-100 (an astroglial
protein), glial fibrallary acidic protein, neurone specific enolase, and neurofilament (68-kD
polypeptide)] were decreased at 4 and/or 12 weeks exposure to 600 ppm; 300 ppm had no effect
on any endpoint.
The effects of exposure to 200-ppm tetrachloroethylene, for 6 hours/day, for 4 days, in
male S-D rats were examined for a number of endpoints (Savolainen et al.. 1977a; Savolainen et
al.. 1977b). Rats were euthanized on the 5th day following a further 0-6 hours of exposure.
Tetrachloroethylene levels were highest in fat, followed by liver, cerebrum, cerebellum, lung,
and blood. Tissue levels increased in all tissues over the 6 hours of exposure. Brain RNA
content decreased, and brain nonspecific cholinesterase was increased on the 5th day, although no
statistical comparisons were performed. Locomotion in an open field was increased immediately
following the end of exposure on the 4th day, with no difference 17 hours after exposure,
although no statistical comparisons were made. Brain protein, GSH, and acid proteinase were
unaffected.
A series of experiments were performed on the effects of tetrachloroethylene on brain
lipid patterns. Exposure to 320 ppm for 90 days (Kyrklund et al.. 1990) or 30 days (Kyrklund et
al.. 1988) in male S-D rats resulted in changes in the fatty acid composition of cerebral cortex,
which persisted after a 30-day recovery period (Kyrklund et al.. 1990). Similar results were
observed in the cerebral cortex and the hippocampus of Mongolian gerbils (sex unspecified) as
well as reduced brain weight after exposure to 320 ppm (Kyrklund et al.. 1987). Exposure of
male Mongolian gerbils to 120 ppm for 12 months also resulted in decreases in long-chain,
linolenic acid-derived fatty acids in the cerebral cortex and the hippocampus (Kyrklund et al..
1984).
The effect of tetrachloroethylene on neurotransmitter levels in the brain was explored in
male S-D rats exposed continuously to 200-, 400-, or 800-ppm tetrachloroethylene for a month
(Honma et al.. 1980a; Honmaetal.. 1980b). The 800-ppm dose produced a decrease in ACh in
striatum, and there was a dose-related increase in a peak containing glutamine, threonine, and
serine in whole brain preparations. GAB A, NE, 5-HT, and other amino acids were not affected.
In a study from the same laboratory , Mongolian gerbils of both sexes were exposed to
60- or 300-ppm tetrachloroethylene for 3 months, followed by a 4-month solvent-free period.
Changes in both S-100 and DNA concentrations in various brain regions were observed at the
higher concentration, and decreased DNA in the frontal cortex was observed after exposure to 60
ppm. The higher concentration also produced decreased brain but not body weight. The results
at 60 ppm were replicated in a follow-up study (Karlsson et al.. 1987).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-46 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
In a related study (Briving et al.. 1986). Mongolian gerbils were exposed to
tetrachloroethylene at 120 ppm for 12 months. At the end of exposure, out of a total of 8 amino
acids assayed, taurine was significantly decreased in the two brain regions assessed
(hippocampus and cerebellum), and glutamine was elevated in the hippocampus.
y-Aminobutyric acid (GABA) levels were unaffected, as was uptake of GABA and glutamate.
Kyrklund and Haglid (1991) exposed pregnant guinea pigs to airborne
tetrachloroethylene continuously from GD 33 through GD 65. The exposure was continuous at
160 ppm except for 4 days at the beginning and end of the exposure period, when it was reduced
to 80 ppm. In the control group, there were three dams with litter sizes of four, three, and two
pups, and in the exposed group, there were three dams with litter sizes of two each. The pup
body weights differed between litters. According to the study authors' analysis, the offspring
had a slightly altered brain fatty acid composition, with a statistically significant reduced stearic
acid content in the tetrachloroethylene treatment group, which is consistent with the study
authors' earlier findings in rats. The statistical analysis, however, relied on pups as the
experimental unit rather than the litters, so the ^-values were likely underestimated. The results
also suggested that tetrachloroethylene reduced the litter size, but a much larger study would be
necessary to establish reduced litter size as an effect of tetrachloroethylene in this study was
relatively small and the reduction was not statistically significant.
Caucasian male and female NMRI mice were exposed to 9-, 37-, 75-, or 150-ppm
tetrachloroethylene continuously for 30 days, to 150-ppm tetrachloroethylene for one of several
exposure periods ranging from 5-30 days, or to 150-ppm tetrachloroethylene for 30 days with
various recovery periods (Ki ell strand et al.. 1984). Other groups were exposed intermittently on
several dosing and exposure regimens, which resulted in a TWA of 150 ppm for 30 days. Motor
activity was assessed following exposure. All concentrations of intermittent exposure increased
motor activity. Results of motor activity following continuous exposure were not reported.
4.1.2.2. Oral and Intraperitoneal Studies
Table 4-4 presents a summary of the oral neurotoxicity animal studies, which are
described in greater detail in the sections that follow. For the six oral neurotoxicity studies in
rodents reviewed here, only one (Tredriksson et al.. 1993) describes effects lasting more than 1
week. In that study, the effect (increased motor activity) was the same at 5 and 320 mg/kg. The
lowest LOAEL occurring in the four remaining studies is 100 mg/kg for delayed onset of
circadian activity in rats (Motohashi et al.. 1993). This LOAEL is based on an i.p.-administered
dose describing transient neurological effects and is not comparable to inhalation or ingestion
LOAELs without pharmacokinetic modeling of an appropriate dose metric. No information is
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-47 DRAFT-DO NOT CITE OR QUOTE
-------
1 available for irreversible neurological effects via the oral route because no studies have evaluated
2 the potential for neurotoxicity following chronic oral exposure.
Table 4-4. Summary of oral neurotoxicity animal studies
Subjects
Effect
NOAEL/LOAEL3 (mg/kg)
Reference
Neurobehavioral studies
S-D rats,
male
9/dose
Pain threshold, pain susceptibility, weight
gain decrement
Interpretation is unclear
Daily dose for 8 wk: 5, 50 mg/kg
Chen et al.
(2002b)
S-D rats,
male,
8-10/dose
Operant responses stopped
immediately after 480-mg/kg dose, then 2/3
of animals recovered by 40 min
Brain PCE concentrations were the same at
both doses
Gavage single dose: 0, 160. 480 mg/kg
Warren et al.
(1996)
ICR mice,
male
8-10/dose
NOAEL/LOAEL:
Righting reflex, 2,000/4,000
Balance, 1,000/2,000
Operant responses, 1,000/2,000
Punishment, 500/1,000
Sinele i.n. doses: 0. 500. 1.000. 2.000.
4,000 mg/kg
Umezu et al.
(1997)
F344 rats,
female
«/dose
Increased reactivity, decreased motor
activity, decreased righting ability,
increased landing foot splay, abnormal gait
after one dose
No effect after repeated doses
Single doses: 150 mg/kg is LOAEL
Repeated dosins for 14 d: 1.500 me/ke
is NOAEL
Moser et al.
(1995)
Wistar rats,
male
«/dose
Transient delay in circadian activity, dose-
related
i.p. doses: 0, 100. 500, 1,000 mg/kg-
day for 3 d
Motohashi et
al. (1993)
Developmental neurotoxicity study
NMRI male
mice,
postnatal
exposure
12 pups/dose
(derived from
3 litters)
Increased locomotion and decreased rearing
at Day 60 in both dose groups
No effect immediately after treatment
Gavage treatment: 5, 320 mg/kg daily
for PNDs 10-16
Fredriksson et
al. (1993)
a Experimental/observational NOAEL is underlined, LOAEL is double-underlined.
«/dosc = Number of animals per dose not clearly defined
4.1.2.2.1. Neurobehavior
3 A study in male S-D rats assessed the acute or short-term effects of tetrachloroethylene
4 by gavage on several screening tests (Chen et al.. 2002a). A single dose of 500 mg/kg in adult
5 rats produced changes on three different tests of pain threshold, locomotor activity, and seizure
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-48 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
susceptibility threshold following pentylenetetrazol infusion, whereas 50 mg/kg resulted in
statistically significant effects only on seizure threshold. In the short-term study, young, 45-50 g
rats were dosed 5 days/week, for 8 weeks, with 5 or 50 mg/kg. Behavioral testing began 3 days
after the last dose. Locomotion was affected only at the high dose, whereas both doses produced
effects on the other four endpoints. The 8-week exposure resulted in retarded weight gain in
both treated groups, which was about 10% at the end of the dosing period. The interpretation of
these results is problematic. The tests required scoring by an observer. The study by Chen et al.
(2002a) does not state whether the observer(s) was blind to the treatment group of the animals, a
condition that is essential for such tests to be valid. Differences in body weight between control
and treated rats add potential bias. Further, the paper does not state whether all animals were
tested by the same person for each task or, if not, whether there was any indication of
interobserver correlation. The potential effect of the difference in weight between the control
and the treated groups on these measures is also unknown. Given that the difference between the
control and the treated groups in response latency to painful stimuli is tenths or hundredths of a
second with no dose-response, these issues are of serious concern.
Various behavioral endpoints were assessed in 8-week-old ICR male mice at the
beginning of an experiment by Umezu et al. (1997). Righting reflex was affected after single-
dose i.p. administration of tetrachloroethylene at 4,000 but not at 2,000 mg/kg or less, and ability
to balance on a wooden rod was decreased at 2,000 but not at 1,000 mg/kg or less. Response rate
on a fixed-ratio 20 (FR20) schedule, which requires 20 responses for each reinforcement, was
affected at 2,000 but not at 1,000 mg/kg or less 30 minutes after administration. In a procedure
in which a thirsty mouse was shocked every 20th lick of a water spout, mice dosed with
500 mg/kg—but not with higher or lower doses—received an increased number of shocks. In an
FR20-FR20 punishment schedule, response in the punishment condition was increased at
1,000 but not at 500 mg/kg or less. A puzzling aspect of the study is the mention in the methods
section of—baeding animals," with no further explanation. If the investigators bred their own
mice, there is no indication of how pups were assigned to treatment groups.
Moser et al. (1995) examined the effects of a number of potentially neurotoxic agents,
including tetrachloroethylene, on a neurotoxicity screening battery in adult female F344 rats
following either a single gavage dose (acute exposure) or repeated gavage doses over 14 days
(subacute exposure). For the acute study, subjects were tested 4 and 24 hours following
exposure. After acute exposure, a LOAEL of 150 mg/kg was identified for increased reactivity
to being handled 4 hours after dosing, with increased lacrimation, decreased motor activity,
abnormal gait, decreased response to an auditory stimulus, decreased righting ability, and
increased landing foot splay at higher doses at 4 and/or 24 hours postdosing. A NOAEL was not
identified. In the subacute study, no endpoint was significantly different from those of controls
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-49 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
at doses of 50-1,500 mg/kg. This presumably represents behavioral acclimation following
repeated exposure to tetrachloroethylene.
Locomotor activity was monitored in NMRI mice gavaged with 5- or 320-mg/kg
tetrachloroethylene for 7 days beginning at 10 days of age (Fredriksson et al.. 1993). Twelve
male pups from three or four litters were assigned to each treatment group. Locomotion, rearing,
and total activity (vibration of the cage) were measured for 60 minutes at 17 and 60 days of age.
A statistically significant increase in locomotor activity and total activity of treated mice in both
dose groups was observed, and rearing behavior decreased as compared with controls for all
three evaluations at 60 days of age, but not at 17 days of age when testing followed shortly after
the last dose. Litter mates were used as independent observations in the statistical analysis,
which tends to underestimate ^-values and thereby overstate statistical significance (i.e., Buelke-
Sam et al.. 1985; Hoi son and Pearce. 1992). However, the magnitude of the effects seen, more
than a twofold increase in locomotion and total activity by the end of the Day 60 evaluation
period, and the persistent effects of subacute developmental exposures in this study raise
concern. Locomotor activity was assessed in 6-week-old male Wister rats following i.p. doses of
100-, 500-, or 1,000-mg/kg tetrachloroethylene for 3 consecutive days, with activity being
monitored for at least 1 week following cessation of administration (Motohashi etal.. 1993).
Animals were monitored 24 hours/day, and locomotor activity (measured as change in electrical
capacitance of a circuit beneath the floor of the cage) was analyzed by time-series analysis and
spectral analysis. All doses of tetrachloroethylene changed circadian rhythm in a dose-
dependent manner, with the increased activity at the start of the dark period delayed by
tetrachloroethylene exposure. Recovery took 3-5 days after cessation of exposure.
Operant performance on a fixed-ratio 40 schedule of reinforcement was assessed in adult
male S-D rats gavaged with 160 or 480 mg/kg tetrachloroethylene immediately before testing
(Warren et al.. 1996). The lower dose produced no effect on response rate over the 90-minute
session, whereas the higher dose produced a transient rate decrease in three of six animals (with
recovery after 20 to 40 minutes) and induced a complete cessation of response in two of the six
animals. Tetrachloroethylene concentrations increased rapidly after administration in blood,
brain, fat, liver, and muscle. For the duration of the 90-minute period of testing, blood
tetrachloroethylene levels were approximately linearly related to the administered dose, but brain
tetrachloroethylene levels were similar for both dose groups. This study did not evaluate the
persistent effects of exposure to tetrachloroethylene on cognitive performance.
4.1.2.2.2. Developmental neurotoxicity
Evidence of potential developmental neurotoxicity was reported by Fredriksson et al.
(1993). In this study (see Section 4.1.2.2), tetrachloroethylene was administered to male NMRI
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-50 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
mice by gavage at dose levels of 0, 5, or 320 mg/kg-day on PNDs 10-16. At PNDs 17 and 60,
spontaneous activity (locomotion, rearing, and total activity) was measured over three, 20-minute
periods. No treatment-related alterations in activity were observed at 17 days of age; however, at
60 days of age, all three measures of spontaneous activity were altered. .
4.1.3. Mode of Action for Neurotoxic Effects
The MOA for the neurotoxic effects of tetrachloroethylene is unknown; however, at
present, the best surrogate for the dose metric for neurotoxicity is blood tetrachloroethylene. The
primary neurobehavioral changes that are observed following tetrachloroethylene exposure are
visual changes, cognitive deficits, and increased reaction time. It is not clear if there are multiple
mechanisms resulting in these outcomes. Additionally, there may be multiple mechanisms or
MO As, which may differ for adult and developmental exposure. The acute effects of
tetrachloroethylene appear to share much in common with those of other chlorinated solvents
such as trichloroethylene and dichloromethane as well as toluene, volatile anesthetics, and
alcohols. It is unknown how these different neurological effects are induced, but there are data
available to help elucidate what areas in the brain and specific molecular targets may be involved
in the resulting neurotoxicological outcome.
Neuropathology and mechanistic studies have been conducted in animal models (rats,
mice, gerbils) to determine how tetrachloroethylene may be producing the observed neurological
effects. Changes in fatty acid composition of the brain following a 30- or 90-day exposure has
been reported, and these changes persist for up to 30 days after the cessation of exposure
(Kyrklund et al.. 1984; Kyrklund etal.. 1987. 1988. 1990). Studies that examined the entire
brains of animals reported decreases in astroglial proteins (GFAP and S-100), decreased brain
RNA content, and decreased levels of glutamine, threonine, and serine (Honma etal.. 1980a;
Honmaetal.. 1980b; Kyrklund etal.. 1984; Kyrklund et al.. 1987. 1988. 1990; Rosengren et al..
1986; Savolainen et al.. 1977b; Wang et al.. 1993). Brain regions examined following
tetrachloroethylene exposure included the frontal cortex, the hippocampus, the striatum, and the
cerebellum (B riving et al.. 1986; Honma et al.. 1980a; Honma etal.. 1980b; Karl s son et al..
1987; Kyrklund et al.. 1984; Wang et al.. 1993). Notable changes include decreased DNA
content in the frontal cortex following continuous exposure of 600 ppm for 4 weeks in rats
(Wang et al.. 1993) or a 60-ppm exposure for 3 months in Mongolian gerbils (Karlsson et al..
1987). Decreased taurine levels were noted in both the cerebellum and hippocampus following a
12-month exposure to 120-ppm tetrachloroethylene in Mongolian gerbils, but there were no
changes in GABA levels or uptake (Briving et al.. 1986). Decreased acetylcholine levels in the
striatum were noted in male rats exposed to 800 ppm for 1 month (Honma et al.. 1980a; Honma
etal.. 1980b).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-51 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Voltage and ligand-gated ion channels have been implicated in many neurological
functions and have been studied as potential neurological targets for tetrachloroethylene and
other structurally related chlorinated solvents (e.g., trichloroethylene, 1,1,1-trichloroethane,
dichloromethane). Table 4-5 summarizes the available in vitro mechanistic studies with
chlorinated solvents. Tetrachloroethylene has been demonstrated to inhibit calcium channel
function (Shafer et al.. 2005) and the neuronal nicotinic acetylcholine receptor (Bale et al..
2005). Based on the structural similarity of tetrachloroethylene to other chlorinated solvents as
well as the similar neurobehavioral and mechanistic findings, it is likely that tetrachloroethylene
also interacts with the other listed targets in Table 4-5. This solvent class has also been shown to
interact with ion channels such as the GABAa and glycine receptors (Beckstead et al.. 2000;
Krasowski and Harrison. 2000; Lopreato et al.. 2003). Overall, these solvents appear to
potentiate the function of inhibitory receptors and inhibit the function of excitatory receptors (see
Bowen et al.. 2006; Bushnell et al.. 2007 for a review). Additionally, this class of solvents
blocks sodium channel (Havdon and Urban. 1983; Shrivastav et al.. 1976) and voltage sensitive
calcium channel function (Shafer et al.. 2005) when the membrane is held at or near the resting
membrane potential.
Based on these findings as well as other mechanistic studies conducted with
tetrachloroethylene, some neurotransmitter systems may be more favorably involved in
neurotoxicological outcomes than others. Also, based on the number of reported molecular
targets, it is more likely that there are several plausible mechanisms responsible for the resultant
neurotoxicological outcome, and those potential mechanisms (as well as a discussion of
plausibility) are summarized below by the major observed outcomes (visual changes, cognitive
deficits, increased reaction time).
4.1.3.1. Visual Function
Although tetrachloroethylene produces changes in visual evoked potentials, there are no
associated mechanistic studies to indicate what receptor systems may be involved. However,
there is a characterization study evaluating the contribution of specific ligand-gated ion channels
(GABAa, NMDA-glutamate, nicotinic acetylcholine receptors) to the generation of the steady
state visual evoked potential (Bale et al.. 2005). The findings suggest that ion channels are
involved in visual function and, specifically, the measured evoked potentials. The only
administered drugs resulting in an effect similar to tetrachloroethylene were NMDA
(NMDA-glutamate receptor agonist) and mecamylamine (nAChR antagonist). Therefore, the
Table 4-5. Summary of in vitro ion channel effects with tetrachloroethylene
and other chlorinated solvents
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-52 DRAFT-DO NOT CITE OR QUOTE
-------
Reference
Cellular system
Ion channel/receptor
Concentration
Effects
Tetrachloro ethylene
Shafer et al.
(2005)
PC12 cells, primary
cortical neurons
Voltage Sensitive Calcium
Channels (VSCCs)
0-325 pM
Shift of VSCC activation to a
more hyperpolarizing potential.
Inhibition of VSCCs at a holding
potential of -70 mV
Bale et al.
(2005)
Xenopus oocytes
Human and rat a4p2, a3p2,
and al receptors
0-65 nM
Inhibition of nicotinic
acetylcholine receptor function
Dichloromethane
Hardon and
Urban (1983)
Squid giant axon
Sodium channels
0, 15, 25 mM
Inhibition of inward sodium
channel currents
T richloroethylene
Shafer et al.
(2005)
PC12 cells, primary
cortical neurons
VSCCs
0-2,100 pM
Shift of VSCC activation to a
more hyperpolarizing potential.
Inhibition of VSCCs at a holding
potential of -70 mV
Beckstead et
al. (2000)
Xenopus oocytes
Human recombinant
glycine receptor al,
GABAareceptors, aipi,
aip2y2L
0, 390 (iM
50% potentiation of the GABAa
receptors; 100% potentiation of
the glycine receptor
Lopreato et
al. C2003)
Xenopus oocytes
Human recombinant
serotonin 3A receptor
0, 390 (iM
Potentiation of serotonin
receptor function
Krasowski
and
Harrison,
(2000)
Human embryonic
kidney 293 cells
Human recombinant
glycine receptor al,
GABAa 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
SMvastav et
al. (1976)
Squid giant axon
Sodium channels
5-80%
saturation
Shift of sodium channel
activation to a more
hyperpolarizing potential.
Inhibition of inward sodium
channel current at -70 mV
1,1,1-Trichloroethane
Cruz, et al.
(2000)
Xenopus oocytes
NMDA-glutamate receptor
NR1/2A, NR1/2B
0.1-10 mM
Inhibition of NMDA-glutamate
receptor function
Beckstead et
al. (2000)
Xenopus oocytes
Human recombinant
glycine receptor al,
GABAareceptors, aipi,
aip2y2L
0.39 mM
Potentiation of GABAa and
glycine receptor function
Beckstead et
al. (2000)
Rat hippocampal
slices
GABAa receptor
0.28 mM
Reversible increase in GABAa-
mediated inhibitory postsynaptic
currents (IPSCs)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-53 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
NMDA-glutamate and the nicotinic acetylcholine receptor systems may be more closely
involved in the visual evoked potential changes resulting from solvent exposure.
With respect to the impact on color vision and visual contrast sensitivity following
tetrachloroethylene exposure, the mechanisms behind these effects are unknown. These visual
changes occur at exposures that are lower than the visual evoked potential changes. Cones at the
level of the retina process color vision, and there may be a change in the function and/or
signaling of the retina to the visual center in the CNS. In visual contrast sensitivity, retinal
ganglion cells have been implicated as a sensitive target in processing changes in contrast
(Beaudoin et al.. 2008). The available literature suggests that NMDA-glutamate receptors
(Manookin et al.. 2010) and calcium channels (Hu et al.. 2009) may be involved in visual
contrast sensitivity changes. It is known that tetrachloroethylene exposure affects calcium
channel function in vitro (Shafer et al.. 2005). and a related chlorinated solvent,
1,1,1-trichloroethane, has been demonstrated to modulate NMDA-glutamate receptor function
(Cruz et al.. 2000).
4.1.3.2. Cognition
The hippocampus is involved in cognitive functions, but only changes in taurine levels
were observed in this brain region following tetrachloroethylene exposure in gerbils (Briving et
al.. 1986). It was demonstrated that tetrachloroethylene inhibits both human and rat recombinant
nicotinic acetylcholine receptors in vitro (Bale et al.. 2005). and perhaps this finding may help
explain why cognitive changes are observed with tetrachloroethylene exposure. However, more
studies need to be conducted with tetrachloroethylene exposure and perhaps incorporating a
challenge with nicotinic agonists and antagonists to determine the involvement of nicotinic
acetylcholine receptors in cognitive function.
4.1.3.3. Reaction Time
Reaction time is a general measure of CNS function. With increased reaction time, it can
be surmised that there is a general CNS decrease in movement. Currently, there are no available
mechanistic studies with tetrachloroethylene that have evaluated neurological systems mediating
reaction time activity. There is one study that has reported that decreased CNS function
(anxiolytic profile) observed with tetrachloroethylene may be due to site-specific action on the
GABAa receptors. Chen et al. (2002a) pretreated rats with tetrachloroethylene (50 or 500
mg/kg, oral gavage), and this pretreatment, following both an acute and a subchronic (5 or
50 mg/kg-day, 5 days/week, 8 weeks) schedule significantly increased the seizure threshold
when challenged with pentylenetetrazole (PTZ), a convulsant that blocks GABAa receptors.
This study suggests that the GABAergic system may be involved in the anxiolytic and general
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-54 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
CNS depressive behavior that is observed following tetrachloroethylene exposure and could be
potentially related to observed increased reaction times in the various tasks.
4.1.4. Summary of Neurotoxic Effects in Humans and Animals
Human and animal studies provide complementary evidence regarding the association of
neurobehavioral deficits and tetrachloroethylene exposure. Tetrachloroethylene exposure in
humans has primarily been shown to affect visual function (including color vision) and
visuospatial memory and other aspects of cognition. Brain weight changes have been measured
in animal studies. A more in-depth discussion of the human neurotoxicological studies can be
found in Section 4.1.1.3, and the animal inhalation and oral or i.p. exposure studies are discussed
in Sections 4.1.2.1 and 4.1.2.2, respectively.
Visual contrast sensitivity deficits as well as color discrimination deficits are commonly
present prior to detectable pathology in the retina or optic nerve head, making them one of the
earliest signs of disease and potentially more sensitive measures than evoked potentials from
visual stimuli (Regan. 1989). Several independent lines of evidence can be found in the
occupational and residential exposure studies to support an inference of visual deficits following
chronic tetrachloroethylene exposure. The studies that observed effects on color vision using the
Lanthony D-15 color vision test include cross-sectional and longitudinal designs in dry-cleaning
settings (Cavalleri et al.. 1994; Gobba et al.. 1998) and residential studies (Schreiber et al..
2002). Decrements in color confusion were reported among 22 dry-cleaning workers exposed to
a mean TWA of 7 ppm for an average of 8.8 years (Cavalleri et al.. 1994). A significant dose-
response relationship between CCI value and tetrachloroethylene concentration (r = 0.52,
p < 0.01) was also seen in Cavalleri et al. (1994). As noted previously, the color vision testing in
this study was blinded to exposure level of the study participants, and the study participants were
well matched in terms of age, smoking, and alcohol use. A follow-up of these workers 2 years
later (Gobba et al.. 1998) showed greater loss in color discrimination in those who were
subsequently exposed to a higher concentration (increase in geometric mean from 1.7 to
4.3 ppm), with no change in those exposed to lower concentrations (decrease in geometric mean
from 2.9 to 0.7 ppm). Although Gobba et al. (1998) demonstrates persistent color confusion
effects in this follow-up evaluation, the study exposures are not clearly characterized over the
course of the 2-year duration. Nakatsuka et al. (1992) did not observe an association with color
vision among dry cleaners in China (n = 64, geometric mean TWA: 11 and 15 ppm in females
and males, respectively), but the relative insensitivity of the specific type of color vision test
used in this study (Lanthony. 1978) is a likely explanation for these results. Effects on color
vision were also seen among 14 dry cleaners in the small study in Malaysia by Sharanjeet-Kaur
et al. (2004). but this study provides little weight to the strength of the evidence because of the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-55 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
lack of exposure information (other than job title), and differences between dry cleaners and
controls regarding test conditions and smoking habits. Two other small studies also reported
lower scores on the Lanthony D-15 color vision test in much lower exposure settings, but the
differences were not statistically significant: in a study of residents living above dry cleaners
(mean tetrachloroethylene exposure during active dry cleaning = 0.4 ppm), the mean CCI scores
were 1.33 and 1.20 in 17 exposed and 17 controls, respectively (p = 0.26); in a study of workers
in a day-care center located in a building with a dry-cleaning business (mean tetrachloroethylene
exposure 0.32 ppm), the mean CCI scores were 1.22 and 1.18 in the exposed day-care workers
and controls, respectively (p = 0.39) (Schreiber et al.. 2002). Another residential exposure study
observed decrements in color vision in children but not in adults (NYSDOH. 2005a). Overall,
the evidence reveals a high degree of consistency in this aspect of visually mediated function.
Visual contrast sensitivity changes were reported in two NYSDOH residential studies. In
a small pilot study (4 children and 13 adults), mean scores for visual contrast sensitivity (using a
near vision visual contrast sensitivity test) across spatial frequencies were statistically
significantly lower in exposed residents than in controls, indicating poorer visual function in the
exposed groups (Schreiber et al.. 2002). Controls were age- and sex-matched to the exposed
group, and both groups were English speaking and predominately Caucasian ethnicity; however,
they were drawn from different geographic areas. In addition, two of the four exposed children
had diagnoses of learning disabilities or developmental delays, which could affect performance
on this type of test. In the larger study (NYSDOH. 2005a. b, 2010). the test (Functional Acuity
Contrast Test, FACT) assessed far vision visual contrast sensitivity, and the test had a low rate of
detecting visual contrast changes. For both contrast vision and color vision, a number of
analyses in NYSDOH (2005a. 2010; Storm et al.. In Press) suggest a vulnerability among
children. However, exposure to >0.015 ppm (>100 |ig/m3) tetrachloroethylene was highly
correlated with race and children's age, and the sample sizes in the highest exposure group,
especially in higher income, nonminority groups, makes it difficult to fully examine possible
effects of income, race, and age on vision. Therefore, while both studies report visual contrast
sensitivity changes with exposed children being more sensitive, there are concerns with the
methodological and analytic approaches in these studies.
Acute human exposure studies reported increased latencies of up to 3.0 ms in visual
evoked potentials (Altmann et al.. 1990) and changes in EEGs (magnitude of effect was not
specified (Hake and Stewart. 1977; Stewart et al.. 1970) at higher exposures ranging from 340 to
680 mg/m3.
In rats, acute inhalation exposure to tetrachloroethylene results in significant changes to
the flash-evoked potential at 800 ppm (Mattsson et al.. 1998). and a decrease in F2 amplitudes of
the steady state visual evoked potential at 250 ppm (Boves et al.. 2009). In a subchronic
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-56 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
exposure study (13 weeks, up to 800-ppm tetrachloroethylene), changes in flash-evoked potential
responses were not observed at tetrachloroethylene exposures up to 200 ppm. In the 800-ppm
group, there was a significant increase in the amplitude and a significant increase in latency
(-3.0 ms) of the mid-flash-evoked potential waveform (N3), but histopathological lesions were
not observed in the examination of the visual system brain structures (e.g.. visual cortex; optic
nerve; Mattsson et al.. 1998).
Effects on visuospatial memory in humans were also reported in each of the studies that
examined this measure (Altmann et al.. 1995; Echeverria et al.. 1994; Echeverria et al.. 1995;
Seeber. 1989). These effects (increased response times) were seen in occupational and
residential studies, and the occupational studies were quite large, involving 101, 65, and 173 dry-
cleaning workers in Seeber (1989). Echeverria et al. (1995). and Echeverria et al. (1994).
respectively. Several different types of tests were used including digit reproduction (Seeber.
1989). switching, pattern memory, and pattern recognition (Echeverria et al.. 1994; Echeverria et
al.. 1995). and the Benton test (Altmann et al.. 1995). Exposure ranges for the increased reaction
time observations (LOAELs) ranged from 4.99 to 102 mg/m3 (Altmann et al.. 1995; Echeverria
et al.. 1995; Ferroni et al.. 1992). The changes in the cognitive tasks were observed at exposures
(LOAELs) ranging from 53.9 to 364.22 mg/m3 (Echeverria et al.. 1995; Seeber. 1989;
Spinatonda et al.. 1997). All of these studies except Altmann et al. (1995) indicate that the
neurobehavioral assessment was blinded to knowledge of the exposure level of the subject, and
all of the studies adjusted for potentially confounding factors. It should be noted, however, that
residual confounding from education level differences between exposed and referent subjects
may still be present in Altmann et al. (1995).
Increased reaction time, increased number of false alarms, and decreased trial
completions in a signal detection task (measures of decreased attention) were reported in an
acute (60 minutes) exposure (6,782 mg/m3 or higher) study in rats (Oshiro et al.. 2008).
Additionally, operant tasks that test cognitive performance have demonstrated performance
deficits in rats and mice following acute tetrachloroethylene oral (Warren et al.. 1996) and i.p.
(Umezu et al.. 1997) exposures. These findings are consistent with observed effects on cognition
and memory in humans. However, no studies, to date, have evaluated the persistent effects of
tetrachloroethylene exposure on cognitive performance deficits in animal models.
An occupational exposure study (n = 60) (Ferroni et al.. 1992) and a residential exposure
study (n = 14) (Altmann et al.. 1995). with mean exposure levels of 15 ppm and 0.7 ppm,
respectively, reported significant increases in simple reaction time of 24 ms (11%) (Ferroni et al..
1992) and 40 and 51.1 ms (15 and 20% increases, respectively, for two separate measurements)
(Altmann et al.. 1995) for the exposed subjects. A third study, Lauwerys et al. (1983). reported
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-57 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
better performance on simple reaction time in 21 exposed workers (mean TWA: 21 ppm)
compared with controls measured before a work shift but not after.
The changes in brain weight, DNA/RNA, and neurotransmitter levels that were observed
in the animal studies are highly supportive of the neurobehavioral changes observed with
tetrachloroethylene exposure. Changes in brain DNA, RNA, or protein levels and lipid
composition were altered following inhalation, with changes observed in the cerebellum, the
hippocampus, and the frontal cortex (Rosengren et al.. 1986; Savolainen et al.. 1977a;
Savolainen et al.. 1977b; Wang et al.. 1993). The replication of these changes in biochemical
parameters and effects in brain weight in both rats and gerbils is pathognomonic. Changes in
neurotransmitters systems (Briving et al.. 1986; Honma et al.. 1980a; Honma et al.. 1980b) and
circadian rhythm (Motohashi etal.. 1993) in animal studies are consistent with neuroendocrine
alterations observed in humans (Ferroni et al.. 1992).
In conclusion, the weight of evidence across the available studies of humans and animals
exposed to tetrachloroethylene indicates that chronic exposure to tetrachloroethylene can result
in decrements in color vision, visuospatial memory, and possibly other aspects of cognition and
neuropsychological function, including reaction time.
4 2 KIDNEY AND BLADDER TOXICITY AND CANCER
4.2.1. Human Studies
4.2.1.1. Kidney Toxicity in Humans
High concentrations of inhaled tetrachloroethylene given acutely as an anesthetic are
associated with symptoms of renal dysfunction, including proteinuria and hematuria (ATSDR.
1997b; Hake and Stewart. 1977). Controlled inhalation exposure to tetrachloroethylene at levels
of 0, 20, 100, or 150 ppm for up to 11 weeks did not affect a number of urine parameters or
blood urea nitrogen (BUN) (a measure of kidney function) in 12 healthy individuals [Stewart et
al. (1977). as reported in ATSDR (1997b)]. Whether renal effects would occur from these acute
exposure levels in a larger, more diverse population than the one studied by Stewart et al. (1977)
is not known.
The evidence for kidney effects from chronic inhalation of tetrachloroethylene is limited
to studies of urinary renal proteins as indicator of kidney function. One study has become
available on end stage renal disease (ESRD) incidence in a cohort of dry cleaners (Calvert et al..
In Press). The ATSDR (ATSDR. 1998a; Lybarger et al.. 1999) recommends a core battery of
kidney function tests including serum creatinine, urinalysis with microscopic examination of
urine sediment, albumin, retinol binding protein (RBP), A-acetyl-P-D-glucosaminidase (NAG),
alanine aminopeptidase (AAP), osmolality, and urine creatinine (Lybarger et al.. 1999). These
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-58 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
indicators evaluate a range of toxicity, from effects on general kidney function to effects on
specific segments of the nephron. For example, the overall integrity of the nephron can be
evaluated from the urinalysis, and albumin is an indicator of the integrity of the glomerulus;
three indicators—RBP, NAG, and AAP—assess damage to the proximal tubules, although it
should be noted that NAG is not a sensitive and specific marker of tubular dysfunction (Lybarger
et al.. 1999). The proximal tubules house P-lyase enzymes and are hypothesized to be a target of
tetrachloroethylene toxicity due to the bioactivation of reactive metabolites produced from the
further metabolism of TCVC (see Section 3). For this reason, altered urinary indicators of
proximal tubule function are consistent with knowledge of metabolic processing.
The epidemiologic studies are suggestive of subtle damage to the renal tubules.
Table 4-6 summarizes the human kidney function studies. Five studies (Lauwerys et al.. 1983;
Mutti et al.. 1992; Solet and Robins. 1991; Trevisan et al.. 2000; Verplanke et al.. 1999) have
examined the three core indicators of tubule function—RBP, NAG, or AAP—in urine of dry
cleaners. Three studies measured RBP, with two of the studies reporting a statistically
significant elevated prevalence of abnormal values among study participants (Mutti et al.. 1992)
or a statistically significant elevated geometric mean concentration of RBP (Verplanke et al..
1999) for tetrachloroethylene-exposed workers as compared with controls. The mean
concentration of RBP for exposed subjects (75.4-[j,g/g creatinine) in the Verplanke et al. (1999)
study is within a normal range.1
As a comparison, Nomiyama et al. (1992) suggest a critical level of RBG of 200-|ig/g
creatinine as indicative of cadmium-induced kidney toxicity. Exposure levels were to a median
of 15 ppm (range: limit of detection to 85 ppm) in Mutti et al. (1992) and 1.2 ppm (range:
0.3-6.5 ppm) in Verplanke et al. (1999). Lauwerys et al. (1983). the only other study to assess
RBP, did not observe any differences in the geometric mean concentration of RBP between dry
cleaners with a mean tetrachloroethylene exposure of 21 ppm and their controls; however, this
study contained fewer exposed subjects with a shorter duration of exposure than did that of Mutti
et al. (1992).
The four studies that measured urinary excretion of NAG (Mutti et al.. 1992; Solet and
Robins. 1991; Trevisan et al.. 2000; Verplanke et al.. 1999) and the one study that measured
AAP (Verplanke et al.. 1999) did not observe any differences between exposed subjects and
controls. These findings are not surprising, given the limitations in terms of sensitivity and
specificity of NAG as a marker of tubular dysfunction (Lybarger et al.. 1999). Mean exposures
1 Lapsley et al. (19981 found a median and an upper 98% confidence limit of 67 and 143 |ig/g creatinine,
respectively, in a survey of 70 adults, and this range closely matches the findings of Topping et al. (19861. who
observed a mean and a 98% upper limit of 64 and 185 |ig/g creatine, respectively, in 118 subjects.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-59 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-6. Summary of human kidney toxicity marker studies of
occupational exposures to dry-cleaning facilities using tetrachloroethylene
Subjects, methods
Exposure levels
Results
Reference(s)
Occupational exposures: dry-cleaning settings
Belgium, 26 dry cleaners,
33 unexposed workers
(controls), B, EA, PA, U
[before and after shift]
Mean TWA = 21
ppm, U-TCA = ND,
mean duration =
6.4 yr
No differences in creatinine-adjusted urinary |32|i-
globulin, retinol-binding protein and albumin.
Lauwerys et
al. (1983)
Italy, 57 dry cleaners
(mostly females) (Group
1), 188 painters (mostly
males) (Group 2), 51
glass-fiber reinforced boat
workers (Group 3), 212
workers exposed to C5-C7
alkanes (Group 4), 30
unexposed workers
(mostly females) (Control
Group 1) and 81
unexposed workers
(mostly males) (Control
Group 2). U [before and
after shift]
Dry cleaners
(Group 1): mean
TWA = 10 ppm
(extrapolated from
postshift U-TCA
according to Ikeda et
al. (1972). mean
duration = 13.9 yr
50% increase in creatinine-adjusted geometric
mean concentration of urinary |32 -glucuronidase
and 100% increase in geometric mean urinary
lysozyme in dry cleaners compared to either
control group. No difference in total protein or
albumin.
Franchini et
al. (1983)
Czech Republic, 22 female
dry cleaners, 15 female
controls (clerical workers).
PA, U [end of shift]
3 shops with mean
TWA <12 ppm, 2
shops with mean
TWA 42 ppm and
47 ppm, mean
duration = 11 yr
Fourfold elevation in geometric mean creatinine-
adjusted urinary concentration of lysozyme. No
difference in albumin, p2(i-globulin and total
protein, or prevalence of subjects whose urinary
proteins above 95th percentile.
Vyskocil et
al. (1990)
United State, 192 dry
cleaners (mostly females),
no controls. PA, U
[collection time varied by
subject]
Mean TWA = 7 ppm,
mean duration =
11.6 yr
No correlation of exposure and creatinine-
adjusted urinary protein, albumin, or \-acctyl-
(3-glucuronidase.
Solet and
Robins
(1991)
Italy, 50 dry cleaners and
ironers (mostly females),
50 controls (blood donors).
B, PA, U [before shift]
Mean TWA = 8.8
ppm, mean duration =
10 yr
1.5- to 4-fold increase in creatinine-adjusted
mean concentration of 8 urinary renal proteins
(albumin, transferrin, 3 brush border antigens,
tissue nonspecific alkaline phosphatase,/) < 0.05;
glycosaminoglycans, Tamm-Horsfall
glycoprotein, p = 0.06) and 2 serum proteins
(anti-glomerular basement membrane, laminin
fragments, p < 0.05) in dry cleaners;
discriminated between dry cleaners and matched
controls (p < 0.05). No difference in 12 other
urinary renal proteins (includes total protein and
7V-acety 1-P-glucuronidase).
Mutti et al.
(1992)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-60 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Table 4-6. Summary of human kidney toxicity marker studies of
occupational exposures to dry-cleaning facilities using tetrachloroethylene
(continued)
Subjects, methods
Exposure levels
Results
Reference(s)
Italy, 40 female dry
cleaners, 45 female
controls (ironers). PA, B,
U [before and after shift]
Mean TWA = 14.8
ppm, mean duration =
15 yr
Positive correlation between preshift urinary PCE
and total solutes and total proteins (p < 0.01) and
postshift urinary PCE and glutamine synthetase
(p < 0.001). No difference in creatinine-adjusted
mean urinary concentration of total solutes, total
protein angiotensin converting enzyme, Y-acctyl-
(3-glucuronidase, or glutamine synthetase.
Trevisan et al.
(2000)
The Netherlands, 101 dry
cleaners (mostly males),
19 controls (seamstresses,
sorters or folders in dry-
cleaning shops or laundry
workers) (mostly females).
PA, U [before shift]
Mean TWA = 8 ppm
(dry cleaners), <2.2
ppm (controls), mean
duration = 3.9 yr
Retinol binding protein (creatinine-adjusted mean
concentration) elevated twofold among dry
cleaners (£>=0.01). No difference in creatinine-
adjusted mean of (3-galactosidase, Y-acctyl-^-
glucuronidase, or alanine aminopeptidase. No
difference in geometric mean albumin or total
protein.
Verplanke et
al. (1999)
A = air sample, not specified area or personal sample; AA = area air samples, B = biological monitoring of blood,
BTX = benzene, toluene, xylene, ND = not detectable, PA = personal air samples, EA = exhaled air samples, PCE =
tetrachloroethylene, U = biological monitoring of urine for trichloroacetic acid.
were 14 ppm in Solet and Robins (1991) and 9 ppm in Trevisan et al. (2000); both studies
assessed exposure from personal monitoring of exhaled breath.
The above findings are further supported by the observation of elevated urinary excretion
of other proteins that are also indicators of damage to the proximal tubules: p2[j,-globulin,
intestinal alkaline phosphatase (IAP), tissue nonspecific alkaline phosphatase (TNAP),
lysozyme, P2-glucuronidase, and glutamine synthetase. Both IAP and TNAP are indicators of
proximal tubule brush border integrity (Price et al.. 1996). whereas lysozyme and P2[j,-globulin
indicate a failure of the tubule to reabsorb protein (Bernard and Lauwerys. 1995; Kok et al..
1998; Lybarger et al.. 1999). Glutamine synthetase is a mitochondrial enzyme located in the
proximal tubules and has been recently suggested as a marker of tubular damage in rats exposed
to 1,3-hexachlorobutadiene (Trevisan et al.. 1999).
Mutti et al. (1992) observed an elevated prevalence of abnormal values for p2[j,-globulin
and brush border antigens, a higher geometric mean concentration of brush border antigens in
urine, and a higher concentration of TNAP in urine among 50 exposed dry cleaners as compared
with 50 blood donors matched by sex and age with the exposed subjects. Furthermore, markers
of renal damage were highly predictive of exposure status in discriminant analysis.
P2[j,-Globulin, however, was not elevated among exposed subjects as compared with controls in
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-61 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
the other two studies that examined this protein (Lauwerys et al.. 1983; Vvskocil et al.. 1990).
although the mean concentration of p2[j,-globulin appeared higher in subjects studied by
Vyskocil et al. (1990) than the mean concentration in controls. Both these studies contained
fewer numbers of exposed subjects than did the study by Mutti et al. (1992). and reduced power
as a consequence of fewer subjects may be a reason for the null observations. Further,
tetrachloroethylene exposure appears to affect reabsorption in the renal tubules. Two studies that
assessed lysozyme or -glucuronidase observed a statistically significant elevated mean
concentration of these proteins among dry cleaners as compared with controls (Franchini et al..
1983; Vvskocil etal.. 1990).
It is not clear whether tetrachloroethylene exposure affects other parts of the kidney. The
study by Mutti et al. (1992) is suggestive of damage to the glomerulus; however, the lack of an
elevated excretion of albumin, an indicator of glomerular function (Lybarger et al.. 1999). in the
study by Verplanke et al. (1999) argues for further assessment of possible glomerular effects. As
some albumin is normally filtered, small increases in the amount of albumin in the urine may
result from tubular damage due to failure to reabsorb the small amount filtered (NRC. 2010).
Calvert et al. examined the incidence of end stage renal disease (ESRD) in a cohort of
1,704 dry cleaners assembled by the National Institute of Occupational Safety and Health
(NIOSH), 618 who had worked only in a shop where tetrachloroethylene was the primary
cleaning solvent (tetrachloroethylene-only subcohort) and 1,086 who worked in a shop that used
tetrachloroethylene but who also had a history of employment in shops where the primary
solvent could not be identified (tetrachloroethylene-plus subcohort) (Ruder etal.. 1994. 2001).
All subjects alive as of 1977 were linked to the Renal Management Information System (RMIS),
a database of individuals receiving Medicare benefits for ESRD, and followed to 2004. Thirty
cases of ESRD were identified over the 27 year period (standardized incidence ratio [SIR]: 1.34,
95% CI: 0.90, 1.91), with 12 ESRD cases in the tetrachloroethylene-only subcohort (SIR: 1.30,
95% CI: 0.67, 2.26). Of these cases, eight were due to hypertensive ESRD (SIR: 2.66,
95% CI: 1.15, 5.23), of whom six cases were female subjects (SIR: 2.86, 95% CI: 1.05, 6.23).
The observed risk estimate for hypertensive ESRD among tetrachloroethylene-only subjects
appears larger than that for the tetrachloroethylene-plus subcohort (SIR: 1.53, 95% CI: 0.62,
3.16). An exposure-response pattern was further suggested as hypertensive ESRD risk was
highest among those in the tetrachloroethylene-only subcohort employed for >5 years (SIR: 3.39,
95% CI: 1.10, 7.92). These findings support an association between tetrachloroethylene
exposure and ESRD, particularly hypertensive ESRD. ESRD-observed risk is likely
underestimated using RMIS records. An examination of cause of death among cohort subjects
who had died by 2004 found five additional workers with chronic renal failure listed as an
underlying cause of death. Medical records for three of these five deaths indicated two subjects
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-62 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
with ESRD. Calvert et al. , moreover, found substantial underreporting of chronic renal disease
on death certificates, suggesting incidence as superior to mortality for assessing
tetrachloroethylene exposure and kidney disease. Of the 23 deaths among the 30 ESRD subjects,
cause of death on death certificates for 11 of these subjects was due to chronic renal disease and
three due to —anal disease not otherwise specified."
Taken together, the epidemiologic studies support an association between
tetrachloroethylene and chronic kidney disease, as measured by urinary excretion of renal
proteins and ESRD incidence. The elevated urinary RBP levels seen in two studies (Mutti et al..
1992; Verplanke et al.. 1999) and lysozyme or p2-glucuronidase in Franchini et al. (1983)
provide some evidence for effects to the proximal tubules from tetrachloroethylene exposure.
Effects are seen in populations of both males and females, and potential differences in
susceptibility due to sex-related differences in rates of metabolism (see Section 3) cannot be
determined from the available evidence. Median exposure levels in the studies that observed
alterations in renal enzymes were 9 ppm (Trevisan et al.. 2000). 10 ppm (Franchini et al.. 1983).
and 15 ppm (Mutti et al.. 1992). representing LOAELs for these studies. Only the study by
Trevisan et al. (2000) observed an exposure-response relationship, a correlation between urinary
tetrachloroethylene and the concentration of glutamine synthetase (p < 0.001). None of the other
studies reported exposure-response relationships, which is a limitation on the inference of an
association between tetrachloroethylene and renal damage. However, as pointed out by Mutti et
al. (1992). this is a common finding among solvent-exposed populations, and inadequate
definition of the dose metric most likely contributes to the null finding. Table 4-6 summarizes
the human kidney toxicity studies. Calvert et al. (In Press) supports association between
inhalation tetrachloroethylene exposure and ESRD, particularly hypertensive ESRD. They
observed a twofold elevated incidence (SIR: 2.66, 95% CI: 1.15, 5.23) among subjects who
worked only in a shop where tetrachloroethylene was the primary cleaning solvent compared to
that expected based on U.S. population rates. An exposure-response pattern was further
suggested as hypertensive ESRD risk was highest among those employed for >5 years
(SIR: 3.39, 95% CI: 1.10, 7.92).
4.2.1.2. Kidney Cancer in Humans
Twenty-seven epidemiologic studies reporting data on kidney cancer and
tetrachloroethylene exposure were identified. This set of studies includes 12 cohort or nested
case-control studies within a cohort (Anderson et al.. 1999: Anttila et al.. 1995: Blair et al.. 2003:
Boice et al.. 1999: Calvert et al.. In Press; Chang et al.. 2003; Ji et al.. 2005b; Lynge and
Thygesen. 1990; Pukkala et al.. 2009; Sung et al.. 2007; Travier et al.. 2002; Wilson et al.. 2008);
12 case-control studies of occupational exposures (Asal et al.. 1988; Auperin et al.. 1994;
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-63 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Delahunt et al.. 1995; Dosemeci etal.. 1999; Lynge et al.. 2006; Mandel et al.. 1995; McCredie
and Stewart. 1993; Mellemgaard et al.. 1994; Parent et al.. 2000b; Pesch et al.. 2000b; Schlehofer
et al.. 1995; Selden and Ahlborg. 2011). and 3 studies of residential exposure through
contaminated drinking water (Aschengrau et al.. 1993; Ma et al.. 2009; Vieira et al.. 2005b).
Some sets of these studies represent overlapping study populations. For example, three papers
examined cancer risk among occupational groups defined by census data in Sweden (Ji and
Hemminki. 2005a; Travier et al.. 2002; Wilson et al.. 2008). one paper used a similar design in
Denmark (Lynge and Thygesen. 1990). two papers were based on census data from Sweden,
Denmark, Finland, and Norway (Andersen et al.. 1999; Lynge et al.. 2006). and a third paper
added data from Iceland (Pukkala et al.. 2009). Cases and controls in another four studies
(Dosemeci et al.. 1999; McCredie and Stewart. 1993; Mellemgaard et al.. 1994; Schlehofer et al..
1995) were included in the National Cancer Institute's (NCI's) multicenter international renal
cell study (Mandel etal.. 1995).
Generally, cohort studies presented risk estimates for —klney and other and unspecified
urinary organs," and case-control studies presented risk estimates for renal cell carcinoma, a
histological type included in the broader kidney and other and unspecified urinary organs
category. The exceptions were two studies that presented risk estimates for cancer of the renal
pelvis (McCredie and Stewart. 1993; Wilson et al.. 2008) and two studies of the same cohort that
presented risk estimates for kidney and urinary (bladder) organs (Chang et al.. 2003; Sung et al..
2007). These 27 studies represent the core studies evaluated by EPA, as described in more detail
below. One other cohort study included information on tetrachloroethylene but did not report
risk estimates for kidney cancer (Radican et al.. 2008). and one case-control study identified only
one exposed case (as a dry-cleaning operator) and did not provide an estimate of the association
(Partanen et al.. 1991). and so were not evaluated further. Appendix B reviews the design,
exposure-assessment approach, and statistical methodology for each study. Most studies were of
the inhalation route of exposure, of occupational exposure, and unable to quantify
tetrachloroethylene exposure.
4.2.1.2.1. Consideration of exposure-assessment methodology
Many studies examine occupational title as dry cleaner, launderer, and presser as
surrogate for tetrachloroethylene, given its widespread use from 1960 onward in the United
States and Europe (Andersen et al.. 1999; Asal et al.. 1988; Auperin et al.. 1994; Blair et al..
2003; Calvert et al.. In Press; Delahunt et al.. 1995; Dosemeci et al.. 1999; Ji et al.. 2005b; Lynge
et al.. 2006; Lynge and Thygesen. 1990; Mandel et al.. 1995; McCredie and Stewart. 1993;
Mellemgaard et al.. 1994; Parent et al.. 2000b; Pukkala et al.. 2009; Selden and Ahlborg. 2011;
Travier et al.. 2002; Wilson et al.. 2008). Seven studies conducted in Nordic countries are based
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-64 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
on either the entire Swedish population or combined populations of several Nordic countries;
strengths of these studies are their use of job title as recorded in census databases and
ascertainment of cancer incidence using national cancer registries (Andersen et al.. 1999; Ji et al..
2005b; Lynge et al.. 2006; Pukkala et al.. 2009; Selden and Ahlborg. 2011; Travier et al.. 2002;
Wilson et al.. 2008). Some variation can be expected within an occupational group among
countries; however, as Lynge et al. (2006) reported, average tetrachloroethylene usage in
1960-1970 in Sweden was higher than in Finland or Norway. Studies examining mortality
among U.S. dry-cleaner and laundry workers (Blair et al.. 2003; Calvert et al.. In Press) are of
smaller cohorts than the Nordic studies, with fewer observed kidney cancer events.
The exposure surrogate in studies of dry-cleaners and laundry workers is a broad
category containing jobs of differing potential for tetrachloroethylene exposure. Thus, these
studies have a greater potential for exposure misclassification bias compared to studies with
exposure potential to tetrachloroethylene assigned by exposure matrix approaches applied to
individual subjects. Three studies used additional information pertaining to work environment to
refine the exposure classification (Calvert et al.. In Press; Lynge et al.. 2006; Selden and
Ahlborg. 2011). Selden and Ahlborg (2011) obtained information about the dry-cleaning
establishment (e.g., washing techniques, chemicals used, number of employees, and work history
of individual employees) in a questionnaire sent to businesses in Sweden in the 1980s. Lynge et
al. (2006). using job titles reported in the 1970 Census, identified subjects based on the
occupational code of —Laundry and di^cleaning worker" or industry code of —Laundry and d>r
cleaning." Additional information used to refine this classification was sought for incident
kidney cancer cases (and cases of cancer of the esophagus, gastric cardia, liver, pancreas, cervix,
bladder, and non-Hodgkin lymphoma) within this defined cohort. Five controls, matched to the
cases by country-, sex- age-, and calendar period, were also included in this study. The
additional information sought by Lynge et al. (2006) included handwritten task information from
the census form from Denmark and Norway, pension databases in Denmark and Finland, and
next-of-kin interviews in Norway and Sweden. Exposure classification categories were dry
cleaner (defined as dry cleaners and supporting staff if employed at a business with <10
workers), other job titles in dry cleaning (launderers and pressers), unexposed (job title reported
on 1970 Census was other than in dry cleaning), or unclassifiable (information was lacking to
identify job title of subject). The unclassifiable category represented 43 of 210 identified kidney
cancer cases (20%) and 241 of the 1,060 controls (22%). Another dry-cleaning study of
unionized dry cleaners in the United States included an analysis of subjects who worked for 1 or
more years before 1960 in a shop known to use tetrachloroethylene as the primary solvent
(Calvert et al.. In Press; Ruder et al.. 1994. 2001). The cohort was stratified into two groups
based on the level of certainty that the worker was employed only in facilities using
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-65 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
tetrachloroethylene as the primary solvent exposure; tetrachloroethylene-only and
tetrachloroethylene-plus. Two of the five observed kidney cancer deaths were among the
tetrachloroethylene-only subset (n = 618) of study subjects.
Only Blair et al. (1993; 2003) used an exposure metric for semiquantitative cumulative
exposure within a dry-cleaning setting. Four other studies presented risk estimates by
employment duration (Ji et al.. 2005b; Lynge et al.. 2006; Mandel et al.. 1995; Travier et al..
2002) Because employment duration does not account for variation in exposure levels, it is a
weaker exposure measure (i.e., more subject to misclassification) compared with one defined as
a semiquantitative measure.
One case-control study used a job exposure matrix (JEM) or one with information on
specific tasks, a job-task exposure matrix (JTEM), with semiquantitative exposure assessment
across a variety of jobs (Pesch et al.. 2000b). and two study centers (Dosemeci etal.. 1999;
Schlehofer et al.. 1995) of the large NCI international renal cell carcinoma study used a JEM and
occupation to assign overall tetrachloroethylene exposure potential to individual subjects. In
Pesch et al. (2000b). the use of the German JEM identified approximately three times as many
cases with any potential tetrachloroethylene exposure (38%) compared to the JTEM (12%) and,
in both approaches, few cases were identified with substantial exposure (6% by JEM and 2% by
JTEM). Pesch et al. (2000b) noted, —expasre indices derived from an expert rating of job tasks
can have a higher agent-specificity than indices derived from job titles." For this reason, the
JTEM approach, with consideration of job tasks, is considered a more robust exposure metric for
examining tetrachloroethylene exposure and kidney carcinoma due to likely reduced potential for
exposure misclassification compared to exposure assignment using only job history and title.
Four other cohorts with potential tetrachloroethylene exposure in manufacturing settings
have been examined. These studies include aerospace workers in the United States (Boice et al..
1999). workers primarily in the metal industry, workers in Finland (Anttila et al.. 1995). and
electronic factory workers in Taiwan (Chang et al.. 2005; Sung et al.. 2007). Boice et al. (1999)
used an exposure assessment based on a job-exposure matrix, and Anttila et al. (1995) used
biological monitoring of tetrachloroethylene in blood to assign potential tetrachloroethylene
exposure to individual subjects. In contrast, the exposures in the Taiwan studies included
multiple solvents, and tetrachloroethylene exposure was not linked to individual workers. These
cohorts also included white-collar workers, who had an expected lower potential for exposure
(Chang et al.. 2003; Sung et al.. 2007).
Three geographic studies focused on residential proximity to drinking water sources
contaminated with tetrachloroethylene and other solvents (Aschengrau et al.. 1993; Ma et al..
2009; Vieira et al.. 2005b). Two other studies in Cape Cod, MA, used either an exposure model
incorporating leaching and characteristics of the community water distribution system to assign a
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-66 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
household-relative dose of tetrachloroethylene (Aschengrau et al.. 1993) or residential proximity
to Superfund sites without identifying specific exposures and a generalized additive model that
incorporates smoothing approaches and adjusts for covariates (Vieira et al.. 2005b). Ma et al.
(2009) is an ecological-designed study examining the rate of hospital discharges with a diagnosis
of kidney cancer and the average number of dry cleaners per square kilometer within New York
City zip codes as an exposure surrogate.
In summary, with respect to exposure-assessment methodologies, nine studies with
kidney cancer data assigned tetrachloroethylene exposure to individuals within the study using a
job exposure matrix (Boice et al.. 1999; Dosemeci et al.. 1999; Pesch et al.. 2000b; Schlehofer et
al.. 1995). or semiquantitative metric (Blair et al.. 2003). biological samples (Anttila et al..
1995). an exposure model (Aschengrau et al.. 1993). information about working conditions
obtained through a questionnaire (Selden and Ahlborg. 2011). or classifying the cohort by
certainty of tetrachloroethylene exposure (Calvert et al.. In Press). One other study based on
occupational census data sought additional information for use in refining potential exposure
within dry-cleaning settings (Lynge et al.. 2006). The relative specificity of these exposure-
assessment approaches strengthens their ability to identify cancer hazards compared to studies
with broader and less sensitive exposure-assessment approaches. The least sensitive exposure
assessments are those using very broad definitions such as working in a plant or factory (Chang
et al.. 2003; Sung et al.. 2007) or density of dry-cleaning establishments by zip code (Ma et al..
2009).
4.2.1.2.2. Summary of results
Seven of the 27 studies evaluated by EPA reported estimated relative risks based on a
large number of observed events: 50 or more deaths/incident cases in cohort studies (Andersen et
al.. 1999; Ji and Hemminki. 2005a; Pukkala et al.. 2009; Travier et al.. 2002). or 50 or more
exposed cases in case-control studies (Dosemeci et al.. 1999; Mandel et al.. 1995; Pesch et al..
2000b). Two of these studies adopted a relatively high quality exposure-assessment approach to
assign tetrachloroethylene exposure potential to individual subjects (Dosemeci etal.. 1999;
Pesch et al.. 2000b). Pukkala et al. (2009) updates the analysis of Andersen et al. (1999). adding
data from a 5th country, Iceland, and extending follow-up to 2005, and is preferred over
Andersen et al. (1999) for these reasons.
The three1 cohort studies with findings based on 50 or more events observed standardized
incidence ratios or odds ratio estimates of 1.15 (95% CI: 0.98, 1.35), 0.94 (95% CI: 0.83, 1.07),
and 1.11 (95% CI: 0.93, 1.33) in Ji et al. (Ji et al.. 2005b). Pukkala et al. (2009) and Travier et al.
1 Andersen et al. (19991 is not included in this summary of the data from the individual studies because it was
updated and expanded in the analysis by Pukkala et al. (20091.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-67 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
(2002). respectively, for the association between kidney cancer risk and ever having a job title of
dry-cleaner or laundry worker (see Table 4-7). The largest case-control study (n = 245 cases
from Australia, Denmark, Germany, Sweden, and the United States) reported an odds ratio for
the association between renal cell carcinoma and ever exposed to dry-cleaning solvents of 1.4
(95% CI: 1.1, 1.7) (Mandel etal.. 1995). Dosemeci et al. (1999). whose subjects were included
in the larger study of Mandel et al. (1995). reported an odds ratio estimate of 1.07
(95% CI: 0.7, 1.6) for the association between overall tetrachloroethylene exposure and renal cell
carcinoma, based on 50 cases exposed to tetrachl or ethylene. The other large case-control study
by Pesch et al. (2000b) also included a high-quality exposure-assessment approach (JTEM) for
tetrachloroethylene. This study observed odds ratio estimates of 1.2 (95% CI: 0.9, 1.7), 1.1 (95%
CI: 0.7, 1.5), and 1.3 (95% CI: 0.7, 2.3) and, 2.2 (95% CI: 0.9, 5.2), 1.5 (95% CI: 0.6, 3.8), and
2.0 (0.5, 7.8) for medium, high, and substantial exposure in males and females, respectively.
This study observed lower odds ratio estimates for the association between kidney cancer and
tetrachloroethylene exposure assigned using a job-exposure-matrix, a less robust exposure-
assessment approach compared to a JTEM.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-68 DRAFT-DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-7. Summary of human studies on tetrachloroethylene exposure and kidney cancer
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Cohort Studies
Biologically monitored workers
Anttila et al. (1995)
All subjects
1.82 (0.22, 6.56)
2
849 Finnish men and women, blood PCE [0.4 |imol/L in females and
0.7 (imol/L in males (median)], follow-up 1974-1992, external
referents (SIR)
Aerospace workers (Lockheed)
Boice et al. (1999)
Routine exposure to PCE
0.69 (0.08, 2.47)
2
77,965 (ri = 2,631 with routine PCE exposure and n = 3.199 with
intermittent-routine PCE exposure), began work during or after 1960,
worked at least 1 yr, follow-up 1960-1996, job exposure matrix
without quantitative estimate of PCE intensity, 1987-1988 8-h TWA
PCE concentration (atmospheric monitoring) 3 ppm (mean) and 9.5
ppm (median), external reference for routine exposure (SMR) and
internal references (workers with no chemical exposures) for routine-
intermittent PCE exposure (RR)
Routine-intermittent exposure duration to PCE
0
1.0a
22
<1 yr
0.49 (0.07, 3.68)
1
1-4 yr
0.56 (0.13,2.41)
2
>5 yr
0.46 (0.10,2.08)
2
Electronic factory workers (Taiwan)
Chang et al. (2003); Sung et al. (2007)
All Subjects
86,868 (n = 70,735 female), follow-up 1979-1997, multiple solvents
exposure, does not identify PCE exposure to individual subjects, cancer
mortality, external referents (SMR) (Chans et al.. 2003);
63,982 females, follow-up 1979-2001, factory employment proxy for
exposure, multiple solvents exposures and PCE not identified to
individual subjects, cancer incidence, external referents, analyses
lassed 5 vr (SIR) (Suns et al.. 2007)
Males
0
1.31 exp
Females
1.18 (0.24, 3.44)b
3
Females
1.10(0.62, 1.82)°
10
Aircraft maintenance workers from Hill Air Force Base
Radican et al. (2008)
Any PCE exposure
Not reported
10,461 men and 3,605 women (total n = 14,066, n = 10,256 ever
exposed to mixed solvents, 851 ever-exposed to PCE), employed at
least 1 yrfrom 1952-1956, follow-up 1973-2000, job exposure matrix
(intensity), internal referent (workers with no chemical exposures) (RR)
a -r
>1
to 0\
2 ^
3
On
VO
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-7. Summary of human studies on tetrachloroethylene exposure and kidney cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Dry-cleaner and laundry workers
Andersen et al. (1999)
All laundry worker and dry cleaners
0.92 (0.73, 1.15)
81
29,333 men and women identified in 1960 Census (Sweden) or 1970
Census (Denmark, Finland, Norway), follow-up 1971-1987 or 1991,
PCE not identified to individual subjects, external referents (SIR)
Males
1.03 (0.66, 1.53)
24
Females
0.88 (0.67, 1.15)
57
Blair et al. (2003)
All subjects
1.0 (0.4, 2.0)
8
5,369 U.S. men and women laundry and dry-cleaning union members
(1945-1978), follow-up 1979-1993, semiquantitative cumulative
exposure surrogate to dry clean solvents, cancer mortality, external
referents (SMR)
Semiquantitative exposure score
Little to no exposure
0.3 (<0.1, 1.6)
1
Medium to high exposure
1.5 (0.6, 3.1)
7
Ji et al. (2005bN)
Laundry workers and dry cleaners in 1960 Census
1.15 (0.98, 1.35)
153
9,255 Swedish men and 14,974 Swedish women employed in 1960
(men) or 1970 (women) as laundry workers or dry cleaners, follow-up
1961/1970-2000, PCE not identified to individual subjects, external
referent (SIR) and adjusted for age, period, and socioeconomic status
Males
0.90 (0.69, 1.14)
61
Females
1.41 (1.13, 1.71)
92
Laundry workers and dry cleaners in both 1960 and 1970 Censuses
Males
Not reported
Females
1.67 (1.07, 2.37)
26
Laundry workers and dry cleaners in 1960, 1970, and 1980 Censuses
Males
Not reported
Females
1.00 (0.90, 1.10)
3
a -r
>1
to 0\
2 ^
3
o
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-7. Summary of human studies on tetrachloroethylene exposure and kidney cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Dry-cleaner and laundry workers (continued)
Lvnac and Thvascn (1990)
All laundry worker and dry cleaners
0.88 (0.44, 1.58)
11
10,600 Danish men and women, 20-64 yr old, employed in 1970 as
laundry workers, dry cleaners, and textile dye workers, follow-up
1970-1980, external referents (SIR)
Males
1.50 (0.55, 3.27)
6
Females
0.58 (0.19, 1.36)
5
Pukkala et al. (2009)
Launderer and dry cleaner
0.94 (0.83, 1.07)
263
Men and women participating in national census on or before 1990, 5
Nordic countries (Denmark, Finland, Iceland, Norway, Sweden), 30-64
yr, follow-up 2005, occupational title of launderer and dry cleaner in
any census, external referents (SIR)
Male
0.89 (0.68, 1.14)
62
Female
0.96 (0.84, 1.10)
201
Calvert et al. (In Press)
All subjects
1.14(0.37,2.67)
5
1,704 U.S. men and women dry-cleaning union members in CA, IL,
MI, NY follow-up 1940-2004 (618 subjects worked for one or more yr
prior to 1960 only at shops where PCE was the primary cleaning
solvent, identified as PCE-only exposure), cancer mortality (SMR)
Exposure duration/time since 1st employment
<5 yr/<20 yr
Not reported
<5 yr/>20 yr
Not reported
>5 yr/<20 yr
Not reported
>5 yr/>20 yr
Not reported
PCE-only subjects
1.35 (0.16,4.89)
2
Selden and Ahlborg (2011)
Dry-cleaners and laundry workers
1.04 (0.69, 1.49)
100
9,440 Swedish men (n = 2,810) and women (n = 9,440) in 461 washing
and dry-cleaning establishments, identified by employer in mid-1980s,
employed 1973-1983, follow-up 1985-2000, exposure assigned using
company self-reported information on PCE usage—PCE (dry cleaners
and laundries with a proportion of PCE dry cleaning), laundry (no PCE
use), and other (mixed exposures to PCE, CFCs, TCE, etc.), external
referents (SIR)
PCE
Not reported
Laundry
Not reported
a -r
>1
to 0\
2 ^
3
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-7. Summary of human studies on tetrachloroethylene exposure and kidney cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Travier et al. (2002)
All subjects, 1960 or 1970 Census in laundry and
dry cleaner or related occupation and industry
1.11 (0.93, 1.33)
121
Swedish men and women identified as laundry worker, dry cleaner, or
presser (occupational title), in the laundry, ironing, or dyeing industry
or related industry in 1960 or 1970 (543,036 person-years); or, as
laundry worker, dry cleaner, or presser (occupational and job title)
(46,933 person-years) in both censuses, follow-up 1971-1989, external
referents (SIR)
All subjects in 1960 and 1970 in laundry and dry
cleaner occupation and industry
1.20 (0.71, 2.02)
14
Wilson et al. ^2008^
All subjects, laundry and dry cleaning occupation
16,512 Swedish men (n = 3,375)and women (n = 13,137) identified in
1960 or 1970 as laundry worker or dry cleaner (occupation) or in
laundry, ironing and dyeing industry, follow-up 1971-1989, external
referents (SIR), cancer of the renal pelvis
Males
Not reported
<2 obs.
Females
1.23 (0.39, 2.86)
5
Case-Control Studies
Asaletal. ri988^
Dry-cleaning industry
315 histologically or radiologically confirmed renal cell carcinoma
cases identified from 29 Oklahoma hospitals, 1981-1984, 336
population controls frequency matched on age and sex and 313 hospital
controls matched by age, sex, race, hospital and time of interview to
cases, in-person interview using questionnaire, longest job held was
exposure surrogate, OR adjusted for age, smoking weight
Males
0.7 (0.2, 2.3)
3
Females
2.8 (0.8, 9.8)
8
Upper Cape Cod, MA (United States)
Aschengrau et al. (1993). Vieira et al. (2005b)
Any PCE
1.08 (0.42, 2.79)
6
35 kidney cancer cases, 1983-1986, Massachusetts Cancer Registry,
777 population controls, residential history, ordinal estimate of PCE-
contaminated water (RDD) from exposure model (Aschensrau et al..
1993) or eeoeratihical information svstem and Droximitv to
groundwater Dlumc (Vieira et al.. 2005b). OR adiusted for sex. ase at
diagnosis, vital status at interview, education, cigarette smoking, and
urinary tract infection or stone (both studies)
RDD >90lh percentile
0
a -r
>1
to 0\
2 ^
3
to
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-7. Summary of human studies on tetrachloroethylene exposure and kidney cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
10 hospitals (France)
Auperin et al. (1994)
Dry cleaning occupation
Not reported
151 histologically confirmed renal cell carcinoma hospital cases,
1987-1991, 161 hospital cancer controls and 186 hospital controls with
nonmalignant disease matched on age, sex, and interviewed to cases,
in-person interview, lifetime occupational title as exposure surrogate,
OR adjusted for age, smoking, weight
Population of New Zealand
Delahunt et al. (1995)
Launderer and dry cleaner occupation
1.92 (0.37, 13.89)
Not reported
710 male histologically confirmed renal cell carcinoma cases, >20 yr of
age, 1978-1986, 12,758 male controls randomly selected from same
cancer registry as cases but with tumor outside urinary tract, interview
method not reported, occupational title (ever employed or usual job title
not reported) as exposure surrogate, Mantel-Haenszel OR stratified by
smoking history and 10-yr age group
International Renal Cell Cancer Study (Australia, Denmark, Germany, Sweden, United
States)
Mandel et al. (1995); Dosemeci et al. (1999); McCredie and Stewart
(1993); Mellemeaard et al. (1994); Schlehofer et al. (1995)
All Centers (Mandel et al.. 1995)
1,732 histologically or cytologically confirmed renal cell carcinoma
cases from 6 studv centers (Mandel et al.. 1995) 1438 renal cell
carcinoma cases from one United States center [Minnesota Cancer
Surveillance Svstem. a SEER rcDortina sitel (Dosemeci et al.. 1999).
368 cases from Denmark (Mellemeaard et al.. 1994). 277 renal cell
carcinoma cases from 10 local urology departments near Heidelberg,
Germanv (Schlehofer et al.. 1995)1. 20-79 vr (20-75 vr. Heidelbers).
1989 1991, identified from hospital surveillance (Germany) national
cancer registries (all other countries), same birth country and cancer
registry (except Australia and the United States), 2,309 population
controls (all countries, with controls >65 yr in the United States
identified from HCFA roles) (Mandel et al.. 1995) 1687 DODulation
controls (Dosemeci et al.. 1999); 396 DODulation referents
(Mellemeaard et al.. 1994). 286 DODulation controls (Schlehofer et al..
1995)1. matched on sex. and aee. in-Dcrson interview with
questionnaire inquiry on specific occupations (4 centers)
Ever exposed to dry-cleaning solvents
1.4(1.1, 1.7)
245
Duration of exposure to dry-cleaning solvents (yr)
1-7
0.2 (0.9, 1.8)
70
8-25
1.7(1.2, 2.4)
78
26-60
1.2(0.9, 1.8)
75
Denmark (Mellemgaard et al.. 1994)
>1 yr exposure duration in dry-cleaning industry,
10 yr before interview
Males
2.3 (0.2, 27)
2
Females
2.9(0.3,33)
2
a -r
>1
to 0\
2 ^
3
LtJ
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-7. Summary of human studies on tetrachloroethylene exposure and kidney cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
cont
New South Wales. Australia (McCrcdic and Stewart. 1993) '
or full occupational history (2 centers); occupation and chemical
grouping as exposure surrogate, OR stratified by sex and adjusted for
age, smoking, BMI, education, and study center, OR reported for males
onlv (Mandel et al.. 1995). In Mellemeaard et al. (1994). OR for
occupational title/exposure >1 year duration and 10 years before
interview and adjusted for age, BMI and smoking. In Dosemeci et al.
(1999). OR rcDortcd for both sexes tosether and separately and adiusted
for age, smoking hypertension, and/or diuretic use, and/or anti-
hvpertension drug use. and BMI. In Schlehofer et al. (1995). OR for
exposure duration >5 years and adjusted for age and smoking
Dry-cleaning industry occupation or job
2.70 (1.08, 6.72)
16
Germany (Heidelberg) (Schlehofer et al.. 1995)
PCE and tetrachlorocarbonate
2.52 (1.23,5.16)
27
United States (Minnesota) (Dosemeci et al.. 1999)
PCE
1.07 (0.7, 1.6)
50
Male
1.12(0.7, 1.7)
42
Female
0.82 (0.3,2.1)
8
Nordic Countries (Denmark, Finland, Norway, Sweden)
Lvnge et al. (2006)
Unexposed
1.00
129
Case-control study among 46,768 Danish, Finnish, Norwegian, and
Swedish men and women employed in 1960 as laundry worker or dry
cleaner, follow-up 1970-1971 to 1997-2001, 210 renal cell carcinoma
cases, 3 controls per case randomly selected from cohort matched on
country, sex, age, calendar period at diagnosis time, occupational task
at 1970 Census proxy for exposure, kidney cancer incidence, RR
adjusted for country, sex, age, calendar period at time of diagnosis
Dry cleaner
0.67 (0.43, 1.05)
29
Other in dry-cleaning
1.15 (0.52,2.53)
9
Unclassifiable
0.76 (0.50, 1.16)
3
Dry cleaner, employment duration, 1964-1979
<1 yr
0.24 (0.03, 2.04)
1
2-4 yr
0.86 (0.28, 2.67)
4
5-9 yr
0.70 (0.32, 1.55)
8
>10 yr
0.75 (0.39, 1.42)
14
Unknown
0.70 (0.15,3.36)
2
a -r
>1
to 0\
2 ^
3
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-7. Summary of human studies on tetrachloroethylene exposure and kidney cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
New South Wales, Australia
McCredie and Stewart CI993)
Dry-cleaning industry occupation or job
6.09 (1.95, 18.9)e
8
147 renal pelvic cancer cases, 20-79 yr, 1989-1990, identified from
hospitals and physicians, 523 population controls, in-person or
telephone interview, job title or industry as exposure surrogate, OR
adjusted for age, sex, and method of interview (for renal cell
carcinomas) and age, sex, interview methods and education (for renal
pelvic cancers)
Finland
Partanen et al. (1991)
Dry-cleaning operator
Not reported
1
338 renal cell carcinoma cases, 20-95 yr, 1977-1987, identified from
Finnish Cancer Registry, 484 population controls matched on birth
year, sex, and survival status at time of interview, mailed interview, job
title or industry for all jobs held 1926-1968, OR adjusted for smoking,
coffee consumption and obesity
Germany, 5 regions
Pesch et al. C2000b)
PCE, JEM
935 histologically confirmed renal cell carcinomal cancer in men and
Medium exposure
1.1 (0.9, 1.4) M
1.2(0.8, 1.8) F
135
28
women, hospital record study, 1991-1995, 4,298 age-sex-matched
population controls, in-person interview, JEM and JTEM for PCE, OR
adjusted for age, study center, smoking
High exposure
1.1 (0.9, 1.4) M
1.3 (0.8, 2.0) F
138
29
Substantial exposure
1.3 (0.9, 1.8) M
0.8(0.3, 1.9) F
55
6
PCE, JTEM
Medium exposure
1.2(0.9, 1.7) M
2.2(0.9, 5.2) F
44
8
High exposure
1.1 (0.7, 1.5) M
1.5 (0.6, 3.8) F
39
6
Substantial exposure
1.3 (0.7, 2.3) M
2.0(0.5, 7.8) F
15
3
a -r
>1
to 0\
2 ^
3
Ltl
~n
H
O
O
-------
a
o
<
s?
cs
>!
o
^5
§•
rs
st
3
>!
sT
St
*3
o
>1
>!
09
o
>!
Si
Si
s
O
I
§•
K) On
2
Table 4-7. Summary of human studies on tetrachloroethylene exposure and kidney cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Montreal, Canada
Parent, 2000
Launderers and dry cleaners
142 histologically confirmed renal cell carcinoma cancer, 1979-1985,
35-70 yr, 533 population control group and 1,900 cancer control group,
in-person interviews, occupational title, OR adjusted age, smoking, and
Any exposure
1.7 (0.6, 4.7)
4
Substantial exposure
0
BMI
Geographic Studies
New York City, NY (United States)
Ma, 2010
Zip codes with number of dry cleaners/km2
10,916 cases with hospital discharge diagnosis of renal or renal pelvis
cancer, 1993-2004, zip code of residential address and dry-cleaner
business number/zip code area as exposure surrogate, crude prevalence
rate ratio (prevalence RR)
0-0.47
1.0a
1,458
0.47-0.90
1.14 (1.03, 1.27)f
2,289
0.90-1.50
1.09 (0.97, 1.21)f
1,838
1.50-2.70
1.17 (1.05, 1.32)f
2,766
2.70-16.43
1.15 (1.01, 1.30)f
2,565
aReferent.
bFor Chang et al. (2003). SMR for kidney and urinary organs.
Tor Sung et al. (20071. SIR for kidney and urinary organs, 10-yr lag period.
dIn McCredie and Stewart (19931. renal cell carcinoma cases from hospitals and physicians in New South Wales, Australia. Of the 489 renal cell carcinoma
cases, 256 were from the Sydney Metropolitan area and were included in the National Cancer Institute's international study (Mandel et al.. 19951.
4^ eIn McCredie and Stewart (19931. OR for renal pelvic cancer.
-^j fIn Ma et al. (20091. rate ratio from negative binomial regression model with main effect for zip code (crude rate ratio). Rate ratios from models with adjustment
for age, race, sex, population density and median household and effect modifiers that vary by exposure category are 1.0 (referent), 1.15 (95% CI: 1.04, 1.27)
[no effect modification], 1.10 (1.00, 1.24) [effect modification by population density], 1.27 (95% CI: 1.13, 1.42) [effect modification by race], and 1.16 (05%
CI: 1.02, 1.33) [effect modification by mean household income and age], for numbers of dry cleaners of 0-0.47, 0.47-0.90, 0.90-1.50, 1.50-2.70, and
2.70-16.43/km2, respectively.
> JEM = job-exposure matrix, HCFA = Health Care Financing Administration, JTEM = job-task-exposure-matrix, PCE = tetrachloroethylene, RDD = relative
^ delivered dose, TWA = time-weighted-average.
O
O
-------
1
2
3
4
5
6
7
8
9
10
11
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
Differences in risk estimates between males and females were reported in three studies;
two studies observed higher point estimates in females (Ji et al.. 2005b; Pesch et al.. 2000b). with
a higher risk estimate for males observed in Dosemeci et al. (1999). Pukkala et al. (2009). in
contrast, did not observe differences in kidney cancer risk estimates between male and female
subjects. It is unclear why apparent differences in sex-specific results were observed in some
studies, although different exposure potentials, different exposure intensities, chance, or residual
confounding are possible alternative explanations (Dosemeci et al.. 1999; NRC. 2010).
In addition to the large cohort and case-control studies, some evidence is found in studies
whose effect estimates are based on fewer observed events and that carry less weight in the
analysis. As expected, the magnitude of the point estimate of the association reported in these
studies is more variable than in the larger studies. Because of the relatively small number of
observed exposed cases in these cohort studies or exposed cases in case-control studies, ranging
from 2 in Antilla et al. (1995) and Boice et al. (1999) to 29 in Selden and Ahlborg (2011). the
statistical power of these lesser-weighted studies is limited. The variation in the association
observed in these studies is consistent with that from studies discussed above that carry greater
weight in the analysis. For the association between kidney cancer and dry cleaning, six studies
reported risk estimates from 0.69 to 0.94 (Andersen et al.. 1999; Asal etal.. 1988)|"malesl; (Boice
et al.. 1999; Lynge et al.. 2006; Lynge and Thygesen. 1990; Pukkala et al.. 2009). three studies
reported risk estimates from 1.0 to 1.08 (Aschengrau et al.. 1993; Blair et al.. 2003; Selden and
Ahlborg. 2011). four studies reported risk estimates from 1.35 to 1.92 (Anttila et al.. 1995;
Calvert. 1976; Delahunt et al.. 1995; Parent et al.. 2000b). and four studies reported risk
estimates from 2.3 to 2.8 (Asal etal.. 1988)rfemalesl; (McCredie and Stewart. 1993;
Mellemgaard et al.. 1994; Schlehofer et al.. 1995).
Several studies had been previously identified based on the relative strengths of their
exposure-assessment methodology. The results from these studies are mixed. Some of these
studies reported no evidence of an increased risk, with relative risks of 0.67 (Lynge et al.. 2006;
dry cleaners). 0.69 (Boice et al.. 1999; routine exposure). 1.04 (Selden and Ahlborg. 2011). and
1.07 (Dosemeci et al.. 1999; tetrachloroethylene exposure). No cases were observed in the group
above the 90th percentile of exposure based on modeling of residential exposure in Aschengrau et
al. (1993). and the overall relative risk for any tetrachloroethylene exposure was 1.08. In
contrast, data from other studies with relatively strong exposure-assessment methods provide
more evidence of an effect, with relative risks of 1.35 (Calvert. 1976; tetrachloroethylene-only).
1.5 (Blair et al.. 2003; medium-high exposure), 1.82 (Anttila et al.. 1995; biological samples).
and 2.52 (Schlehofer et al.. 1995; tetrachloroethylene or tetrachlorocarbonate exposure). The
data from Pesch et al. (2000b). as described earlier, do not indicate a pattern of increasing risk
with increasing exposure among males (odds ratio [OR]: 1.2, 1.1, and 1.3 for medium, high, and
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-77 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
substantial exposure, respectively), or among females, although the overall risk pattern is
stronger among women (OR: 2.2, 1.5, and 2.0 for medium, high, and substantial exposure,
respectively).
Two studies of the same population, an electronics factory in Taiwan, which did not use
an exposure-assessment approach that allowed individual-level classification of exposure,
observed standardized mortality ratios (SMRs) for kidney and urinary organ cancer of 1.18
(95% CI: 0.24, 3.44) (Chang et al.. 2003) and 1.10 (95% CI: 0.62, 1.82) (Sung et al.. 20081
respectively. A geographic-based study reported a relatively constant prevalence rate ratio for
the association between hospital discharge diagnoses for kidney cancer and density of dry
cleaners by zip code of residence (Ma et al.. 2009).
The two studies reporting findings for cancer of the renal pelvis and tetrachloroethylene
were each based on 10 or fewer observations, with the standardized incidence ratio or odds ratio
estimates in these studies of 1.23 (95% CI: 0.39, 2.86) and 6.09 (95% CI: 1.95, 8.9) in Wilson et
al. (2008) and McCredie and Stewart (1993). respectively.
Establishment of an exposure or concentration-response relationship can add to the
weight of evidence for identifying a cancer hazard, but only limited data pertaining to exposure-
response relationships for kidney cancer and tetrachloroethylene exposure are available. Seven
studies presented risk estimates for increasing exposure categories. Four studies used exposure
duration as a proxy (Boice et al.. 1999: Ji et al.. 2005b: Lynge et al.. 2006)(Mandel et al., 1994);
one of these included only five cases in three exposure categories (Boice et al.. 1999). which
limits the potential of this study to assess trends. Three studies used a semiquantitative exposure
surrogate (Blair et al.. 2003: Ma et al.. 2009: Pesch et al.. 2000b). but one of these was a
relatively nonspecific and nonsensitive measure based on zip code area-based density of dry
cleaners (Ma et al.. 2009). A monotonic increasing trend in relative risk with increasing
exposure surrogate was not seen in any of the larger occupational exposure studies with three or
more exposure categories (Lynge et al.. 2006: Mandel et al.. 1995: Pesch et al.. 2000b). In a
smaller study, Blair et al. (2003) reported a higher risk in the higher of two exposure categories
(SMR: 0.3 for little-to-no exposure and 1.5 for medium-to-high exposure). One other study
provided data pertaining to the effect of duration of work. Ji et al. (2005b) reported a higher, but
more imprecise, SIR for females employed as laundry workers and dry cleaners in the 1960 and
1970 Swedish Censuses (SIR: 1.67, 95% CI: 1.07, 2.37) compared to those who were classified
in this type of work only in 1960 (SIR: 1.41, 95% CI: 1.13, 1.71). Neither of the two studies of
renal pelvis cancer reported odds ratio estimates by exposure gradients.
Statistical analyses in all case-control studies except McCredie and Stewart (1993) and
Lynge et al. (2006) controlled for cigarette smoking, a known risk factor for kidney cancer (Asal
et al.. 1988: Aschengrau et al.. 1993: Auperin et al.. 1994: Delahunt et al.. 1995: Dosemeci et al..
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-78 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
1999; Mandel et al.. 1995; Mellemgaard et al.. 1994; Parent et al.. 2000b; Pesch et al.. 2000b;
Schlehofer et al.. 1995). Fewer studies also controlled for body mass index, another risk factor
for kidney cancer (Dosemeci et al.. 1999; Mandel et al.. 1995; Mellemgaard et al.. 1994; Parent
et al.. 2000b). Direct examination of possible confounders is less common in cohort studies
relying on company-supplied or census work history data compared to case-control studies
where information is obtained from study subjects or their proxies. In cohort studies, however,
use of internal controls rather than an external referent group (e.g., national mortality rates) can
minimize effects of potential confounding due to smoking or socioeconomic status, because
exposed and referent subjects are drawn from the same target population. However, only one of
the available cohort studies included an analysis using internal controls, and that study is limited
by the observation of only two kidney cancer cases with routine tetrachloroethylene exposure in
the cohort (Boice et al.. 1999). Effect of smoking as a possible confounder may be assessed
indirectly through examination of risk ratios for other smoking-related sites such as lung cancer.
Several studies observed roughly a 30% increase in lung cancer risk among dry cleaners (Blair et
al.. 2003; Calvert et al.. In Press; Ji et al.. 2005b; Pukkala et al.. 2009; S el den and Ahlborg.
2011). Any expected contribution of smoking to kidney cancer risk will be smaller than that for
lung cancer.
In conclusion, the epidemiologic data provide limited evidence pertaining to
tetrachloroethylene exposure and kidney cancer risk. The studies that support this finding
include the largest international case-control study (245 exposed cases from Australia, Denmark,
Germany, Sweden, and the United States), which reported a relative risk of 1.4
(95% CI: 1.1, 1.7) for any exposure to dry-cleaning solvents (Mandel et al.. 1995). This study
was able to adjust for smoking history, BMI, and other risk factors for kidney cancer. The large
cohort studies, using a more general exposure classification based on national census occupation
data, present more variable results, with relative risks of 0.94, 1.11, and 1.15 in Pukkula et al.
(2009). Travier et al. (2002). and Ji et al. (2005b). respectively. One difference among these
cohort studies is that Travier et al. (2002) and Ji et al. (Ji et al.. 2005b) were based on data from
Sweden, while Pukkula et al. (2009) used data from Sweden, Denmark, Finland, Norway, and
Iceland. Differences between these countries in tetrachloroethylene usage, as was noted by
Lynge et al. (2006). may have introduced an additional source of exposure misclassification in
this multicountry analysis. In addition to the large cohort studies, evidence also comes from
cohort and case-control studies, whose effect estimates are based on fewer observed events.
Smaller studies that do not also have a more sensitive or specific exposure metric carry lesser
weight in the analysis. Eight studies were identified that used a relatively specific exposure-
assessment approach to refine classification of potential tetrachloroethylene exposure in dry-
cleaning settings (Blair et al.. 2003; Calvert et al.. In Press; Lynge et al.. 2006). the aerospace
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-79 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
industry (Boice et al.. 1999). or within a variety of workplaces (Anttila et al.. 1995; Dosemeci et
al.. 1999; Pesch et al.. 2000b; Schlehofer et al.. 1995) or a residential area setting (Aschengrau et
al.. 1993). The results from these studies are mixed, with some studies reporting little or no
evidence of an association (Aschengrau et al.. 1993; Boice et al.. 1999; Lynge et al.. 2006; Pesch
et al.. 2000b XDosemici et al., 1991), and other studies reported elevated risks (Anttila et al..
1995; Blair et al.. 2003; Calvert et al.. In Press; Schlehofer et al.. 1995). An increasing trend in
relative risk with increasing exposure surrogate was not seen in any of the larger occupational
exposure studies with three or more exposure categories (Lynge et al.. 2006)(Mandel et al.,
1994), but some indication of higher risk with higher exposure (or duration) was seen in other
studies (Blair et al.. 2003). As expected, the results from sixteen other studies using a relatively
nonspecific exposure measure (broad occupational title of launderers and dry cleaners, all
workers at factory, density of dry-cleaning establishments by zip code) are more variable and
less precise, reflecting a greater potential for misclassification bias.
4.2.1.3. Bladder Cancer in Humans
Thirty-two epidemiologic studies reporting data on bladder cancer and
tetrachloroethylene exposure were identified. This set of studies includes 13 cohort or nested
case-control studies within a cohort (Andersen et al.. 1999; Blair et al.. 2003; Boice et al.. 1999;
Calvert et al.. In Press; Chang et al.. 2005; Ji and Hemminki. 2005a; Lynge et al.. 2006; Lynge
and Thygesen. 1990; Pukkala et al.. 2009; Selden and Ahlborg. 2011; Sung et al.. 2007; Travier
et al.. 2002; Wilson et al.. 2008). 16 case-control studies of occupational exposures (Burns and
Swanson. 1991; Colt et al.. 2011; Dryson et al.. 2008; Gaertner et al.. 2004; Kogevinas et al..
2003; Pesch et al.. 2000b; Reulen et al.. 2007; Schoenberg et al.. 1984b; Siemiatvcki. 1991;
Silverman et al.. 1989a; Silverman et al.. 1989b; Smith et al.. 1985; Steineck et al.. 1990;
Swanson and Burns. 1995; Teschke et al.. 1997; Zheng et al.. 2002). and 3 studies of residential
exposure through contaminated drinking water (Aschengrau et al.. 1993; Mallin. 1990; Vieiraet
al.. 2005b). These 32 studies represent the core studies evaluated by EPA, as described in more
detail below. Two other cohort studies and one case-control study included information on
tetrachloroethylene but did not report risk estimates for bladder cancer (Anttila et al.. 1995; Colt
et al.. 2004; Radican et al.. 2008). and so were not evaluated further. The peer-reviewed
literature also contains a meta-analysis that examined dry cleaning and bladder cancer (Reulen et
al.. 2007).
There is some overlap in the study populations among these studies: Travier et al. (2002)
used occupational data from the Swedish national census, and Lynge and Thygsen (1990) used a
similar design in Denmark; Andersen et al. (1999) and Lynge et al. (2006) expanded these
studies to include Denmark, Finland, and Norway in addition to Sweden, and Pukkala et al.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-80 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
(2009) added Iceland to this set. Pesch et al. (2000b) is a large case-control study examining
urothelial cancers, a grouping of bladder, ureter, and renal pelvis neoplasms, with exposure
information on tetrachloroethylene. Kogevinas et al. (2002), a pooled analysis of 11 studies
conducted in European countries between 1976 and 1996, includes the dry cleaning but not the
tetrachloroethylene exposure observations in males in Pesch et al. (2000b). Kogevinas does not
provide information on women; 't Mannetje et al. (1999) pooled observations in women in these
11 studies but did not report findings on dry-cleaner and laundry workers.
Appendix B reviews the design, exposure-assessment approach, and statistical
methodology for each study. Most studies were of the inhalation route, of occupational
exposure, and unable to quantify tetrachloroethylene exposure.
4.2.1.3.1. Consideration of exposure-assessment methodology
Many studies examine occupational titles such as dry cleaner, launderer, and presser as
surrogate for tetrachloroethylene, given its widespread use from 1960 onward in the United
States and Europe (Andersen et al.. 1999; Blair et al.. 2003; Burns and Swanson. 1991; Calvert et
al.. In Press; Colt et al.. 2011; Dryson et al.. 2008; Gaertner et al.. 2004; Ji and Hemminki.
2005a; Lynge et al.. 2006; Lynge and Thygesen. 1990; Pukkala et al.. 2009; Reulen et al.. 2007;
Reulen et al.. 2008; Schoenberg et al.. 1984a; Silverman et al.. 1990; Silverman et al.. 1989a;
Silverman et al.. 1989b; Smith et al.. 1985; Steineck et al.. 1990; Swanson and Burns. 1995;
Teschke et al.. 1997; Travier et al.. 2002; Wilson et al.. 2008; Zheng et al.. 2002)(Kogevinas et
al., 2002). Six studies conducted in Nordic countries are either based on the entire Swedish
population or combined populations of several Nordic countries; strengths of these studies are
their use of job titles as recorded in census databases and ascertainment of cancer incidence
using national cancer registries (Andersen et al.. 1999; Ji and Hemminki. 2005a; Lynge et al..
2006; Pukkala et al.. 2009; Travier et al.. 2002; Wilson et al.. 2008). Studies examining
mortality among U.S. dry-cleaner and laundry workers (Blair et al.. 2003; Calvert et al.. In Press)
are of smaller cohorts than the Nordic studies, with fewer observed bladder cancer events.
The exposure surrogate in studies of dry-cleaners and laundry workers is a broad
category containing jobs of differing potential for tetrachloroethylene exposure. Thus, these
studies have a greater potential for exposure misclassification bias compared to studies with
exposure potential to tetrachloroethylene assigned by exposure matrix approaches. Two studies
used additional information pertaining to work environment to refine the exposure (Calvert et al..
In Press; Lynge et al.. 2006). Lynge et al. (2006). using job titles reported in the 1970 Census,
identified subjects based on an occupational code of—liandry and dry-cleaning worker" or an
industry code of—laundry and dry cleanig." Additional information to refine this occupational
classification was sought for incident cancer cases, including bladder cancer, within this defined
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-81 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
cohort. Five controls, matched to the cases by country, sex, age, and calendar period, were also
included in the study. The additional information included handwritten task information from
the census forms from Denmark and Norway, pension databases in Denmark and Finland, and
next-of-kin interviews in Norway and Sweden. Exposure classification categories were dry
cleaner (defined as dry cleaners and supporting staff if employed in a business of <10 workers),
other job titles in dry cleaning (launderers and pressers), unexposed (job title reported on 1970
census was other than dry cleaning), or unclassifiable (information was lacking to identify job
title of subject). The unclassifiable category represented 57 of 351 bladder cancer cases (16%)
and 234 out of 1,482 controls (16%). The study by Calvert et al. included an analysis of
subjects who worked for one or more years before 1960 in one or more shops known to use
tetrachloroethylene as the primary solvent (Calvert et al.. In Press). The cohort was stratified
into two groups based on the level of certainty that the worker was employed only in facilities
using tetrachloroethylene as the primary solvent exposure; tetrachloroethylene-only and
tetrachloroethylene plus. However, there were no bladder cancer deaths among this subset
(n = 618) of tetrachloroethylene-only subjects. Three additional studies used a semiquantitative
or quantitative exposure metric. Blair et al. (2003) used an exposure metric for semiquantitative
cumulative exposure between dry-cleaning and laundry workers. The case-control study by
Siemiatycki (1991) used a job exposure matrix (JEM) based on occupational titles for
tetrachloroethylene, and another case-control study used a JEM and one with information on
specific tasks, a job-task exposure matrix (JTEM), with semiquantitative exposure assessment
across a variety of jobs (Pesch et al.. 2000b).
Two other cohorts with potential tetrachloroethylene exposure in manufacturing settings
have been examined. These studies include aerospace workers in the United States (Boice et al..
1999) and electronic factory workers in Taiwan (Chang et al.. 2005; Sung et al.. 2007). Boice et
al. (1999) used an exposure assessment based on a job-exposure matrix to classify exposures. In
contrast, the exposures in the Taiwan studies included multiple solvents, and tetrachloroethylene
exposure was not linked to individual workers (Chang et al.. 2005; Sung et al.. 2007).
Three geographic studies focused on residential proximity to drinking water sources
contaminated with tetrachloroethylene and other solvents. Mallin (1990) examines incidence
and mortality by county in Illinois, with the exposure surrogate assigned uniformly to all
subjects. Two other studies in Cape Cod, MA, used either an exposure model incorporating
tetrachloroethylene leaching and characteristics of the community water distribution system
(Aschengrau et al.. 1993) or residential proximity to Superfund sites and a generalized additive
model that incorporates smoothing approaches and adjusts for covariates (Vieira et al.. 2005b).
In summary, four studies with bladder cancer data assigned tetrachloroethylene exposure
to individuals within the study using a job exposure matrix (Blair et al.. 2003; Boice et al.. 1999;
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-82 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Pesch et al.. 2000b)or an exposure model (Aschengrau et al.. 1993). One other study sought
additional data using a questionnaire for use in refining potential exposure within dry-cleaning
settings (Lynge et al.. 2006). The relative specificity of these exposure-assessment approaches
strengthens their ability to identify cancer hazards compared to studies with broader and less
sensitive exposure-assessment approaches.
4.2.1.3.2. Summary of results
Seven studies evaluated by EPA reported estimated relative risks based on a large
number of observed events; 50 or more deaths/incident cases in cohort studies (Andersen et al..
1999; 2005a; Pukkala et al.. 2009; Travier et al.. 2002; Wilson et al.. 2008). or 50 or more
exposed cases in case-control studies (Lynge et al.. 2006; Pesch et al.. 2000b). with sufficient
power to detect a twofold elevation in estimated risk. Pukkala et al. (2009) updates the analysis
of Andersen et al. (1999) adding data from a 5th country, Iceland, and extending follow-up to
2005, and is preferred over Andersen et al. (1999) for these reasons. The five1 large cohort
studies observed a standardized incidence ratio or odds ratio estimate of 1.01 (95% CI: 0.86,
1.19), 1.08 (95% CI: 0.98, 1.23), 1.14 (95% CI: 0.89, 1.45), 1.27 (95% CI: 1.08, 1.48), and 1.44
(95% CI: 1.07, 1.93) in Travier et al. (2002). Pukkala et al. (2009). Wilson et al. (2008). Ji et al.,
and Lynge et al. (2006). respectively, for the association between bladder cancer risk and ever
having a job title of dry cleaner or laundry worker (see Table 4-8). The Lynge et al. (2006)
results were slightly higher among the subgroup from Denmark and Norway, in which the
number of unclassifiable data was negligible (relative risk 1.69, 95% CI: 1.18, 2.43). The large
case-control study by Pesch et al. (2000b) reported an odds ratio of 0.8 (95% CI: 0.6, 1.2), 1.3
(95% CI: 0.9, 1.7), and 1.8 (95% CI: 1.2, 2.7) for medium, high, and substantial exposure,
respectively, compared to low exposure, based on the JTEM approach.
Additional evidence is found in studies whose effect estimates are based on fewer
observed events and that carry lesser weight in the analysis. As expected, the magnitude of the
point estimate of the association reported in these studies is more variable than in the larger
studies: 4 studies report relative risks between 0.7 and 0.91 (Colt et al.. 2011 Tmalesl) (Boice et
al.. 1999; Dryson et al.. 2008; Lynge and Thygesen. 1990). 10 studies report relative risks
between 1.2 and 1.9 (Blair et al.. 2003) [females]; (Aschengrau et al.. 1993; Burns and Swanson.
1991; Colt et al.. 2011; Gaertner et al.. 2004; Schoenberg et al.. 1984; Siemiatvcki. 1991; Smith
et al.. 1985; Steineck et al.. 1990)(Kogevinas et al., 2002), and 3 studies report relative risk
estimates >2.0 (Reulen et al.. 2007; Teschke et al.. 1997; Zheng et al.. 2002). Except for the
estimate from Reulen et al. (2007) (RR: 2.7, 95% CI: 1.1, 6.6), all of the 95% CIs of these
1 Andersen et al. (1999) is not included in this summary of the data from the individual studies because it was
updated and expanded in the analysis by Pukkala et al. (2009).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-83 DRAFT-DO NOT CITE OR QUOTE
-------
1 estimates overlap 1.0. Because of the relatively small number of observed cases in these cohort
2 studies or exposed cases in case-control studies, ranging from 2 in Boice et al. (1999) to 19 in the
3 pooled study of Kogevinas et al. (2002); the statistical power of these lesser-weighted studies is
4 limited.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-84 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-8. Summary of human studies on tetrachloroethylene exposure and bladder cancer
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Cohort Studies
Biologically monitored workers
Anttila et al. (1995)
All subjects
Not reported3
849 Finnish men and women, blood PCE [0.4 |imol/L in females
and 0.7 |imol/L in males (median)], follow-up 1974-1992,
external referents (SIR)
Aerospace workers (Lockheed)
Boice et al. (1999)
Routine exposure to PCE
0.70 (0.09, 2.53)
2
77,965 (n = 2,631 with routine PCE exposure and n = 3,199 with
intermittent-routine PCE exposure), began work during or after
1960, worked at least 1 yr, follow-up 1960-1996, job exposure
matrix without quantitative estimate of PCE intensity, 1987-1988
8-h TWA PCE concentration (atmospheric monitoring) 3 ppm
[mean] and 9.5 ppm [median], external reference for routine
exposure (SMR) and internal references (workers with no
chemical exposures) for routine-intermittent PCE exposure (RR)
Routine-Intermittent exposure to PCE
Not reported13
Electronic factory workers (Taiwan)
Chans et al. (2005); Suns et al. (2007)
All Subjects
86,868 (n = 70,735 female), follow-up 1979-1997, multiple
solvents exposure, does not identify PCE exposure to individual
subjects, cancer incidence, external referents (SIR) (Chang et al..
2005);
63,982 females, follow-up 1979-2001, factory employment proxy
for exposure, multiple solvents exposures and PCE not identified
to individual subjects, cancer incidence, external referents,
analvses lassed 5 vr (SIR) (Suns et al.. 2007)
Males
1.06 (0.45, 2.08)°
8
Females
1.09 (0.56, 1.91)°
12
Females
0.34 (0.07, 1.00)
12
Aircraft maintenance workers from Hill Air Force Base
Radican et al. (2008)
Any PCE exposure
Not reported
10,461 men and 3,605 women (total n = 14,066, n = 10,256 ever
exposed to mixed solvents, 851 ever-exposed to PCE), employed
at least 1 yrfrom 1952 to 1956, follow-up 1973-2000, job
exposure matrix (intensity), internal referent (workers with no
chemical exposures) (RR)
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-8. Summary of human studies on tetrachloroethylene exposure and bladder cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Dry-cleaner and laundry workers
Andersen et al. (1999)
All laundry worker and dry cleaners
1.00 (0.83, 1.21)
119
29,333 men and women identified in 1960 Census (Sweden) or
1970 Census (Denmark, Finland, Norway), follow-up 1971-1987
or 1991, PCE not identified to individual subjects, external
referents (SIR)
Males
1.14(0.87, 1.46)
62
Females
0.89 (0.68, 1.16)
57
Blair etal. (2003^
All subjects
1.3 (0.7, 2.4)
12
5,369 U.S. men and women laundry and dry-cleaning union
members (1945-1978), follow-up 1979-1993, semiquantitative
cumulative exposure surrogate to dry clean solvents, cancer
mortality, external referents (SMR)
Semiquantitative exposure score
Little to no exposure
1.4 (0.4, 3.2)
5
Medium to high exposure
1.5 (0.6, 3.1)
7
Ji et al. C2005a*)
Male laundry workers and dry cleaners in 1960
Census
1.27 (1.08, 1.48)
157
9,255 Swedish men employed in 1960 as laundry worker or dry
cleaner, follow-up 1961-2000, PCE not identified to individual
subjects, external referent (SIR) and adjusted for age, period and
socioeconomic status
Male laundry workers and dry cleaners in 1960
Census
1.13 (0.96, 1.3 l)d
157
Male laundry workers and dry cleaners in both 1960
and 1970 Censuses
1.03 (0.80, 1.29)d
67
Male laundry workers and dry cleaners in 1960, 1970
and 1980 Censuses
0.86 (0.51, 1.28)d
19
Female laundry workers and dry cleaners
Not reported
Lvnge and Thvgsen (1990)
All laundry worker and dry cleaners
0.74 (0.41, 1.25)
14
10,600 Danish men and women, 20-64 yr old, employed in 1970
as laundry worker, dry cleaners and textile dye workers, follow-up
1970-1980, external referents (SIR)
Males
0.62 (0.23, 1.35)
6
Females
0.88 (0.38, 1.73)
8
a -r
>1
to 0\
2 ^
3
00
On
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-8. Summary of human studies on tetrachloroethylene exposure and bladder cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Pukkala et al. (2009)
Launderer and dry cleaner
1.08 (0.98, 1.23)
434
Men and women participating in national census on or before
1990, 5 Nordic countries (Denmark, Finland, Iceland, Norway,
Sweden), 30-64 yr, follow-up 2005, occupational title of launderer
and dry cleaner in any census, external referents (SIR)
Male
1.10(0.95, 1.27)
186
Female
1.07 (0.95, 1.22)
248
Calvert et al. (In Press)
All subjects
1.81 (0.87, 3.33)
10
1,704 U.S. men and women dry-cleaning union member in CA, IL,
MI, NY follow-up 1940-2004 (618 subjects worked for one or
more years prior to 1960 only at shops where PCE was the
Exposure duration/time since 1st employment
<5 yr/<20 yr
0
primary cleaning solvent, identified as PCE-only exposure), cancer
mortality (SMR)
<5 yr/>20 yr
0.53 (0.03, 2.52)
1
>5 yr/<20 yr
0
>5 yr/>20 yr
4.08 (2.13,7.12)
9
PCE-only subjects
0
Selden and Ahlborg (2011)
Dry-cleaners and laundry workers (females)
0.92 (0.65, 1.26)
38
9,440 Swedish men (n = 2,810) and women (n = 9,440) in 461
washing and dry-cleaning establishments, identified by employer
in mid-1980s, employed 1973-1983, follow-up 1985-2000
Travier et al. (2002)
All subjects, 1960 or 1970 Census in laundry and dry
cleaner occupation and industry
1.01 (0.86, 1.19)
145
Swedish men and women identified in 1960, 1970, or both
Censuses as laundry worker, dry cleaner, or presser (occupational
title) or in the laundry, ironing, or dyeing industry, follow-up
1971-1989, separates launders and dry cleaners form pressers,
external referents (SIR)
All subjects in 1960 and 1970 in laundry and dry
cleaner occupation and industry
1.00(0.61, 1.63)
16
a -r
>1
to 0\
2 ^
3
00
<1
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-8. Summary of human studies on tetrachloroethylene exposure and bladder cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Wilson etal. (2008)
All subjects, laundry and dry cleaning occupation
1.14 (0.89, 1.45)
68
Swedish men and women identified in 1960 or 1970 as laundry
worker or dry cleaner (occupation) or in laundry, ironing and
dyeing industry, follow-up 1971-1989, external referents (SIR),
transitional cell carcinoma
Males
1.23 (0.83, 1.74)
31
Females
1.07 (0.75, 1.47)
37
Case-Control Studies
Upper Cape Cod, MA (United States)
Aschengrau et al. (1993). Vieira (2005b)
Any PCE
1.39 (0.67, 2.91)
13
63 bladder cancer cases, 1968-1980, Massachusetts Cancer
Registry, 852 population controls, residential history, ordinal
estimate of PCE-contaminated water (RDD) from exposure model
(Aschensrau et al.. 1993) or seosranhical information system and
Droximitv to groundw ater dIluhc (Vieira et al.. 2005b). OR
adjusted for sex, age at diagnosis, vital status at interview,
education, cigarette smoking, and urinary tract infection (both
studies), and. oast occupational exposure (Aschensrau et al.. 1993)
RDD >90lh percentile
4.03 (0.65,25.10)
4
—Htospot" SW of MMR
-2.5 (CI not
reported)
Metropolitan Detroit, MI (United States)
Burns and Swanson (1991); Swanson and Burns (1995)
Usual occupation as dry-cleaning workers
1.9 (0.7, 4.9)
8
2,160 histologically confirmed bladder cancer cases in men and
women, 40-84 yr old, Metropolitan Detroit Cancer Surveillance
System, 3.979 rectal or colon cancer controls, telephone interview,
longest period (usual) employed in occupation or industry, OR
adjusted for cigarette smoking, race, sex, and age at diagnosis
Males
Not reported
2
Females
2.0 (0.7, 6.2)
6
Usual industry in dry cleaner and laundry
1.2 (0.6, 2.4)
15
New Hampshire (United States)
Colt et al. (2004)
Launderers and dry cleaners
459 bladder cancer cases, 1994-1998, New Hampshire State
Cancer Registry, 25-74 yr, 665 populations controls, 1993-1997,
occupation as exposure surrogate, OR adjusted for 5-yr age group
and smoking
Males
Not reported
5
Females
0
a -r
>1
to 0\
2 ^
3
00
00
~n
H
o
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-8. Summary of human studies on tetrachloroethylene exposure and bladder cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Maine, Vermont, and New Hampshire (United States)
Coltetal. (2011)
Occupation: Laundering and dry-cleaning machine
operators and tenders
1,158 patients, aged 30-79, newly diagnosed with histologically
confirmed bladder cancer, 2001-2004, ascertained from hospital
pathology departments, hospital cancer registries and state cancer
registries, 1,402 population controls frequency matched by age
(within 5 yr), state and gender, occupational histories through
interview coded by occupation (SOC 7658)) and industry (SIC
721), OR for occupation or industry category compared to other
never employed in that category, adjusted for age, race, Hispanic
ethnicity, state, smoking status, and employment in a high risk
occupation
Males
Not reported
5
Females
0.45 (0.03, 7.46)
1
Industry: Laundry, cleaning and garment services
Males
0.91 (0.41,2.03)
14
Females
1.50 (0.50, 4.50)
10
New Zealand
Drvson et al. (2008)
Textile bleaching, dyeing and cleaning machine
operators
0.81 (0.19, 3.54)
3
213 bladder cancer cases, 25-70 yr, 2003-2004, New Zealand
Cancer Registry, 471 population controls, occupational title, OR
adjusted for sex, smoking, SES
Canada, 7 Provinces
Gaertner et al. (2004)
Drycleaner
1.24 (0.23, 6.64)
5
887 histologically confirmed bladder cancer, 20-74 yr, 2,847
population controls, Province Cancer Registry, mailed
questionnaire, occupational title as exposure surrogate, OR
adjusted for age, province, race, smoking status, consumption of
fruit, fried food, and coffee, and past occupational exposure.
European Pooled Study (Denmark, France, Germany, Greece, Italy, Spain)
Kogevinas et al., 20021
Launderers, dry cleaners and pressers
1.24 (0.67, 2.31)
19
Pooled study of 3,346 male bladder cancer cases, 30-79 yr, study-
specific groups of 6,840 controls, occupational title, OR adjusted
for age, smoking, and study center
Nordic Countries (Denmark, Finland, Norway, Sweden)
Lvnge et al. (2006)
Unexposed
1.00
188
Case-control study among 46,768 Danish, Finnish, Norwegian,
and Swedish men and women employed in 1960 as laundry
Dry cleaner
1.44 (1.07, 1.93)e
93
a -r
>1
to 0\
2 ^
3
00
VO
~n
H
o
O
-------
a
o
<
s:
rs
>!
o
^5
§•
rs
st
3
>!
sT
St
o
>1
>!
Oq
o
>!
Si
Si
s
O
Sfc
I
§•
K) On
2
Table 4-8. Summary of human studies on tetrachloroethylene exposure and bladder cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Other in dry-cleaning
1.08 (0.55, 2.1 l)e
12
worker or dry cleaner, follow-up 1970-1971 to 1997-2001, 351
Unclassifiable
1.24 (0.83, 1.83)e
57
bladder cancer cases, 3 controls per case randomly selected from
cohort matched on country, sex, age, calendar period at diagnosis
time, occupational task at 1970 Census proxy for exposure,
bladder cancer incidence (excluding in-situ), RR adjusted for
matching criteria
Dry cleaner
1.69 (1.18, 2.43)ef
15
Other in dry-cleaning
1.13 (0.51, 2.50)e'f
6
Unclassifiable
Not reportede
1
Dry cleaner, smoking adjusted
1.25 (0.79, 1.98)8
Dry cleaner, employment duration, 1964-1979
<1 yr
1.50 (0.57, 3.96)e
6
2-4 yr
2.39 (1.09, 5.22)e
10
5-9 yr
0.92 (0.52, 1.59)e
17
>10 yr
1.57 (1.07 2.29)e
53
Unknown
1.97 (0.64, 6.05)e
6
vo
o
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-8. Summary of human studies on tetrachloroethylene exposure and bladder cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Germany, 5 regions
Pesch et al. (2000b)
PCE, JEM
1,035 histologically confirmed urothelial cancer in men and
Medium exposure
1.1 (0.9, 1.3) M
1.8(1.0, 3.0) F
162
21
women, hospital record study, 1991-1995, 4,298 population
controls, in-person interview, JEM and JTEM for PCE, OR
adjusted for age, study center, smoking
High exposure
1.2(1.0, 1.5) M
1.0(0.6, 1.9) F
172
16
Substantial exposure
1.4(1.0, 1.9) M
0.7 (0.2, 2.5)
71
3
PCE, JTEM
Medium exposure
0.8 (0.6, 1.2)
47
High exposure
1.3 (0.9, 1.7)
74
Substantial exposure
1.8(1.2,2.7)
36
Belgium, Limburg Region
Reulen et al. (2007)
Domestic helpers, cleaners, and launderers
2.7(1.1,6.6)
14
202 histologically confirmed transitional cell carcinoma cases,
40-96 yr, Limburg Cancer Registry, 390 population controls, in-
person interview, occupational title, OR adjusted for age, sex,
smoking status, number cigarettes, years smoked, education
New Jersey (United States)
Schoenberg et al. (1984a)
Dry-cleaning workers
1.33 (0.50, 3.58)
7
Histologically confirmed bladder cancer cases (658 Caucasian
men), 1978-1979, 21-84 yr, age-stratified population controls
(1,258 Caucasian men) identified through RDD or HCFA register,
in-person interview with questionnaire, industry and job title
surrogate exposure metric, OR adjusted for age and cigarette
smoking
a -r
>1
to 0\
2 ^
3
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-8. Summary of human studies on tetrachloroethylene exposure and bladder cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Montreal, Canada
Siemiatvcki (199 Is)
Launderers and dry cleaners
Histologically confirmed bladder cancer, 1979-1985, 35-70 yr,
population control group and cancer control group, in-person
interviews, occupational title and JEM for PCE, OR adjusted age,
Any exposure
1.6(0.9,3.1)
10
Substantial exposure
1.9 (0.9, 4.2)
7
family income, and cigarette index, 90% CI
National Bladder Cancer study
Silverman et al. (1990; 1989a: 1989b): Smith etal. (1985)
Laundry and dry cleaners, males and females
Histologically confirmed bladder cancer cases (2,226 men, 733
women), 1977-1978, 21-64 yr, 5,757 population controls, in-
person interview, occupational title as exposure surrogate, OR
adiusted for smokins (Silverman et al.. 1990) and cmolov merit in
other hish-risk occupation (Silverman et al.. 1989a) and ase. sex.
and smokins (<20/d. >20 to <40/d.. >40/d (Smith et al.. 1985)
Nonsmoker
1.31 (0.85,2.03)
Not reported
Former smoker
2.99 (1.80, 4.97)
Not reported
Current smoker
3.94 (2.39, 6.51)
Not reported
Laundry and dry cleaners, non-Caucasian males
2.8(1.1,7.4)
11
<5 yr employment duration
5.3 (CI not reported)
7
>5 yr employment duration
1.8 (CI not reported)
4
p-valuc for linear trend
p = 0.016
Laundry and dry cleaners, females
1.4 (08, 2.6)
23
Stockholm, Sweden
Steineck et al. (1990)
Dry cleaner
1.2 (0.2, 9.2)
2
Bladder cancer cases in males, birth years, 1911-1945 and living
in County of Stockholm 1985-1987, population controls, mailed
questionnaire, occupational title as surrogate, OR adjusted for
birth year and smoking
British Columbia, Canada
Teschke et al. (1997)
Laundry and dry-cleaner workers
2.3 (0.4, 13.9)
5
Histologically confirmed bladder cancer cases (excluding in situ)
from British Columbia Cancer Agency in men and women,
1990-1991, >19 yr, population controls, in-person or telephone
interviews, occupation and industry as surrogates, OR adjusted for
sex, age, cigarette smoking
Exposure surrogate lagged 20 yr
1.8(0.3, 11.3)
4
Dry cleaners
Not reported
3
a -r
>1
to 0\
2 ^
3
\o
to
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-8. Summary of human studies on tetrachloroethylene exposure and bladder cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Iowa, United States
Zheng et al. (2002)
Laundering and dry cleaning occupation
Histologically confirmed in situ and invasive bladder cancer from
Iowa state health registry records in men and women, 1986-1989,
40-85 yr, population controls, in-person interview, occupation and
industry as surrogate, OR adjusted for age, lifetime pack-years of
cigarette smoking, and first-degree relative with bladder cancer
Males
Not reported
Females
9.3 (0.9, 94.8)
3/1
Duration of employment
<10 yr
2/0
>10 yr
2.1 (0.1, 36.9)
1/1
Geographic Studies
Illinois, 8 NW counties
Mallin (19901
Winnebago County
712 bladder cancer cases in Caucasian men and women,
1978-1985, residence as exposure surrogate, solvent-contaminated
municipal drinking water wells in Winnebego County [multiple
Males
0.96 (0.8, l.l)h
250
1.39(1.1, l.lf
76
solvents including PCE, <1-5.1 ppb], incidence and mortality
rates of U.S. population as referent (SIR, SMR)
Females
1.03 (0.8, 1.3)h
96
1.40(1.0, 1.9)a
35
Meta-analysis
Laundry and dry-cleaning workers
1.27 (0.95, 1.71)J
Reulen et al. (20081
Cohort studies
0.82 (0.54, 1.25)
Case-control studies
1.66 (1.23, 2.24)
a -r
>1
to 0\
2 ^
3
\o
LtJ
~n
H
O
O
-------
a to ^ Table 4-8. Summary of human studies on tetrachloroethylene exposure and bladder cancer (continued)
cs ^ <3'
S ^ incidence.
11 S bFor Boice et al. (19991. Relative risks for employment duration from Poisson regression with internal referents of factory workers not exposed to any solvent
5 ; s and with adjustment for date of birth, date first employed, date of finishing employment, race, and sex.
~ 5 5 Tor Chang et al. (20051. SIR for urinary organ neoplasms given bladder cancer SIR is not identified separately.
§ §_ ^ dSmoking-corrected SIR obtained by dividing SIR by 35% of the excess of lung cancer risk (assumed proportion of risk between lung and bladder cancer
•2 g, associated with smoking 20 cigarettes/d).
"o H ^ cIn Lynge et al. (2006), odds ratio from logistic regression adjusted for country, sex, age, and calendar period at time of diagnosis.
S- f In Lynge et al. (20061. odds ratio—Norway and Denmark, countries with better exposure information.
2 8In Lynge et al. (20061. smoking adjusted odds ratio for subjects from Norway and Sweden,
g 5^ Mortality.
^ 'In Kogevinas et al. (2002) includes the following case-control studies—Claude et al. (19881. Cordier et al. (19931. Hours et al. (19941. Gonzalez et al. (19891.
^ Jensen et al. (1987). Pesch et al. (2000b). Pohlabeln et al. (1999). Porru et al. (1996). Rebelakos et al. (1985). Serra et al. (2000). and Vineis et al. (1985).
includes Andersen et al. (1999). Burns et al. (1991). Bouchardy et al. (2002). Colt et al. (2004). Gaertner et al. (2004). Schoenberg et al. (1984a). Siemiatycki
(1991). Silverman et al. (1989b). Silverman et al. (1990). Steineck et al. (1990). Swanson et al. (1995). Teschke et al. (1997) Travier et al. (2002). and Zheng et
al. (2002).
HCFA = Health Care Financing Administration, JEM = job-exposure matrix, MMR = Massachusetts Military Reservation, NCI = National Cancer Institute,
PCE = tetrachloroethylene, RDD = random digit dialing.
vo
~n
H
O
O
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Five studies had been previously identified based on the relative strengths of their
exposure-assessment methodology. The results from four of these studies provide additional
evidence of an association, with relative risks of 1.44 (Lynge et al.. 2006). 1.5 (Blair et al., 2006)
(medium-high exposure), 4.03 (Aschengrau et al.. 1993) (>90th percentile exposure), and the
exposure-response gradient seen in Pesch et al. (2000b). Although a SMR of 2.59 (95%
CI: 1.24, 4.76) was reported among workers with exposure to tetrachloroethylene and possibly
other dry-cleaning solvents (10 exposed cases), no bladder cancer deaths were observed among a
subgroup with a higher certainty of exposure only to tetrachloroethylene (Calvert et al.. In Press).
Statistical analyses in all case-control studies controlled for cigarette smoking, a known
risk factor for bladder cancer. The potential effect modification by smoking history is also an
important issue but has been examined in only one study (Smith et al.. 1985). In the analysis
stratified by smoking status, adjusted ORs for the association between laundry or dry-cleaning
work (based on occupational title from interview data) and bladder cancer incidence of 1.31
(95% CI: 0.85, 2.03) among nonsmokers, 2.99 (95% CI: 1.80, 4.97) among former smokers, and
3.94 (95%) CI: 2.39, 6.51) among current smokers were seen.
Three studies of weaker exposure-assessment approaches observed odds ratio or
standardized incidence ratio estimates of 0.34 (95%> CI: 0.07, 1.00), 1.39 (95%> CI: 1.1, 1.7;
males) and 1.40 (95%> CI: 1.0, 1.9; females), and 2.5 (CI not reported) for the association
between bladder cancer and employment in a manufacturing plant (Sung et al.. 2007) or
residential proximity to groundwater contamination (Mallin. 1990: Vieira et al.. 2005b). These
studies carry lower weight in the analyses because of their low level of detail on
tetrachloroethylene exposure.
Reulen et al.(2008)'s meta-analysis of occupational titles and bladder cancer included 14
studies reporting relative risk estimates for dry-cleaners and laundry workers. The pooled
relative risk estimate for employment in these industries was 1.27 (95%> CI: 0.95, 1.71). While
Reulen et al. (2008)included many of the studies identified above, they do not include the
cohorts of Calvert et al. , Blair et al. (2003). Ji et al.(2005b). and Pukkala et al. (2009). or the
case-control studies of Kogevinas et al. (2002) and Lynge et al. (2006). Other differences
between Reulen et al. (2008) and this analysis are the inclusion of Bouchardy et al. (2002) who
reported a odds ratio estimate for the association between bladder cancer and cleaning, and
personal services—a broad category that included dry cleaners, laundry workers, chimney
sweeps, hairdressers, and other cleaning occupations not included in the EPA analysis due to the
lack of data specific for dry-cleaners and laundry workers. Despite the differences in the specific
studies included in this analysis, the results are similar to that of the EPA's evaluation, indicating
a small (10—40%>) increased risk.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-95 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Establishment of an exposure or concentration-response relationship can add to the
weight of evidence for identifying a cancer hazard, but only limited data pertaining to exposure-
response relationships for bladder cancer and tetrachloroethylene exposure are available. As
described previously, effect estimates of 0.8 (95% CI: 0.6, 1.2), 1.3 (95% CI: 0.9, 1.7), and 1.8
(95%) CI: 1.2, 2.7) for medium, high, and substantial exposure, respectively, based on JTEM
exposure data were reported in the large case-control study by Pesch et al. (2000b). Some
additional information on exposure-response relationships comes from lesser-weighted studies.
Two of the smaller studies with semiquantitative exposure surrogates observed larger effect
measures for the highest exposure category than for overall exposure. In Aschengrau et al.
(1993). the adjusted OR was 4.03 (95%> CI: 0.65, 25.10) for the >90th percentile of the relative
delivered dose, compared with 1.39 (95%> CI: 0.67, 2.91) for any tetrachloroethylene exposure.
Siemiatycki (1991) reported an adjusted OR of 1.9 (95%> CI: 0.9, 4.2) for substantial exposure
and 1.6 (95%> CI: 0.9, 3.1) for any exposure. In the third study with semiquantitative exposure
measurement, the SMR in Blair et al. (2003) was 1.5 (95%> CI: 0.6, 3.1) for the medium-to-high
cumulative exposure, 1.4 (95%> CI: 0.4, 3.2) for the little-to-no exposure category, and 1.3
(95%o CI: 0.7, 2.4) among all cohort members (laundry and dry-cleaning union members). Other
studies examined duration of laundry or dry-cleaning work. Two studies did not observe
increasing patterns of risk with increasing employment durations as measured by census
occupation codes from two or more periods (Ji and Hemminki. 2005a: Travier et al.. 2002). and
one study observed a lower risk with higher duration of laundry and dry-cleaning work based on
employment duration data collected in interviews with cases and controls [trendp-value = 0.016
for the adjusted OR estimate of 5.3 for <5 years and 1.8 for >5 years duration in laundry and
drying cleaning work, respectively (Silverman et al.. 1989a)l. Another study using 1960 and
1970 Census data from Nordic countries reported a nonmonotonic pattern of increasing risk, with
adjusted relative risks of 1.50, 2.39, 0.92, and 1.57 for duration of dry-cleaning work from
1964-1979 of <1, 2-4, 4-9, and >10 years, respectively, compared to subjects never employed
as a dry cleaner or in a shop with <10 employees1 (Lynge et al.. 2006). For the job held in 1970,
Lynge et al. (2006) relied upon a biography of dry-cleaning shop owners, the yellow pages of
local telephone books for self-employed persons and national pension system records to assess
length of employment for Danish subjects, national pension records for Finnish subjects,2 and
self-reported information using questionnaires for Norse and Swedish subjects. Several potential
sources of exposure misclassification for these data should be noted, however, such as would be
introduced by changing employers, starting dry-cleaning work at a later time period, employment
1 Lynge et al. (20061. an analysis based only on the employment periods from 1965 through 1978, gave the
following RRs: 0-1 year = 1.43 (95% CI, 0.52-3.97); 2-4 years = 2.38 (95% CI, 1.08-5.24); 5-9 years = 1.21
(95% CI, 0.58-2.50); >10 years = 2.84 (95% CI, 0.97-8.35); unknown = 2.12 (95% CI, 0.65-6.85).
2 Finnish pension records started in 1962 for dry cleaning employees and in 1970 for self-employed persons.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-96 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
during a time period outside the examined range or before recordkeeping began, or imperfect
recall by proxy respondents on questionnaires. Moreover, exposure duration examined in all of
these studies is a poorer surrogate than a semiquantitative or quantitative exposure metric
because it does not account for potential temporal decreases in tetrachloroethylene intensity
resulting from improved tetrachloroethylene recovery and technological changes (Gold et al..
2008) or for variation in tetrachloroethylene concentration across shops (Lynge et al.. 2006). A
fourth study that examined exposure duration and, also, time since first employment observed
statistically significant associations with both increasing time since first employment and with
increasing duration of exposure (Calvert et al.. In Press).
Known risk factors for bladder cancer include smoking, aromatic amine dyes, chronic
inflammation, infection with the parasite Schistosoma heamatobium, and pelvic irradiation
(Kaufman et al.. 2009). Of these identified risk factors, potential confounding related to smoking
is most important to consider in the evaluation of bladder cancer and tetrachloroethylene in
studies of occupational and residential exposures, as exposure to other known risk factors is
much less common. Statistical control for smoking effects was used in all case-control studies,
including those informing the hazard identification analysis and those contributing lesser weight
(Aschengrau et al.. 1993; Burns and Swanson. 1991; Colt et al.. 2011; Dryson et al.. 2008;
Gaertner et al.. 2004; 2000b; Reulen et al.. 2007; Schoenberg et al.. 1984a; Siemiatvcki. 1991;
Silverman et al.. 1990; Silverman et al.. 1989a; Silverman et al.. 1989b; Smith et al.. 1985;
Steineck et al.. 1990; Teschke et al.. 1997; Vieira et al.. 2005b; Zheng et al.. 2002)(Kogevinas et
al., 2002). Lynge et al. (2006). a case-control study with subjects from four Nordic countries,
presented smoking-adjusted and unadjusted effect measures for subjects from two countries for
which smoking histories were obtained through interviews. Adjustment made little difference
(<10%) in the magnitude of the effect measure, indicating that smoking history is not a strong
confounder of the observed risk estimates [smoking unadjusted, 1.34, 95% CI: 0.86, 2.08;
smoking adjusted, 1.25, 95% CI: 0.79, 1.98 (Lynge et al.. 2006)1.
Direct examination of possible confounders is less common in cohort studies relying on
company-supplied or census work history data compared to case-control studies where
information is obtained from study subjects or their proxies. In cohort studies, however, use of
internal controls rather than an external referent group (e.g., national mortality rates) can
minimize effects of potential confounding due to smoking or socioeconomic status, because
exposed and referent subjects are drawn from the same target population. However, only one of
the available cohort studies included an analysis using internal controls, and that study is limited
by the observation of only two bladder cancer cases in the cohort (Boice et al.. 1999). Effect of
smoking as a possible confounder may be assessed indirectly through examination of risk ratios
for other smoking-related sites such as lung cancer. Several studies observed roughly a 30%
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-97 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
increase in lung cancer risk among dry cleaners (Blair et al.. 2003; Calvert et al.. In Press; Ji et
al.. 2005a; Pukkala et al.. 2009)(Ji et al., 2000a) employed a method that assumed smoking
accounted for 35% of their lung cancer observations and adjusted the bladder cancer
standardized incidence ratio by this proportion. This method reduced slightly the effect measure
for dry-cleaner and laundry workers (smoking unadjusted, 1.27, 95% CI: 1.08, 1.48; smoking
adjusted, 1.13, 95% CI: 0.96, 1.31) (Ji and Hemminki. 2005a). Blair et al. (2003) addressed
potential confounding by smoking and noted that if the magnitude of the difference in smoking
for dry cleaners compared with the general population is in the range of 10% of less,
confounding from smoking in their study of dry-cleaners and laundry workers was unlikely to
result in increased excess of over >20%. In the case of bladder cancer in this study, smoking
may explain the excess risk reported for overall exposure (SMR: 1.3). In contrast, the meta-
analysis of Reulen et al. (2008) examined studies that did or did not adjust for smoking and
found a stronger effect estimate with the smoking adjustment: the bladder cancer metarelative
risk estimates for launderers and bladder cancer were 1.72 (95% CI: 1.25, 2.37) in studies that
adjusted for smoking and 0.86 (95% CI: 0.59, 1.26) in studies that did not adjust for smoking. In
conclusion, while smoking may potentially confound, to a small degree, observations in some
cohort studies controlling for its effect in statistical analyses (Aschengrau et al.. 1993; Gaertner
et al.. 2004; Lynge et al.. 2006; Pesch et al.. 2000b; Reulen et al.. 2007; Siemiatvcki. 1991;
Silverman et al.. 1989a; Silverman et al.. 1989b; Smith et al.. 1985; Teschke et al.. 1997; Zheng
et al.. 2002)(Kogevinas et al., 2002), these studies do provide evidence of an association with
tetrachloroethylene or with holding a job as a dry cleaner or a laundry worker, a surrogate for
tetrachloroethylene exposure potential.
In conclusion, the pattern of results from this collection of studies is consistent with an
elevated risk for tetrachloroethylene of a relatively modest magnitude. The results from four of
the five studies with the relatively high quality exposure-assessment methodologies provide
evidence of an association, with relative risks of 1.44 to 4.03 (Blair et al.. 2003; Calvert et al.. In
Press; Lynge et al.. 2006; Pesch et al.. 2000b; substantial exposure. JTEM approach). The Lynge
et al. (2006) results were slightly higher among the subgroup from Denmark and Norway, in
which the number of unclassifiable data was negligible (relative risk 1.69, 95% CI: 1.18, 2.43).
An exposure-response gradient was seen in a large case-control study by Pesch et al. (2000b)
using a semiquantitative cumulative exposure assessment, but not in Lynge et al. (2006) using
employment duration without consideration of exposure concentration. In addition, relative risk
estimates between bladder cancer risk and ever having a job title of dry-cleaner or laundry
worker in four large cohort studies ranged from 1.01 to 1.44 (Ji and Hemminki. 2005a; Pukkala
et al.. 2009; Travier et al.. 2002; Wilson et al.. 2008). Confounding by smoking is an unlikely
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-98 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
explanation for the findings, given the adjustment for smoking by Pesch et al. (2000b) and other
case-control studies.
4.2.2. Animal Studies
Kidney toxicity and cancer has been observed in laboratory animals exposed to
tetrachloroethylene in multiple studies. The sections below describe studies of kidney toxicity
(see Section 4.2.2.1) and cancer (see Section 4.2.2.2). These studies are summarized in
Tables 4-9 and 4-10, respectively.
4.2.2.1. Kidney Toxicity in Animals
Tetrachloroethylene causes renal toxicity across multiple species, including several
strains of rats and mice (for reviews, see ATSDR. 1997b; Cal/EPA. 2001; NYSDOH. 1997; U.S.
EPA. 1985b). Adverse effects on the kidney have been observed in studies of animals exposed
to high concentrations of tetrachloroethylene by inhalation, oral intake, and i.p. injection. These
effects increased kidney-to-body weight ratios, hyaline droplet formation, cast formation,
glomerular —nephron," karyomegaly (enlarged nuclei), and other lesions or indicators of renal
toxicity. These nephrotoxic effects mainly occurred following relatively high subchronic
(400-800 ppm) or chronic tetrachloroethylene exposures (100-200 ppm).
4.2.2.1.1. Inhalation
A long-term inhalation study examined the effects of tetrachloroethylene exposure in
male and female rats by observation throughout the lifetime of the animals (0, 300, 600 ppm,
6 hours/day, 5 days/week, for 12 months) (Rampy et al.. 1978). No increase in tumors compared
to controls was observed in any animals in this study; however, an increase in mortality related
to renal failure was observed in male rats starting at 5 months exposure in the high-dose group.
No effects were observed in hematologic parameters measured (hemoglobin concentration, WBC
counts) or various urinalysis endpoints (specific gravity, pH, presence of ketones, bilirubin, or
blood, or sugar and albumin concentrations). The authors state that clinical chemistry
measurements are not useful because most animals were deceased or moribund at the end of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-99 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-9. Summary of rodent kidney toxicity studies
Species/strain/
sex/number
Exposure level/duration
Effects
Reference
Mouse, B6C3F1; both
sexes (49 or 50 of each
sex per dose group, total
of -300 mice)
0, 100, 200 ppmfor
104 wk, inhalation
Karyomegaly and cytomegaly of the
proximal tubules in all exposed mice;
nephrosis was observed in exposed females,
casts increased in all exposed males and in
high-dose females
NTP (1986b)
Rat, F344, both sexes
(50 of each sex per dose
group, total of -300
mice)
0, 200, 400 ppm for
104 wk, inhalation
Karyomegaly and cytomegaly of the
proximal tubules in all exposed rats
NTP (1986b)
Mouse, Crj/BDFl (both
sexes, 50 of each sex per
dose group, total of
400 mice)
0, 10, 50, 250 ppmfor
110 wk, inhalation
Increased relative kidney weights and
karyomegaly in the proximal tubules in 250-
ppm exposed male and female mice; atypical
tubular dilation in 250-ppm male and female
mice but was not statistically significant
JISA (1993)
Rat, F344/DuCij (both
sexes, 50 of each sex per
dose group, total of
400 rats)
0, 50, 200, 600 ppm for
110 wk, inhalation
Increased relative kidney weights and
karyomegaly in the proximal tubules in 200
and 600-ppm exposed male and female rats;
atypical tubular dilation in 600-ppm male
and female rats; exacerbation of chronic
renal disease in male rats only at 600 ppm
JISA, (1993)
Osborne-Mendel rats,
both sexes, 50 of each
sex per dose group;
B6C3Fi mice, both
sexes, 50 of each sex per
dose group
0, 475, 950 mg/kg-day
(rats); 0, 536,
1,072 mg/kg-day (male
mice); 0, 386,
772 mg/kg-day (female
mice) by oral gavage in
corn oil for 78 wk,
observed for 32 wk (rats) or
12 wk (mice) following
exposure
Toxic nephropathy observed in all exposed
animal groups, with an increased incidence
in rats as compared to mice
NCI, (1977)
Sprague-Dawley rats
(both sexes); 96 per sex
per exposure group;
192 per sex per control
group
0, 300, 600 ppm for 6 h/d,
5 d/wk for 12 mo; observed
for the lifetime of the rat
(up to 31 mo total)
Increased mortality related to renal failure in
male rats exposed to 600 ppm starting at
5 mo of exposure
Rampy et al.
(1978)
Rat, F344; and mouse,
B6C3Fi; both sexes
(5 of each sex per group)
0, 200 (28 d only), and
400 ppm (14, 21,28 d) for
6 h/d, inhalation
Analysis in mice was limited to pooled
tissue but showed slight increases in
(3-oxidation in mouse kidney; modest
increases in PCO observed in male rat
kidneys at 200 ppm for 28 d only, but
elevated in female rat kidneys at all doses
and times
Odum et
al.,( 1988b)
Mouse, Swiss-Webster,
male (4/group)
0, 150, 500, and
1,000 mg/kg-day, aqueous
gavage for 30 d
No kidney injury or dysfunction was
observed in this study
Philip et al.
(2007)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-100 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-9. Summary of rodent kidney toxicity studies (continued)
Species/strain/
sex/number
Exposure level/duration
Effects
Reference
Rat, Wistar, female only
(10 rats in each control
group; 5 rats in each
treatment group)
0, 600, and 2,400
mg/kg-day for 32 d, corn
oil gavage; alone or in
combination with other
compounds
(trichloroethylene,
hexachloro-l,2-butadiene,
1,1,2-trichloro-
3,3,3 -trifluoropropene)
Relative kidney weight was increased on
exposure to PCE alone and in combination
with other nephrotoxicants; nephrotoxic
effects noted at high dose (urea, total protein,
albumin, NAG); karyomegaly was also
observed in high dose animals
Jonker et al.
(1996)
F344 rats (male only,
5/group) and B6C3Fi
mice (male only,
5/group)
0 or 1,000 mg/kg-day for
10 d, corn oil gavage
Increased kidney weight in exposed rats;
increased PCO activity in all exposed mice
Goldsworthy
and Popp
(1987)
F344 rats (both sexes)
0 or 1,000 mg/kg-day for
10 d, corn oil gavage
Increases in a2(i-hyaline droplets in exposed
male, but not female, rats, correlated with
increased cell proliferation and protein
droplet nephropathy
Goldsworthy
et al. (1988)
F344 rats (both sexes, 12
per group)
0, 500 mg/kg-day daily for
4 wk, corn oil gavage
Increases in a2(i-hyaline accumulation in
proximal tubule cells
Bergamaschi
et al. (1992)
Mouse, Swiss, both
sexes; 6 groups of 6
each (1996); male onlv;
8 groups of 6 each
(2001)
0 or 3,000 mg/kg-day for
15 d, sesame oil gavage
Significant increase in kidney weight;
decreased blood glucose (glucose effects
mitigated by coexposures to 2-deoxy-
D-glucose and vitamin E [1996])
Decreased membrane-bound
Na+K+-ATPases and Mg2+-ATPases activity
but increased Ca-ATPase activity; mitigated
by coexposure to 2-deoxy-D-glucose and
vitamin E, and taurine; hypercellular
glomeruli in PCE-exposed only
Ebrahim et al.
(1996; 2001)
F344 rats (both sexes)
and B6C3Fi mice (both
sexes); 10 per group for
oral studies, 5 per group
for inhalation studies
0, 1,000, or 1,500
mg/kg-day daily by corn oil
gavage for 42 d; 0 or
1,000 ppmfor 10 d
Accumulation of a2(i-globulin in proximal
tubules of male rats; nephrotoxicity also
observed in male rats (formation of granular
tubular casts and evidence of tubular cell
regeneration)
Inhalation exposure demonstrated formation
of hyaline droplets in kidneys of male rats
Green et al.
(1990)
Sprague-Dawley rats
(both sexes, 20 per
group)
0, 14, 400, or 1,400
mg/kg-day for 90 d
Increased kidney weight observed in
exposed animals; nephrotoxicity observed at
400 mg/kg-day
Hayes et al.
(1986)
Sprague-Dawley rats
(male only; 4 per group)
0, 115, 230 junol/kg of
TCVC or TCVCS bw in
saline by one i.p. injection,
sacrificed 24 h
postexposure
High-dose exposed animals showed visible
kidney necrosis; all other rats showed
histological markers for mild acute tubular
necrosis (TCVC) or severe acute tubular
necrosis (TCVCS); prior exposure to AOAA
increased toxicity
Elfarra et al.
(2007)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-101 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-10. Kidney tumor incidence in laboratory animals exposed to
tetrachloroethylene
Bioassay
Doses/exposures
Sex
Tumor incidence (%)
Administered
Continuous equivalent
Kidney adenomas and carcinomas
NCI (1977)a
B6C3Fi mice
Gavage:
5 d/wk,
78 wk
Vehicle control
450 mg/kg-day
900 mg/kg-day
0
332 mg/kg-day
663 mg/kg-day
Male
0/20 (0)
1/49 (2)
0/48 (0)
Vehicle control
300 mg/kg-day
600 mg/kg-day
0
239 mg/kg-day
478 mg/kg-day
Female
0/20 (0)
0/48 (0)
0/45(0)
NCI (1977)a
Osborn-Mendel rats
Gavage:
5 d/wk,
78 wk
Vehicle control
500 mg/kg-day
1,000 mg/kg-day
0
471 mg/kg-day
941 mg/kg-day
Male
3/20 (5)
1/49 (2)
0/50 (0)
Vehicle control
500 mg/kg-day
1,000 mg/kg-day
0
474 mg/kg-day
974 mg/kg-day
Female
0/20 (0)
0/50 (0)
1/50 (2)
NTP (1986b)
B6C3Fi mice
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
100 ppm
200 ppm
0
18 ppm
36 ppm
Male
0/49 (0)
1/49 (2)
0/50 (0)
0 ppm
100 ppm
200 ppm
0
18 ppm
36 ppm
Female
0/48 (0)
0/50 (0)
0/48 (0)
NTP (1986b)
F344/N rats
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
200 ppm
400 ppm
0
36 ppm
72 ppm
Male
1/49 (2)
3/47 (6)
4/50 (8)
0 ppm
200 ppm
400 ppm
0
36 ppm
72 ppm
Female
0/50 (0)
0/50 (0)
0/50 (0)
JISA (1993)
Cij:BDFl mice
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
10 ppm
50 ppm
250 ppm
0
1.8 ppm
9.0 ppm
45 ppm
Male
0/50 (0)
1/50 (2)
1/50 (2)
0/50 (0)
0 ppm
10 ppm
50 ppm
250 ppm
0
1.8 ppm
9.0 ppm
45 ppm
Female
0/50 (0)
0/47 (0)
0/49 (0)
0/50 (0)
JISA (1993)
F344/DuCij rats
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
50 ppm
200 ppm
600 ppm
0
9 ppm
36 ppm
108 ppm
Male
1/50 (2)
2/50 (4)
1/50 (2)
2/50 (4)
0 ppm
50 ppm
200 ppm
600 ppm
0
9 ppm
36 ppm
108 ppm
Female
1/50 (2)
0/50 (0)
0/50 (0)
1/50 (2)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-102 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
study, and the study details show only measurements in a limited number of animals (1 male per
group, 5 females per group). Although the authors conclude limited tetrachloroethylene toxicity,
due to the large amount of morbidity in this study, it is difficult to make any conclusions as to the
toxicity and/or carcinogenicity of tetrachloroethylene from this study.
Acute, subchronic, and chronic exposures to tetrachloroethylene were examined in male
and female F344 rats and B6C3Fi mice (NTP. 1986b). Single exposure studies and 14-day
studies were performed, but no kidney effects were observed, with the first kidney effects
observed in the subchronic (13 week) study. Groups of 10 rats and mice of each sex were
exposed to air containing tetrachloroethylene for 6 hours/day, 5 days/week, for 13 weeks (0, 100,
200, 400, 800, or 1,600 ppm). Some rats in the high-dose group died before the end of the
studies (4/10 male, 7/10 female), but no kidney effects were observed. In mice, 2/10 males and
4/10 females in the high-dose group died before the end of the studies, and karyomegaly (nuclear
enlargement) of the renal tubule epithelial cells was observed in all but the lowest dose group.
Toxicity was observed in a 2-year cancer bioassay performed on groups of 50 F344 rats
of each sex (0-, 200-, or 400-ppm tetrachloroethylene), or groups of 49 or 50 mice (0-, 100-, or
200-ppm tetrachloroethylene) exposed for 6 hours/day 5 days/week for 103 weeks (NTP. 1986b).
Karyomegaly and cytomegaly changes were observed in both sexes of rats at all doses but not in
unexposed controls. These lesions were present primarily in the proximal convoluted tubules of
the inner half of the cortex but not limited to this area. In mice, nephrosis (generally defined as
noninflammatory degenerative disease of the kidney) was observed in exposed females, casts
(cylindrical structures formed from cells and protein released from the kidney) were increased in
exposed male and high-dose females, and karyomegaly of the tubular cells was observed in all
dosed mice, with severity of lesions being dose related. Therefore, the LOAEL for renal toxicity
reported in both mice and rats in this study is 100 ppm (678 mg/m3) for inhalation exposure in
mice and 200 ppm (1,356 mg/m3) in rats (NTP. 1986b).
Nephrotoxicity was observed in a second, 2-year inhalation cancer bioassay also
performed in 50 male and female Fischer rats (0, 50, 200, or 600 ppm) and Crj:BDFl mice (0,
10, 50, or 250 ppm) in each treatment group (6 hours/day, 5 days/week, for 104 weeks) (JISA.
1993). Survival compared to controls was decreased in all high-dose exposure groups, which
was believed to be treatment related. Relative kidney weight was increased in male and female
rats exposed to tetrachloroethylene (200 or 600 ppm) and in male and female mice (250 ppm).
Karyomegaly in the proximal tubules of the kidneys was observed among males and females
(200 and 600 ppm in male rats [23/50 and 48/50]; 600 ppm in female rats [18/50]; 50 and 250
ppm in male mice [6/50 and 38/50]; 250 ppm in female mice [49/50]), and an increase in atypical
tubular dilation of the proximal tubules [male and female rats, 600 ppm (24/50 males, 6/50
females) and exacerbation of chronic renal disease in male rats only (600 ppm) was observed
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-103 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
with tetrachloroethylene exposure (JISA. 1993). Atypical tubular dilation was also observed in
mice but was not statistically significant (250 ppm in male mice [1/50] and female mice [6/50]).
The role of peroxisome proliferation in tetrachloroethylene-induced kidney toxicity and
cancer was examined in male and female F344 rats and B6C3Fi mice exposed to
tetrachloroethylene by inhalation (400 ppm, 6 hours/day, for 14, 21, or 28 days, or 200 ppm,
6 hours/day, for 28 days) in a study by Odum et al. (1988b). Five animals per group were
exposed. Insufficient mouse kidney tissue limited the analysis to pooled samples. Slight
increases were observed in P-oxidation in mouse kidney (maximum 1.6-fold increase at 21 days,
400-ppm exposure). Modest palmitoyl-CoA oxidation (PCO) increases were seen in the kidney
of male rats at 200 ppm at 28 days (1.3-fold) but not 400 ppm at 14, 21, or 28 days. In female rat
kidney, PCO was elevated (approximately 1.6-fold) at all doses and times. However,
peroxisome proliferation was not seen in rat or mouse kidney upon microscopy, suggesting that
this does not play a role in kidney carcinogenesis. Short-term inhalation exposure to 1,000-ppm
tetrachloroethylene for 10 days resulted in the formation of hyaline droplets in the kidneys of
male rats. Although granular casts and tubule cell regeneration were not observed, the time
period may have been too short to allow progression to this stage (Green et al.. 1990).
4.2.2.1.2. Oral
Hayes et al. (1986) reported renal effects in rats exposed to 400-mg/kg-day
tetrachloroethylene in drinking water for 90 days. Tetrachloroethylene was administered in the
drinking water at 14, 400, and 1,400 mg/kg bw per day for 90 days, with no deaths reported
before the end of the study. Increased kidney weight was observed.
A lifetime animal carcinogenicity study in which tetrachloroethylene was administered to
50 of each sex of Osborne-Mendel rats and B6C3Fi mice by oral gavage in corn oil for 78 weeks
resulted in clear evidence of kidney toxicity in both species (NCI. 1977). The TWA doses
(mg/kg-day) used in the bioassay were 471 and 941 for male rats, 474 and 949 for female rats,
536 and 1,072 for male mice, and 386 and 772 for female mice. Animals were observed for 32
weeks (rats) or 12 weeks (mice) following the last dose. Toxic nephropathy was observed in
almost all test animals, with a high incidence observed in treated rats, including those that died
early in the study (as early as Week 20 in male rats, Week 28 in female rats). Similar results
were observed in exposed mice, with no nephropathy observed in control mice. Therefore, the
LOAEL for renal toxicity following oral exposure is 471 mg/kg-day in male rats and
474 mg/kg-day in female rats based on toxic nephropathy. The LOAEL for mice is
536 mg/kg-day for males and 386 mg/kg-day in females based on toxic nephropathy.
In a study by Jonker et al. (1996). tetrachloroethylene nephrotoxicity was observed in
female Wistar rats administered tetrachloroethylene (600 or 2,400 mg/kg-day) in corn oil by
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-104 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
daily oral gavage for 32 days. Relative kidney weight was increased upon exposure to
tetrachloroethylene alone and in combination with other nephrotoxicants (trichloroethylene
[TCE], hexachloro-l,2-butadiene, and l,l,2-trichloro-3,3,3-trhifluoropropene [TCTFP]). One
high-dose animal died as a result of tetrachloroethylene treatment, and one animal exposed to the
high-dose combination of TCE, tetrachloroethylene, and TCTFP also died as a result of
treatment. Nephrotoxic effects were noted at 2,400 mg/kg. Significant changes were observed
following exposure to tetrachloroethylene at 2,400 mg/kg-day in all clinical chemistry markers
related to kidney function (urea, total protein, albumin, NAG) as measured in the urine at the end
of Week 1 or Week 4 except for urinary density, glucose, and creatinine. Karyomegaly was also
observed at the high dose (2,400 mg/kg-day) in four of five animals exposed (p < 0.01) (Jonker
etal.. 1996V
Philip et al. (2007) exposed male 6-7 week old Swiss Webster mice via aqueous gavage
to three dose levels (150, 500, and 1,000 mg/kg-day) for 30 days. At the highest exposure,
mortality was 10% due to apparent CNS toxicity (tremors and ataxia). Neither kidney injury nor
dysfunction was observed following tetrachloroethylene exposure during the course of this study.
Goldsworthy and Popp (1987) administered tetrachloroethylene (1,000 mg/kg-day) by
corn oil gavage to 5 male F344 rats and 5 male B6C3Fi mice for 10 days. In
tetrachloroethylene-exposed rats, PCO was modestly although not significantly elevated in the
liver (1.4-fold increase) and kidney (1.7-fold increase). In mice, tetrachloroethylene exposure
increased PCO activity 4.3-fold in liver and by 2.3-fold in kidney. Relative liver weight was
increased in rats and mice with tetrachloroethylene exposure, but relative kidney weight was
unaffected. A comparison of corn oil with methyl cellulose revealed no effect of the gavage
vehicle on tetrachloroethylene-induced PCO. A less-than-additive effect of trichloroethylene
(1,000 mg/kg) administered together with tetrachloroethylene on PCO induction was seen.
Oral administration of tetrachloroethylene in sesame oil (3,000 mg/kg-day for 15 days) to
male and female albino Swiss mice caused a significant increase in kidney weight (p < 0.001)
and decreased blood glucose levels (p < 0.01) as compared to control animals exposed to sesame
oil alone as well as increases in glomerular nephrosis (Ebrahim et al.. 1996). This study was
designed to give support to the beneficial effect of 2-deoxy-D-glucose (2DG) and vitamin E on
tetrachloroethylene-induced kidney damage. Based on previous experimental mouse tumor
studies, administration of 2DG or vitamin E is hypothesized to have a beneficial effect on
tetrachloroethylene-induced kidney damage, either by inhibition of tumor growth (2DG) or the
auto-catalytic process of lipid peroxidation (vitamin E). In this study, concurrent administration
of 2DG (500 mg/kg-day i.p.) or vitamin E (400 mg/kg-day oral gavage) prevented
tetrachloroethylene-induced biochemical and pathological alterations. Tetrachloroethylene
exposure alone led to a decrease in blood glucose levels, which was returned to near normal with
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-105 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
concomitant exposure to 2DG and vitamin E. Elevated levels of glycolytic and gluconeogenic
enzymes following exposure to tetrachloroethylene were also observed to return to near normal
with exposure to 2DG and vitamin E. Histopathology of the kidney showed hypercellular
glomeruli following exposure to tetrachloroethylene, but this was not observed in animals treated
with tetrachloroethylene and 2DG, or tetrachloroethylene and vitamin E (Ebrahim et al.. 1996).
A follow-up study by this group further examined the potential protective properties of 2DG and
vitamin E as well as taurine against tetrachloroethylene-induced membrane damage (Ebrahim et
al.. 2001). This study exposed male albino Swiss mice to the same doses used in the previous
study with the addition of a taurine-exposed group (tetrachloroethylene in sesame oil
3,000 mg/kg-day for 15 days orally by intubation; tetrachloroethylene plus 2DG 500 mg/kg-day
by i.p. injection once a day for 15 days; tetrachloroethylene plus vitamin E 400 mg/kg-day by
oral intubation once a day for 15 days; and tetrachloroethylene plus taurine 100 mg/kg-day by
oral intubation once a day for 15 days). As compared to control cells in the kidney, membrane-
bound Na+K+-ATPases and Mg2+-ATPases activity was significantly decreased (p < 0.001),
while Ca-ATPases activity was increased (p < 0.001), following exposure to tetrachloroethylene
alone. These levels remained near normal in the animals exposed to tetrachloroethylene along
with 2DG, vitamin E, or taurine. This return to normal levels following exposure to vitamin E
and taurine may be due to their antioxidant abilities, and reduced oxidative stress in exposed
cells.
Goldsworthy et al. (1988) observed increases in a2[j.-hyaline droplets in exposed male but
not female F344 rats following 10 days of gavage with 1,000-mg/kg tetrachloroethylene. This
finding was correlated with both protein droplet nephropathy (crystalloid accumulation) and
increases in cellular proliferation. Cell replication was enhanced in the male rats specifically in
damaged P2 segments, suggesting a link between the a2[j,-globulin accumulation and kidney
tumors. These investigators reported similar findings for pentachloroethane in the same study,
but at a dose of 150 mg/kg for 10 days. Trichloroethylene has a similar structure but did not
cause any a2[j,-accumulation or increase in protein droplets, nor did it stimulate cellular
proliferation in either male or female rats in this study when a dose of 1,000 mg/kg was
administered for 10 days. Bergamaschi et al. (1992) also demonstrated a2[j,-accumulation in P2
segments of rat proximal tubule cells resulting from a daily exposure of rats to 500 mg/kg
tetrachloroethylene in corn oil for 4 weeks.
In short-term, high-dose studies, Green et al. (1990) found that the oral administration of
1,000 to 1,500 mg/kg of tetrachloroethylene daily for up to 42 days caused an accumulation of
a2[j,-globulin in the proximal tubules of male rats. The animals were sacrificed within 24 hours
of the last dose of tetrachloroethylene. The effect was accompanied by evidence of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-106 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
nephrotoxicity, with the formation of granular tubular casts and evidence of tubular cell
regeneration. These effects were not observed in female rats or in mice.
4.2.2.1.3. Intraperitoneal injection
The role of the glutathione metabolites, particularly TCVC and TCVCS, in kidney
toxicity was examined by Elfarra et al. (2007) in vivo. This study exposed two groups of four
male Sprague-Dawley rats to a single i.p. injection of TCVC or TCVCS (115 or 230 [j,mol/kg bw
in saline). Animals were sacrificed 24 hours following exposure. Serum was analyzed for BUN,
and urine samples were analyzed for GGTP activity as markers of nephrotoxicity. Rats exposed
to the high-dose of TCVCS showed visible signs of kidney necrosis, while all other exposed
groups did not. Histologically, kidneys from rats exposed to low-dose TCVC or TCVCS showed
slight-to-mild acute tubular necrosis. Analysis of kidneys at 24 hours postexposure showed
mild-to-moderate acute tubular necrosis in animals exposed to high-dose (230 (j,mol/kg) TCVC,
and severe tubular necrosis in animals exposed to high-dose (230 (j.mol/kg) TCVCS. Similar to
the pattern of toxicity described above, significant increases in BUN (fourfold) were observed in
rats exposed to 230-[j,mol/kg TCVCS as compared to control, but no significant increases were
observed following exposure to TCVC. Variable increases were observed following exposure to
TCVC or TCVCS in urine glucose levels and GGTP activity. A second part of this experiment
involved a preexposure to a P-lyase inhibitor (AOAA) (500 [j,mol/kg bw) by i.p. injection
30 minutes prior to administration of 230-[j,mol/kg TCVC. Exposure to AOAA prior to exposure
to TCVCS resulted in increased toxicity. In a third study, three groups of four rats were exposed
to saline, TCVC, or TCVCS (230 [j,mol/kg bw) and sacrificed 2 hours after administration. The
kidneys were removed at sacrifice and examined for NPT and NPT disulfide concentrations as a
measure of thiol status in the kidney. Although no changes were observed in NPT status,
histological examination of these kidneys showed scattered foci of mild acute tubular necrosis
(TCVC) or widespread acute tubular necrosis, intratubular casts, and interstitial congestion and
hemorrhage (TCVCS). These results suggest that while both TCVC and TCVCS are
nephrotoxicants, TCVCS is more potent than TCVC.
In summary, exposure to tetrachloroethylene from all routes studied (oral, inhalation, i.p.)
led to nephrotoxicity in multiple strains of rats and mice. These studies demonstrate
karyomegaly, increased kidney weights, and atypical tubular dilation following subchronic high-
dose exposures or lower dose chronic exposures. Limited studies have also examined the
potential role for peroxisome proliferation or a2[j,-globulin in nephrotoxicity. Exposure to
tetrachloroethylene glutathione conjugation metabolites led to similar effects in rats (mice not
tested). Further studies examining the impact of concomitant antioxidant exposures with
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-107 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
tetrachloroethylene in mice suggests a role for oxidative stress in tetrachloroethylene-induced
nephrotoxicity.
4.2.2.2. Kidney Cancer in Animals
4.2.2.2.1. Inhalation
In the studies conducted by NTP (1986b. described above), groups of 50 male and
50 female F344/N rats were exposed for 6 hours/day, 5 days/week, for 103 weeks by inhalation
to atmospheres containing 0-, 200-, or 400-ppm tetrachloroethylene. Tubule cell hyperplasia
was observed in male rats (control, 0/49; low dose, 3/49; high dose, 5/50) and in one high-dose
female rat. Renal tubule adenomas and adenocarcinomas were observed in male rats (control,
1/49; low dose, 3/49; high dose, 4/50). In the same study (doses described above), one renal
tubule adenocarcinoma was observed in a low-dose male mouse, but no other neoplastic lesions
were observed.
The spontaneous incidence rate for renal tubule tumors in F344/N rats, the strain used in
the NTP bioassay, as well as for other rat strains reported by NTP, was less than 1%. Thus, the
appearance of tubule neoplasms in 8% of the treated animals in the NTP study (low-dose and
high-dose groups combined) provided convincing evidence of a treatment-related effect
(Goodman et al.. 1979; Solleveld et al.. 1984; U.S. EPA. 1986a. 1991b). Also notable is the fact
that no malignant renal tubule neoplasms had ever been observed in any control rats examined
by NTP—including chamber controls from the performing laboratory and the untreated controls
and vehicle controls from gavage studies—whereas two of the tumors observed in high-dose
animals in the NTP study were carcinomas. The probability of two rare carcinomas appearing by
chance in a group of 50 animals has been calculated to be less than 0.001 (NTP. 1986b; U.S.
EPA, 1987, 1991b).
In addition, when compared with historical control incidences of renal tubule tumors at
the NTP, a statistically significant dose-related positive trend exists, and tumor incidences in
both low-dose and high-dose groups are significantly elevated. Standard statistical analyses of
tumor incidence data did not reveal a significant increase in kidney tumors, and the tumor
incidence is not statistically significant when compared with concurrent controls; however, when
the incidences of tubule cell hyperplasia and neoplasms and tumor severity are all considered, a
dose-response relationship is apparent.
No increase in renal cell cancers was observed in a second 2-year inhalation cancer
bioassay was also performed in 50 male and female Fischer rats (0, 50, 200, or 600 ppm) and
Crj:BDFl mice (0, 10, 50, or 250 ppm) in each treatment group (6 hours/day, 5 day/week, for
104 weeks) (JISA. 1993). Survival compared to controls was decreased in all high-dose
exposure groups, which is believed to be treatment related. Renal cell adenoma was observed in
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-108 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
male rats (1/50, control; 2/50, 50 ppm; 1/50, 200 ppm; 2/50, 600 ppm) and male mice (1/50, 50
ppm) but only in control female rats (1/50, control) and not in exposed female mice. Renal cell
carcinoma was not observed in male rats or female mice, but was observed in the high-dose
female rats (1/50, 600 ppm) and male mice (1/50, 50 ppm). As described above for the NTP
study (1986b), these tumors are rare in Fischer rats, but the reported results are similar to those
historical control rates for this study group CJISA, 1993).
The study authors reported a slight increase in renal tumors with tetrachloroethylene
exposure in a study reporting increased mortality related to renal failure in male rats starting at
5 months exposure in the high-dose group (Rampy et al.. 1978). This lifetime observation study
exposed male and female rats to 0, 300, 600 ppm, 6 hours/day, 5 days/week, for 12 months
(Rampy et al.. 1978). The authors stated that most animals were deceased or moribund at the
end of study, rendering difficult, clear conclusions regarding the renal carcinogenicity of
tetrachl oroethy lene.
4.2.2.2.2. Oral
No significant increased incidence of neoplastic lesions was observed in treated rats
following oral exposure to tetrachloroethylene in a lifetime carcinogenicity bioassay (NCI. 1977;
doses described above). However, a high rate of death occurred in the high-dose groups of both
sexes, so the authors of the study determined carcinogenicity could not be evaluated. Only one
kidney tumor was observed in mice in this study (high dose; doses described above), but this was
a tumor that had metastasized from the liver.
In summary, an increase in rare kidney tumors was reported in one inhalation cancer
bioassay of tetrachloroethylene (0, 200, or 400 ppm) in F344/N rats (NTP. 1986b). The JISA
(1993) rat inhalation bioassay of tetrachloroethylene (50, 200, and 600 ppm) reported no
treatment-related increase in the incidence of kidney tubular cell adenoma or carcinoma in
excess of that in the concurrent or historical control animals at administered concentrations.
Another inhalation study, the interpretation of which is limited by high morbidity and mortality,
reported a slight increase in renal tumors in male S-D rats (Rampy et al.. 1978). Although the
renal tumors were not significantly increased compared with controls, morbidity related to renal
failure was increased in male rats beginning at 5 months of exposure. The NCI (1977) oral
gavage bioassay of tetrachloroethylene (0, 475, 950 mg/kg-day) reported a high rate of death in
the high-dose groups of both sexes, and, thus, carcinogenicity could not be evaluated in this
study.
Other evidence supporting the conclusion of renal carcinogenicity of tetrachloroethylene
includes low incidences of tubule neoplasms in male rats in NTP bioassays of other chlorinated
ethanes and ethylenes (NTP. 1983. 1986a. b, 1987. 1988. 1989. 1990a). In particular, the closely
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-109 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
related compound trichloroethylene also induces low increases in the incidence of rare renal
tumors in rats and in humans (U.S. EPA. 2009b).
4.2.2.2.3. In vitro
Lash et al. (1998) examined the role of glutathione conjugation of tetrachloroethylene
inrats and mice in isolated renal cortical cells and hepatocytes from male and female F344 rats.
All cells were exposed to tetrachloroethylene (0.5, 1, or 2 mM) and assayed for TCVG formation
at 0, 15, 30, and 60 minutes. This study demonstrated that GSH metabolites from
tetrachloroethylene are formed in kidney cells as well as hepatocytes in both species; however,
the amount of TCVG produced varied depending on sex, species, and tissue assayed. TCVG
formation was higher in male rats and mice as compared to their female counterparts and was
also higher in hepatocytes as compared to kidney cells. Although rats are more susceptible to
nephrocarcinogenicity as compared to mice (see Section 4.5.2.2), isolated mouse kidney and
liver cells had a greater amount of TCVG formation (7- to 10-fold and 2- to 5-fold, respectively)
as compared to rat cells (Lash et al.. 1998). To further examine the species- and sex-dependent-
differences in tetrachloroethylene cytotoxicity, Lash et al. (2002) measured acute cytotoxicity
following exposure to tetrachloroethylene or TCVG (0.1 to 10 mM) in isolated rat kidney cells
and renal mitochondria from rats and mice. Exposure to tetrachloroethylene or TCVG led to a
marked increase in LDH release in isolated kidney cells from male but not female rats, but no
significant effects were observed in rat hepatocytes from either gender (Lash et al.. 1998).
Isolated mitochondria from rats and mice showed a pattern of sensitivity similar to the kidney
cell effects, with increased inhibition of respiration in isolated mitochondria from male rats as
compared to their female counterparts. Inhibition of respiration was observed equally in male
and female mice exposed to tetrachloroethylene or TCVG. The results of this in vitro study
support those of the in vivo studies, which demonstrate increased nephrotoxicity in male rats
following exposure to tetrachloroethylene or TCVG.
Lash et al. (2007) examined the effect of modulation of renal metabolism on toxicity of
tetrachloroethylene in isolated cells and microsomes from male F344 rat kidney and liver.
Oxidative-dependent metabolism of tetrachloroethylene was more than 30-fold increased in liver
microsomes than in kidney. Pretreatment of rats with a P450-inhibitor had little to no effect on
the tetrachloroethylene metabolism in either kidney or liver. Pretreatment of rats with a P450
inducer increased tetrachloroethylene metabolism by over twofold in the kidney microsomes,
with no effect observed in liver. Following exposure to modulating chemicals, lactate
dehydrogenase (LDH) was measured as a marker of cytotoxicity, and the presence of specific
metabolites was documented (TCVG, TCOH, and CH). Tetrachloroethylene metabolism in
kidney cells was slightly (but significantly) increased by the nonspecific inhibitors of P450s but
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-110 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
not affected by the pretreatment with the CYP2E1-specific inhibitor. Increased cytotoxicity in
kidney cells was observed following exposure to tetrachloroethylene (2 or 10 mM, 3 hours), and
this was not affected by pretreatment with CYP inhibitors or inducers. However, increases in
GSH concentrations in the kidney cells led to increased cytotoxicity following exposure to
tetrachloroethylene, but no effect was observed following pretreatment with GSH inhibitors.
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.
Tetrachloroethylene effects in kidney cells have also been demonstrated in a variety of
genotoxicity assays. Exposing kidney cells and/or microsomal fractions from kidneys to
tetrachloroethylene or its some of its metabolites led to low levels of DNA binding (Mazzullo et
al.. 1987). micronuclei induction (Wang et al.. 2001). single-stranded DNA breaks (Walles.
1986). unscheduled DNA synthesis (Vamvakas et al.. 1989b). and gene mutations (Dekant et al..
1986d; Vamvakas et al.. 1987; Vamvakas et al.. 1989c). Negative studies were observed in
kidney cells from exposed animals for DNA damage (Cederberg et al.. 2010; Potter et al.. 1996).
and DNA adduct formation (Toraason et al.. 1999).
Limited DNA binding to calf thymus DNA was observed in the presence of microsomal
fractions from mice and rats (Mazzullo etal.. 1987). Binding to DNA in the in vitro study
increased in the presence of microsomal fractions from both mouse and rat liver, but not kidney,
lung, or stomach. Cytosolic fractions from rat and mouse liver, kidney, lung, and stomach, all
induced binding of tetrachloroethylene to calf thymus DNA, with enzymes from both mouse and
rat livers and mouse lung being the most efficient.
Wang et al. (2001) examined micronuclei induction following exposure to
tetrachloroethylene (-63 ppm in culture medium at peak) in vitro in a closed system. Chinese
hamster ovary (CHO-K1) cells were plated in a petri dish surrounding a glass dish of
tetrachloroethylene and incubated for 24 hours. Tetrachloroethylene exposure led to a dose-
dependent significant increase in micronuclei induction (p < 0.001) (Wang et al.. 2001).
Vamvakas et al. (1989a) reported concentration-related increases in unscheduled DNA
synthesis (UDS) in LLC-PK1 (a porcine kidney cell line) exposed to TCVC, with the effect
abolished by a P-lyase inhibitor. This effect was observed at exposure to 5 x 10 6—10 5 M
TCVC for 24 hours.
TCVG produced from tetrachloroethylene in isolated perfused rat liver and excreted into
bile, in the presence of a rat kidney fraction, was mutagenic in Salmonella, as was purified
TCVG (Vamvakas et al.. 1989c). This study performed the Ames assay in Salmonella
typhimurium TA100, TA98, and TA2638 with tetrachloroethylene, TCVG, and bile from liver
perfusate following tetrachloroethylene exposure in rats and demonstrated that the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-111 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
GST-metabolites or tetrachloroethylene in the presence of bile containing GST led to gene
mutations in S. typhimurium TA100. Dreessen (2003) also demonstrated for TCVG an
unequivocal dose-dependent mutagenic response in the TA100 strain in the presence of the rat
kidney S9-protein fraction; TCVC was mutagenic without metabolic activation in this strain. In
a separate study, the tetrachloroethylene metabolite TCVC (1-10 nmol/plate) was also positive
in Salmonella strains TA98 and TA100 but not strain TA2638, and inhibition of P-lyase activity
was blocked by addition of amionoxyacetic acid (AOAA) (Dekant et al.. 1986d). A subsequent
study from this same group indicated that Salmonella also was capable of deacetylating the
urinary metabolite NAcTCVC (50-100 nmol/plate) when TA100 showed a clear positive
response in the Ames assay without exogenous activation (Vamvakas et al.. 1987).
In summary, the limited in vitro studies performed in kidney cells exposed to
tetrachloroethylene or its GSH conjugation metabolites demonstrate an increase in cytotoxicity.
This cytotoxic effect was sex- and species-dependent, with increases observed in male rats and
mice compared to their female counterparts, with rats showing the most cytotoxicity. Limited
genotoxicity studies demonstrated the potential for tetrachloroethylene mutagenicity in
Salmonella strains in the presence of the kidney S9 fraction, or in Salmonella exposed to
GSH-conjugation metabolites (TCVC, TCVG, or NacTCVC) without activation.
4.2.3. Summary of Kidney Effects in Humans and Animals
Taken together, the epidemiologic studies support an association between inhalation
tetrachloroethylene exposure and chronic kidney disease, as measured by urinary excretion of
renal proteins and ESRD. The elevated urinary RBP levels seen in two studies (Mutti et al..
1992; Verplanke et al.. 1999) and lysozyme or P-glucuronidase in Franchini et al. (1983) provide
some evidence for effects to the proximal tubules from tetrachloroethylene exposure. Exposures
in the studies that observed renal toxicity were 1.2 ppm, 10 ppm, and 15 ppm (means),
representing an observational LOAEL for human kidney effects. An exposure-response
relationship was reported in one study (Trevisan et al.. 2000) but not in the other human studies
that examined renal function, an important limitation of the available data. However, as pointed
out by Mutti et al. (1992). this is a common finding among solvent-exposed populations, and
inadequate definition of the dose metric most likely contributes to the absence of exposure-
response relationships. Calvert et al. supports association between inhalation tetrachloroethylene
exposure and ESRD, particularly hypertensive ESRD. They observed a twofold elevated
incidence (SIR: 2.66, 95% CI: 1.15, 5.23) among subjects who worked only in a shop where
tetrachloroethylene was the primary cleaning solvent compared to that expected based on U.S.
population rates. An exposure-response pattern was further suggested because hypertensive
ESRD risk was highest among those employed for >5 years (SIR: 3.39, 95% CI: 1.10, 7.92). No
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-112 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
human studies investigating drinking water or other oral exposures on kidney toxicity have been
published.
Positive associations between kidney cancer (renal cell carcinoma) and exposure to dry-
cleaning and laundry workers or to tetrachloroethylene specifically were observed in several
well-conducted studies (Mandel et al.. 1995). The results from the other studies using a relatively
specific exposure-assessment approach to refine classification of potential tetrachloroethylene
exposure in dry-cleaning settings are mixed, with some studies reporting little or no evidence of
an association (Aschengrau et al., 1993; Dosemici et al., 1999; Boice et al., 1999; Lynge et al.,
2006; Pesch et al., 2000), and other studies reported elevated risks (Anttila et al., 1995; Blair et
al., 2003; Calvert et al., In Press; Schlehofer et al., 1995). An increasing trend in relative risk
with increasing exposure surrogate was not seen in any of the larger occupational exposure
studies with three or more exposure categories (Mandel et al., 1995)(Lynge et al.. 2006). but
some indication of higher risk with higher exposure (or duration) was seen in other studies (Blair
et al.. 2003). As expected, the results from sixteen other studies using a relatively nonspecific
exposure measure (broad occupational title of launderers and dry cleaners, all workers at factory,
density of dry-cleaning establishments by zip code) are more variable and less precise, reflecting
a greater potential for misclassification bias.
Adverse effects on the kidney have been observed in studies of animals exposed to high
concentrations of tetrachloroethylene by inhalation (JISA. 1993; NTP. 1986b). oral gavage
(Ebrahim et al.. 1996; Ebrahim et al.. 2001; Goldsworthy et al.. 1988; Green et al.. 1990; Jonker
et al.. 1996; NCI. 1977). and i.p. injection of tetrachloroethylene metabolites (Elfarra and
Krause. 2007). The nephrotoxic effects include increased kidney-to-body weight ratios, hyaline
droplet formation, glomerular —nephrosis' karyomegaly (enlarged nuclei), cast formation, and
other lesions or indicators of renal toxicity. Increased incidences of relatively rare renal tumors
have been observed in one bioassay of male rats exposed to tetrachloroethylene by inhalation
(NTP. 1986b). The renal effects occurred following very high (or chronic, relatively high) doses
of tetrachloroethylene exposures. Overall, multiple lines of evidence support the conclusion that
tetrachloroethylene causes nephrotoxicity in the form of tubular toxicity, mediated potentially
through the tetrachloroethylene GSH conjugation products TCVC and TCVCS.
4.2.4. Hypothesized Mode(s) of Action for Kidney Carcinogenicity
There are multiple hypothesized MO As for kidney carcinogenicity induced with
tetrachloroethylene exposure, including a2[j,-globulin accumulation, peroxisome proliferation,
genotoxicity, and cytotoxicity unrelated to a2[j,-globulin. These MO As are addressed in the
sections that follow.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-113 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
4.2.4.1. Role of Metabolism in Kidney Carcinogenicity
Except for a2[j,-globulin accumulation, which is more likely due to tetrachloroethylene
itself (Lash and Parker. 2001). other mechanisms hypothesized to contributed to
tetrachloroethylene-induced renal carcinogenicity are thought to be mediated by
tetrachloroethylene metabolites rather than by the parent compound. Metabolites from the GSH
conjugation pathway are posited to induce renal tumorigenicity, as opposed to (or to a greater
extent than) the metabolites resulting from oxidative CYP processing. The glutathione
conjugation of tetrachloroethylene in the kidney, discussed in Section 3, leads sequentially to
TCVG and TCVC. TCVC can be further processed by P-lyase to yield an unstable thiol,
1,2,2-trichlorovinylthiol, which may give rise to a highly reactive thioketene, a chemical species
that can form covalent adducts with cellular nucleophiles including DNA. TCVC can also
undergo FM03- or P450-oxidation to reactive intermediates; additionally, sulfoxidation of both
TCVC and its A-acetylated product occurs, resulting in reactive metabolites (Ripp et al.. 1999;
Ripp et al.. 1997; Werner et al.. 1996).
4.2.4.2. a2ji-Globulin Accumulation
Generally, kidney tumors observed in cancer bioassays are assumed to be relevant for
assessment of human carcinogenic potential. However, male rat-specific kidney tumors that are
caused by the accumulation of a2[j,-globulin are not generally considered relevant to humans.
Accumulation of a2[j,-globulin in hyaline droplets initiates a sequence of events that leads to
renal nephropathy and, eventually, renal tubular tumor formation. The phenomenon is unique to
the male rat because female rats and other laboratory mammals administered the same chemicals
do not accumulate a2[j.-globulin in the kidney and do not subsequently develop renal tubule
tumors (Doi et al.. 2007; Swenberg and Lehman-McKeeman. 1999; U.S. EPA. 1991a).
4.2.4.2.1. Identification of key events
The histopathological sequence of events in mature male rats is hypothesized to consist
of the following:
• Excessive accumulation of hyaline droplets containing a2[j,-globulin in renal proximal
tubules
• Subsequent cytotoxicity and single-cell necrosis of the tubule epithelium
• Sustained regenerative tubule cell proliferation
• Development of intralumenal granular casts from sloughed cellular debris associated with
tubule dilatation and papillary mineralization
• Foci of tubule hyperplasia in the convoluted proximal tubules
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-114 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
• Renal tubule tumors
4.2.4.2.2. Data requirements for establishing the MOA
The EPA (1991a) Risk Assessment Forum Technical Panel report provides specific
guidance for evaluating chemical exposure-related male rat renal tubule tumors for the purpose
of risk assessment, based on an examination of the potential involvement of a2[j,-globulin
accumulation. In particular, the following information from adequately conducted studies of
male rats is used for demonstrating that the a2[j,-globulin process may be a factor in any observed
renal effects. An affirmative response in each of the three categories is required. If data are
lacking for any of the criteria in any one category, the available renal toxicity data should be
analyzed in accordance with standard risk assessment principles. The three categories of
information and criteria are as follows:
• Increased number and size of hyaline droplets in the renal proximal tubule cells of
treated male rats. The abnormal accumulation of hyaline droplets in the P2 segment
helps differentiate a2[j,-globulin inducers from chemicals that produce renal tubule
tumors by other modes of action.
• Accumulating protein in the hyaline droplets is a2/u-globulin. Hyaline droplet
accumulation is a nonspecific response to protein overload, and, thus, it is necessary to
demonstrate that the protein in the droplet is, in fact, a2[j,-globulin.
• Additional aspects of the pathological sequence of lesions associated with a2/u-globulin
nephropathy are present. Typical lesions include single-cell necrosis, exfoliation of
epithelial cells into the proximal tubular lumen, formation of granular casts, linear
mineralization of papillary tubules, and tubule hyperplasia. If the response is mild, not
all of these lesions may be observed. However, some elements consistent with the
pathological sequence must be demonstrated to be present.
4.2.4.2.3. Induction of hypothesized key events by tetrachloroethylene
Three studies show that doses of tetrachloroethylene in excess of those observed to
induce tumorigenesis are capable of precipitating hyaline droplet nephropathy in male rats
(Bergamaschi et al.. 1992; Goldsworthy et al.. 1988; Green et al.. 1990); see Table 4-11.
Goldsworthy et al. (1988) observed increases in a2[j,-hyaline droplets in exposed male, but not
female, F344 rats following 10 days of gavage with 1,000-mg/kg tetrachloroethylene. This
finding was correlated with both protein droplet nephropathy (crystalloid accumulation) and
increases in cellular proliferation. The cell replication was enhanced in the male rats specifically
in damaged P2 segments, suggesting a link between the a2[j,-globulin accumulation and kidney
tumors. Bergamaschi et al. (1992) also demonstrated a2[j,-accumulation in P2 segments of rat
proximal tubule cells resulting from a daily exposure of rats to 500-mg/kg tetrachloroethylene in
corn oil for 4 weeks. In short-term, high-dose studies, Green et al. (1990) found that the oral
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-115 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
administration of 1,000 to 1,500 mg/kg of tetrachloroethylene daily for up to 42 days caused an
accumulation of a2[j,-globulin in the proximal tubules of male rats. These effects were not
observed in female rats or in mice.
Table 4-11. Renal a2ji-globulin formation in tetrachloroethylene-exposed
rodents
Species/strain/
sex/number
Exposure
level/duration
Effects
Reference
Mouse, B6C3Fi, both
sexes (groups of 49 or
50 mice of each sex per
dose group, total of
-300 mice)
0, 100, 200 ppm for
104 wk, inhalation
Karyomegaly and cytomegaly of the
proximal tubules in all exposed mice;
nephrosis was observed in exposed
females, casts increased in all exposed
males and in high-dose females
NTP (1986b)
Rats, F344, both sexes
(groups of 50 mice of
each sex per dose group,
total of -300 mice)
0, 200, 400 ppm for
104 wk, inhalation
Karyomegaly and cytomegaly of the
proximal tubules in all exposed rats
NTP (1986b)
F344 rats (both sexes,
5 per group)
0 or 1,000 mg/kg-day for
10 d, corn oil gavage
Increases in «2|i-hyalinc droplets in
exposed male but not female rats.
Correlated to increased cell proliferation
and protein droplet nephropathy
Goldsworthy
et al. (1988)
F344 rats (both sexes,
12 per group)
0, 500 mg/kg-day daily
for 4 wk, corn oil gavage
Increases in a2(i-hyaline accumulation in
proximal tubule cells
Bergamaschi
et al. (1992)
F344 rats (both sexes) and
B6C3Fi mice (both
sexes); 10 per group for
oral studies, 5 per group
for inhalation studies
0, 1,000 or 1,500
mg/kg-day daily by corn
oil gavage for 42 d; 0 or
1,000 ppm for 10 d
Accumulation of a2(i-globulin in proximal
tubules of male rats; nephrotoxicity also
observed in male rats (formation of
granular tubular casts and evidence of
tubular cell regeneration)
Inhalation exposure demonstrated
formation of hyaline droplets in kidneys of
male rats
Green et al.
(1990)
Green et al. (1990) tested lower inhaled tetrachloroethylene doses in rats—up to 400 ppm
for 6 hours/day for 28 days, with the animals being sacrificed within 18 hours of termination of
the final exposure—but found no evidence of hyaline droplet formation; however, there may
have been time for recovery prior to sacrifice. Green et al. (1990) proposed the possibility that
longer-term exposure to the 400 ppm concentration of tetrachloroethylene is required for the
hyaline droplet accumulation in the kidney of rats. a2[j.-Globulin accumulation can be
demonstrated, however, after only short-term exposures (even a single administration) to several
agents, such as d-limonene, decalin, unleaded gasoline, and trimethylpentane (Charbonneau et
al.. 1987; NTP. 1990b).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-116 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Lack of hyaline droplet formation, increase in a2[j,-globulin, or signs of the characteristic
renal nephropathy at the high dose level of the NTP inhalation study (NTP. 1986b) may, thus,
diminish the likelihood that the renal tumors associated with exposure to tetrachloroethylene are
induced through this mechanism (Green et al.. 1990). NTP did not report the presence of hyaline
droplets in rats that had been exposed to either 200- or 400-ppm tetrachloroethylene for up to
2 years. These doses were associated with the production of renal tubule neoplasms in male rats.
However, the fact that NTP did not report the presence of hyaline droplets in the 14-day, 90-day,
or 2-year studies is not definitive, because the NTP protocol at that time was not designed
specifically to detect hyaline droplets or a2[j,-globulin accumulation in the kidney (NTP. 1990b).
Thus, the procedures followed at the time of the study were not necessarily conducive to
detecting hyaline droplets. For example, in the chronic study of tetrachloroethylene, at least 1
week elapsed between the final tetrachloroethylene exposure and the scheduled sacrifice of the
surviving animals. It is possible that had hyaline droplets been present, they could have
regressed. Also, the nephropathy observed at the end of a 2-year bioassay could be difficult to
distinguish from the old-age nephropathy that occurs in these rats.
In contrast, the renal pathology reported in the NTP bioassay is not entirely consistent
with the results generally found for chemicals where there is a2[j,-globulin accumulation (NTP.
1986b)(letter from Scot Eustis, National Toxicology Program, to William Farland, Director,
Office of Health and Environmental Assessment, U.S. EPA, 1988). For example, there was no
mineralization in the inner medulla and papilla of the kidney, a frequent finding in bioassays of
chemicals that induce a2[j,-globulin accumulation (e.g., for pentachloroethane, the incidence of
renal papillar mineralization was 8% in controls, 59% in the low-dose group, and 58% in the
high-dose group). In addition, it is important to note that some aspects of toxic tubular
nephropathy were also observed in female rats and male mice exposed to tetrachloroethylene,
clearly contrary to sex and species specificity.
In the NCI gavage study of tetrachloroethylene (NCI. 1977). toxic nephropathy, which
was not detected in the control animals, occurred in both male and female Osborne-Mendel rats
administered tetrachloroethylene. Tetrachloroethylene also clearly caused nephropathy in both
sexes of mice in the study. Unfortunately, animal survival in the rat study was not adequate to
support any conclusions about tetrachloroethylene carcinogenicity.
In summary, although a few studies show an increase in hyaline droplets in the proximal
tubule cells of treated male rats, other studies demonstrate nephrotoxicity in both male and
female rats and mice without hyaline droplet formation. Further, the studies that demonstrate
hyaline droplet formation do not also have additional aspects of nephrotoxicity associated with
a2[j,-globulin formation. The a2[j,-globulin response reported following exposure to
tetrachloroethylene is relatively modest, and the fact that renal tumors have been observed at
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-117 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
doses lower than those shown to cause the a2[j,-globulin response is inconsistent with this
phenomenon being responsible for tumorigenesis. Chronically induced tetrachloroethylene
nonneoplastic kidney lesions exhibit neither species nor sex specificity. Unlike with other
chemicals that induce a2[j,-globulin accumulation and have been tested by NTP in chronic
carcinogenicity bioassays, renal lesions occurring in animals exposed to tetrachloroethylene were
not limited to the male rat. Although the female rat did not develop any renal tubule tumors, the
incidence of karyomegaly was significantly elevated in the female rat as well as in the male rat;
1 of 50 female rats exposed at the high dose developed tubule cell hyperplasia. Therefore, based
on the criteria described above, there are insufficient data to demonstrate renal toxicity or
cancers are caused by a2[j,-globulin formation.
4.2.4.3. Genotoxicity
A hypothesized mutagenic MOA entails the following key events leading to
tetrachloroethylene-induced kidney tumor formation: following metabolism of
tetrachloroethylene to one or more mutagenic intermediates, the genetic material is altered in a
manner that permits changes to be transmitted during cell division through one or more
mechanisms (gene mutations, deletions, translocations, or amplification); the resulting mutations
advance acquisition of the multiple critical traits contributing to carcinogenesis. This MOA may
apply to multiple cancer types.
The genotoxic potential of tetrachloroethylene is addressed in Section 4.8. To
summarize, the results of a large number of in vitro genotoxicity tests in which
tetrachloroethylene was the test agent support the conclusion that tetrachloroethylene does not
exhibit direct mutagenic activity in the absence or presence of the standard S9 fraction (Bartsch
et al.. 1979; Connor et al.. 1985; DeMarini etal.. 1994; Greim et al.. 1975; Hardin et al.. 1981;
Ha worth et al.. 1983; Kringstad et al.. 1981; Milman et al.. 1988; NTP. 1986b; Roldan-Ariona et
al.. 1991; Shimada et al.. 1985; Warner et al.. 1988; Watanabe et al.. 1998). However, the few in
vitro mutagenicity studies of tetrachloroethylene under conditions that would generate the GSH
conjugate were positive ("Vamvakas et al.. 1989b; Vamvakas et al.. 1989c). While most of these
intermediates have not been characterized for mutagenic potential, TCVG (Dreessen et al.. 2003;
Vamvakas et al.. 1989c) and jV-acetyl-,S'-(l,2,2-trichlorovinyl)-Z-cysteine (NAcTCVC)
(Vamvakas et al.. 1987) are mutagenic in the presence of activation while TCVC was mutagenic
even in the absence of activation (Dekant et al.. 1986d; Dreessen et al.. 2003). The metabolite
DCA is the most potent mutagen of the P450-derived metabolites, exhibiting mutagenic activity
in a number of assays. A putative P450-derived metabolite, 1,1,2,2-tetrachloroethylene oxide, is
also mutagenic; the mutagenicity of this epoxide would be predicted from structure-activity
relationships. Studies of chromosomal aberrations following exposure to tetrachloroethylene are
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-118 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
mostly negative, but positive results have been reported from in vitro studies with enhanced
metabolic activation (Dohertv et al.. 1996).
The limited in vivo studies of tetrachloroethylene are inconsistent, with only negative
(Bronzetti et al.. 1983; NTP. 1986b) or equivocal (Beliles et al.. 1980; Cederberg et al.. 2010)
genotoxicity assay results demonstrated following inhalation or oral exposure. These include the
finding that tetrachloroethylene at higher concentrations induces at most modest increases in
DNA damage in liver tissue (Cederberg et al.. 2010). Following in vivo exposures,
tetrachloroethylene induces SSB and DNA binding in kidney (Mazzullo et al.. 1987; Potter et al..
1996; Walles. 1986). Intraperitoneal injection assays have demonstrated both negative (NTP.
1986b) as well as positive results for different genotoxicity endpoints in other tissues (Murakami
andHorikawa. 1995). Assays of clastogenic effects following inhalation exposure in humans
have shown inconsistent results, and are suggested to be related to coexposures (Ikeda et al..
1980; Seiii etal.. 1990).
Thus, although tetrachloroethylene has largely yielded negative in standard genotoxicity
assays, uncertainties remain with respect to the possibility that genotoxicity contributes to renal
carcinogenesis. Not all metabolites have been identified or characterized, but several known
metabolites including those derived from P450 as well as GSH pathways are mutagenic in the
standard battery of tests. Tetrachloroethylene is mutagenic in bacterial assays in the presence of
GST and GSH whereas the standard S9 fraction has typically yielded negative results.
Tetrachloroethylene at higher concentrations also induces modest increases in DNA damage and
DNA binding in liver tissue (Cederberg et al.. 2010; Murakami andHorikawa. 1995). Given the
demonstrated mutagenicity of several tetrachloroethylene metabolites, the hypothesis that
mutagenicity contributes to the MOA for tetrachloroethylene carcinogenesis cannot be ruled out,
although the specific metabolic species or mechanistic effects are not known.
4.2.4.4. Peroxisome Proliferation
The PPARa-agonism MOA is also hypothesized to induce rat kidney tumorigenesis.
According to this hypothesis, the key events leading to tetrachloroethylene-induced kidney tumor
formation constitute the following, after activation of tetrachloroethylene to one or more reactive
metabolites: the PPARa receptor is activated, which then causes alterations in cell proliferation
and apoptosis, followed by clonal expansion of initiated cells.
Limited data exist to support increased peroxisome proliferation in rodent kidney
following exposure to tetrachloroethylene and are summarized in Table 4-12 (Goldsworthy and
Popp. 1987; Odum et al.. 1988b). The role of peroxisome proliferation in tetrachloroethylene-
induced kidney toxicity and cancer was examined in male and female F344 rats and B6C3Fi
mice exposed to tetrachloroethylene by inhalation (400 ppm, 6 hours/day, 14, 21, or 28 days or
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-119 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
200 ppm, 6 hours/day, 28 days) in (Odum et al.. 1988b). Five animals per group were exposed.
Insufficient mouse kidney tissue limited the analysis to pooled samples. Slight increases were
observed in P-oxidation in mouse kidney (maximum 1.6-fold increase at 21 days, 400-ppm
exposure). Modest palmitoyl-CoA oxidation (PCO) increases were seen in the kidney of male
rats at 200 ppm at 28 days (1.3-fold) but not 400 ppm at 14, 21, or 28 days. In female rat kidney,
PCO was elevated (approximately 1.6-fold) at all doses and times. However, peroxisome
proliferation was not seen in rat or mouse kidney upon microscopy, suggesting that this does not
play a role in kidney carcinogenesis.
Table 4-12. Renal peroxisome proliferation in tetrachloroethylene-exposed
rodents
Species/strain/sex/number
Effect
Dose
Time
Rat, F344; and mouse,
B6C3Fi; both sexes
(5/group)
Odumetal. (1988b)
Mice of both sexes: Analysis in mice was limited
to pooled tissue, but showed slight increases in
P-oxidation in mouse kidney
200, and
400 ppm,
inhalation
14, 21,28 d
Rats of both sexes: Modest increases in PCO
observed in male rat kidneys at 200 ppm for 28 d
only, but elevated in female rat kidneys at all
doses and times
200, and
400 ppm,
inhalation
14, 21,28 d
F344 rats (male only,
5/group) and B6C3Fi mice
(male only, 5/group)
Goldsworthy and Popp
(1987)
Mice: Increased PCO activity in all exposed mice
1,000 mg/kg-day
for 10 d, corn oil
gavage
10 d
Rats: Increased kidney weight in exposed rats
1,000 mg/kg-day
for 10 d, corn oil
gavage
10 d
Goldsworthy and Popp (1987) administered tetrachloroethylene (1,000 mg/kg-day) by
corn oil gavage to 5 male F344 rats and 5 male B6C3Fi mice for 10 days. In
tetrachloroethylene-exposed rats, PCO was modestly although not significantly elevated in the
liver (1.4-fold increase) and kidney (1.7-fold increase). In mice, tetrachloroethylene exposure
increased PCO activity 4.3-fold in liver and by 2.3-fold in kidney. Relative liver weight was
increased in rats and mice with tetrachloroethylene exposure, but relative kidney weight was
unaffected. A comparison of corn oil with methyl cellulose revealed no effect of the gavage
vehicle on tetrachloroethylene-induced PCO. A less-than-additive effect of trichloroethylene
(1,000 mg/kg) administered together was tetrachloroethylene on PCO induction was seen.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-120 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
4.2.4.5. Cytotoxicity/Sustained Chronic Nephrotoxicity Not Associated with a2ji-
Globulin Nephropathy
The hypothesis is that renal neoplasms induced by tetrachloroethylene arise secondary to
renal cytotoxicity and subsequent cellular proliferation without regard to a2[j,-accumulation.
This MOA entails the following key events leading to tetrachloroethylene-induced kidney tumor
formation: following metabolism of tetrachloroethylene to one or more reactive intermediates,
toxicity to the kidney ensues and is sustained; via a variety of potential mechanisms (damage to
and alteration of macromolecules, cell signaling alterations, etc.), the acquisition of the multiple
critical traits contributing to carcinogenesis is advanced.
The kidney is a major target organ for tetrachloroethylene-induced toxicity through the
reactive metabolites produced subsequent to GSH conjugation. Renal tubule neoplasia is
observed to occur only in male rats. This species- and sex-specific response would not be
expected based on the hypothesized MOA, because tetrachloroethylene has been reported to
produce nephrotoxicity across species, and in both sexes. Signs of tetrachloroethylene-induced
kidney damage appeared in both rats and mice during the early phases of the NTP inhalation
study, for example, indicating that animals of both species surviving to the scheduled termination
of the study had long-standing nephrotoxicity. Although the female rats did not develop any
renal tubule tumors, the incidence of karyomegaly was significantly elevated in females as well
as in males, and 1/50 female rats exposed at the high dose developed tubule cell hyperplasia
(NTP. 1986b).
In the NTP study of the mouse, —nephrcis" was observed at increased incidences in
dosed females, casts were observed at increased incidences in dosed males and high-dose
females, and karyomegaly of the tubule cells was observed at increased incidences in both sexes
of treated mice (NTP. 1986b). The severity of the renal lesions was dose related, and one low-
dose male had a renal tubule cell adenocarcinoma. In the NCI gavage study of B6C3Fi mice and
Osborne-Mendel rats exposed to tetrachloroethylene, toxic nephropathy was not detected in
control animals but did occur in both male and female rats as well as in mice (NCI. 1977).
Mechanistic studies of tetrachloroethylene nephrotoxicity are relatively sparse. Most
studies performed to elucidate information related to understanding tetrachloroethylene renal
toxicity have concentrated on the GSH pathway metabolites rather than on the parent chemical;
this is because much of the available data for both tetrachloroethylene and trichloroethylene
suggest that it is flux through this pathway that generates reactive chemical species responsible
for nephrotoxicity. Vamvakas et al. (1989a; 1989d) have shown the tetrachloroethylene
conjugate metabolites TCVG and TCVC to cause dose-related cytotoxicity in renal cell
preparations and prevention of this toxicity by P-lyase enzyme inhibitor. Renal P-lyases are
known to cleave TCVC to yield an unstable thiol, 1,2,2-trichlorovinylthiol, that may give rise to
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-121 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
a highly reactive thioketene, a chemical species that can form covalent adducts with cellular
nucleophiles. Additionally, sulfoxidation of both TCVC and its A'-acetylated product occurs,
resulting in toxic metabolites (Ripp et al.. 1999; Ripp et al.. 1997; Werner et al.. 1996). Findings
using in vitro models studied by Lash et al. (2002) suggest a marked sex difference between
male and female rats in the severity of acute renal toxicity caused by both tetrachloroethylene
and its TCVG metabolite. Tetrachloroethylene and TCVG also produced signs of toxicity in
mitochondria; i.e., mitochondrial dysfunction, such as inhibition of state 3 respiration by specific
inhibition of several sulfhydryl-containing enzymes in both sexes of mice (Lash et al.. 2000;
Lash and Parker. 2001; Lash et al.. 2002).
4.2.4.6. Summary
The kidney is a target organ in mammalian species for tetrachloroethylene and other
related chlorinated ethanes and ethylenes, and tetrachloroethylene causes kidney cancer in male
rats. It is likely that several mechanisms contribute to tetrachloroethylene-induced kidney
cancer. Mutagenicity, peroxisome proliferation, a2[j,-globulin nephropathy, and cytotoxicity not
associated with a2[j,-globulin accumulation are MO As that have been investigated. Except for
a2[j,-globulin accumulation, which is more likely due to tetrachloroethylene itself (Lash and
Parker. 2001). other mechanisms hypothesized to contributed to tetrachloroethylene-induced
renal carcinogenicity are thought to be mediated by tetrachloroethylene metabolites rather than
with the parent compound.
Metabolites from the GSH conjugation pathway are posited to induce renal
tumorigenicity, as opposed to or to a greater extent than the metabolites resulting from oxidative
CYP processing. The glutathione conjugation of tetrachloroethylene in the kidney, discussed in
Section 3, leads sequentially to TCVG and TCVC. TCVC can be further processed by P-lyase to
yield an unstable thiol, 1,2,2-trichlorovinylthiol, that may give rise to a highly reactive
thioketene, a chemical species that can form covalent adducts with cellular nucleophiles
including DNA. TCVC can also undergo FM03 or P450 oxidation to reactive intermediates;
additionally, sulfoxidation of both TCVC and its A-acetyl ated product occurs, resulting in
reactive metabolites (Ripp et al.. 1999; Ripp et al.. 1997; Werner et al.. 1996). While most of
these intermediates have not been characterized for mutagenic potential, TCVG, TCVC, and
NAcTCVC are clearly mutagenic in Salmonella tests. In addition, tetrachloroethylene exhibited
mutagenicity in Salmonella in the few studies of conditions that could generate GSH-derived
metabolites. Tetrachloroethylene, following in vivo exposures, also binds to kidney DNA and
induces SSB in kidney. Given the known mutagenicity of the GSH-derived tetrachloroethylene
metabolites that are formed in the kidney, and the observed in vitro mutagenicity of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-122 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
tetrachloroethylene under conditions that would generate these metabolites, a mutagenic MOA
contributing to the development of the kidney tumors cannot be ruled out.
Due to tetrachloroethylene's nephrotoxic effects, it has been suggested that the low-level
renal tumor production observed in exposed rats is secondary to sustained cytotoxicity and
necrosis leading to activation of repair processes and cellular regeneration. However,
—nephrtoxicity" occurs in both sexes of rats and mice whereas cell replication and
tumorigenesis occurs in male but not in female rats. In addition, tetrachloroethylene induces
kidney tumors at lower doses than those required to cause a2[j,-globulin accumulation, raising
serious doubt that a2[j,-globulin plays a key role—especially any major role—in the rat kidney
tumor formation.
Because tetrachloroethylene has been shown to induce peroxisome proliferation, an
indicator of PPAR-activation, the possibility exists that certain responses resulting from
activation of PPAR receptors might be involved in cancer-causing activity leading to
tetrachloroethylene-induced renal tumors. However, the chemical-specific data are limited and
show only modest effects at exposures exceeding those required for renal carcinogenesis. There
is no evidence causally linking PPARa-activation to kidney tumorigenesis for
tetrachloroethylene or other compounds.
In summary, the complete mechanisms of tetrachloroethylene-induced renal
carcinogenesis are not yet understood. Given the known mutagenicity of the GSH-derived
tetrachloroethylene metabolites that are formed in the kidney, and the observed in vitro
mutagenicity of tetrachloroethylene under conditions that would generate these metabolites, a
mutagenic MOA contributing to the development of the kidney tumors cannot be ruled out.
4 3 LIVER TOXICITY AND CANCER
4.3.1. Human Studies
A number of hepatotoxic effects, including hepatomegaly, hepatocellular damage, and
elevations of several hepatic enzymes and bilirubin degradation byproducts, have been observed
after acute high-level exposure to tetrachloroethylene (levels not identified; Meckler and Phelps
(1966); Coler and Rossmiller (1953); Hake and Stewart (1977); Saland (1967); Stewart et al.
(1961). as reported in ATSDR (1997b)). One case report noted obstructive jaundice and
hepatomegaly in an infant exposed orally to tetrachloroethylene [1 mg/dL; Bagnell and
Ellenberger (1977). as reported in ATSDR (1997b)].
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-123 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
4.3.1.1. Liver Damage
Four cross-sectional studies were available that evaluated the prevalence of liver damage
among dry-cleaner populations (Brodkin etal.. 1995; Cai et al.. 1991; Gennari et al.. 1992;
Lauwerys et al.. 1983). These studies assessed serum concentration of a number of hepatic
enzymes in dry-cleaner and control populations. Additionally, sonographic changes to hepatic
parenchymal tissue were examined in one study (Brodkin et al.. 1995). An elevated
concentration of the serum enzyme GGT and mild hepatic changes were notable observations in
two studies (Brodkin et al.. 1995; Gennari et al.. 1992).
Gennari et al. (1992) measured the electrophoretic fractionation patterns of serum GGT
isozymes among 141 tetrachloroethylene-exposed dry cleaners and 130 nonexposed controls
selected from staff and students from the academic institution of the principal investigators.
Both the exposed subjects and the controls had similar lifestyle (smoking, alcohol consumption)
and clinical medical histories. The TWA tetrachloroethylene concentration in the dry-cleaning
facilities was 11.3 ppm. Total GGT was higher in exposed workers (exposed: mean of 12.4
international units per liter [U/L; standard deviation, 6.9 U/L]; controls: 8.8 U/L [4.9 U/L],
p < 0.01). The GGT-2 isoenzyme component was higher in exposed workers (6.8 U/L [5.7 U/L]
in exposed vs. 3.5 U/L [3.3 U/L] in controls,/? < 0.01) and the GGT-4 component was detectable
in exposed workers but not measurable in controls. The authors regarded a GGT-2/GGT-3 ratio
of greater than 1 as a sensitive index of the reciprocal behavior of the two isoenzymes. GGT-2 is
generally associated with activation of liver microsomal enzymes. GGT-4 is common in liver
diseases and indicates hepato-biliary impairment.
This study excluded individuals who presented values for GGT, or other liver enzymes
above a normal range, and individuals who had past or current liver disease. None of the
workers showed any clinical symptoms of liver disease, and their enzymatic profiles, including
GGT, aspartase amino transaminase (AST), alanine amino transaminase, 5'-nucleotidase, and
alkaline phosphatase, were within the clinically normal reference limits. Given the study's
exclusion criteria, it is not surprising that liver enzyme concentrations were within a normal
range. The authors stated that more research is required to develop this GGT fractionation assay
into a clinically useful method of measuring liver function. Nevertheless, the study showed that
these dry cleaners had markers of tetrachloroethylene oxidative metabolism (GGT-2) and liver
impairment (GGT-4).
The study by Brodkin et al. (1995) examined liver function and carried out sonography
measurements in a population of 27 dry cleaners and 26 nonexposed laundry workers. Dry
cleaners were older and had a longer duration of employment than did laundry workers. The
mean TWA exposure (8 hours) among all dry cleaners was 15.8 ppm (range: 0.4-83 ppm). The
investigators found a higher prevalence of abnormal hepatic sonograms among the dry cleaners
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-124 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
(67%) than among laundry workers (38% ;p< 0 .05), the control group. The noninvasive imaged
penetration of ultrasound into liver tissue can reveal the presence of fat accumulation and fibrous
structures. Hepatic parenchymal changes were graded as mild, moderate, or severe. The
prevalence of hepatic parenchymal changes increased both with increasing current concentration
and with cumulative exposure (p < 0.05). Subjects with serological evidence of active hepatitis
infection were excluded from these analyses.
Brodkin et al. (1995) fit logistic regression models to examine possible associations
between mild or greater parenchymal changes and tetrachloroethylene exposure. These analyses
included adjustment for the effects of ethanol consumption within the past 6 months, sex, body
mass index, age, and serological evidence of active and past hepatitis infection. Subjects with
serological evidence of active hepatitis infection were included in the logistic regression analysis
due to the ability of the statistical method to account for the effects associated with this factor.
These analyses showed subjects exposed during older wet or dry-to-dry transfer processes
(average concentration: 19.8 ppm; range: 1.8-83 ppm) was strongly—but imprecisely—
associated with mild or greater sonographic changes (OR: 4.2, 95% CI: 0.9-20.4) as compared
with controls. No association was shown with subacute exposure in new dry-to-dry operations
(OR: 0.7, 95% CI: 0.1-5.9). An inverse dose-response association was found with cumulative
exposure after adjustment for age due to a strong but imprecise association between
tetrachloroethylene exposure and hepatic sonographic changes in younger workers (workers less
than 35 years of age, OR: 15; 95% CI: 1.33-170).
Only 2P/o of the exposed study subjects who had changes graded as mild or greater had
increases in any hepatic enzyme (Brodkin et al.. 1995). Mean concentrations of GGT, AST, and
alanine transferase (ALT) tended to be higher among the dry cleaners as compared with laundry
workers; however, the differences were not statistically significant and all mean values were
within the normal range of reference values. However, all of the subjects who had elevated ALT
concentrations had moderate or severe sonographic changes. Hence, sonographic imaging of the
liver appeared to be a more sensitive indicator of toxicity than was measurement of serum
hepatic enzymes.
Lauwerys et al. (1983) performed behavioral, renal, hepatic, and pulmonary tests on 22
subjects exposed to tetrachloroethylene in six dry-cleaning shops and compared the results with
those obtained for 33 subjects nonoccupationally exposed to organic solvents. The mean TWA
concentration was 21 ppm. The investigators found no statistically significant differences in
mean serum hepatic enzyme concentration between exposed subjects and controls, but this study
is poorly reported and the authors did not describe the statistical methods used to test for
differences between the exposed and control groups.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-125 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Cai et al. (1991) investigated subjective symptoms, hematology, serum biochemistry, and
other clinical signs in 56 dry cleaners exposed to tetrachloroethylene at 20 ppm (as a geometric
mean of 8-hour TWA) and compared the results with findings for 69 nonexposed controls from
the same factories. Exposure-related increases were observed in the prevalence of subjective
symptoms during the workday as well as in the past 3-month period, whereas no significant
changes in hematology were seen. There was no effect on liver and kidney function, as
measured by enzyme activities, blood urea nitrogen (BUN), and creatinine in the serum.
Table 4-13 presents a summary of the human liver toxicity studies in dry cleaners. Two
of the four studies (Brodkin et al.. 1995; Gennari et al.. 1992) showed clinical signs of liver
toxicity, namely, sonographic changes in the liver and higher serum concentrations of liver
enzymes indicative of liver injury in the absence of frank toxicity. Subjects in these two studies
were exposed to tetrachloroethylene for a longer duration than were subjects in Cai et al. (1991)
or Lauwerys et al. (1983). and for this reason these two studies carry greater weight in this
analysis. Moreover, the studies by Brodkin et al. (1995) and Gennari et al. (1992) assessed
potential liver damage using a different set of markers than those of Cai et al. (1991) or
Lauwerys et al. (1983).
Table 4-13. Summary of studies of human liver toxicity
Subjects
Effects
Exposure
Author
27 PCE-exposed dry cleaners
26 nonexposed laundry
workers
Sonographic scattering of fat in
liver (in vivo)
Severity greater with higher
cumulative exposure
No liver toxicity
Group mean TWA =
15.8 ppm
Mean duration of exposure = 12 yr
Brodkin et al.
(1995)
141 PCE-exposed dry cleaners
130 controls
Elevation of total GGT due to
GGT-2
GGT-4 detected in exposed but
not in control workers
Mean TWA = 11.3 ppm
Mean duration of exposure = 20 yr
Gennari et al.
(1992)
24 PCE-exposed dry cleaners
33 controls nonoccupational^
exposed to organic solvents
No effect on serum hepatic
enzymes
Mean TWA = 21 ppm
Mean duration of exposure = 6 yr
Lauwerys et
al. (1983)
56 PCE-exposed dry cleaners
69 nonexposed factory controls
Increased subjective symptoms
No effects on serum indicators
of liver and kidney toxicity
Geometric mean TWA = 20 ppm
Mean duration of exposure = 3 yr
Cai et al.
(1991)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-126 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Biological markers of liver effects permit the early identification of adverse effects of
xenobiotic exposure. They are an important link between biological markers of exposure and
frank liver toxicity, and they offer the most potential for clinical intervention before irreversible
effects have occurred (NR.C. 1995). The observations of Brodkin et al. (1995) and Gennari et al.
(1992) support the indication that tetrachloroethylene exposure affects liver function; hence, the
lowest-observed-adverse-effect level (LOAEL) for liver effects in humans can be established as
a range from 12 to 16 ppm (TWA).
4.3.1.2. Liver Cancer
Eighteen epidemiologic studies reporting data on liver cancer and tetrachloroethylene
exposure were identified. This set of studies includes 13 cohort studies on liver cancer
(Andersen et al.. 1999; Blair et al.. 2003; Boice et al.. 1999; Bond et al.. 1990; Calvert et al.. In
Press; Ji and Hemminki. 2005c; Lindbohm et al.. 2009; Lynge et al.. 1995; Lynge and Thygesen.
1990; Pukkala et al.. 2009; Selden and Ahlborg. 2011; Sung et al.. 2007; Travier et al.. 2002).
three liver cancer case-control studies of occupational exposures (Lynge et al.. 2006; Stemhagen
et al.. 1983; Suarez et al.. 1989). and two liver cancer case-control studies of residential exposure
(Lee et al.. 2003; Vartiainen et al.. 1993). Two other cohort studies included information on
tetrachloroethylene but did not report risk estimates for liver cancer (Anttila et al.. 1995; Radican
et al.. 2008). as well as an earlier report of mortality by Chang et al. (2003) for subjects in Sung
et al. (2007). did not provide an estimate of the association for liver cancer. Additionally, three
liver cancer case-control studies that examined occupational exposure did not report an odds
ratio for holding an occupation or for work in a dry cleaner and laundry (Austin et al.. 1987;
Ferrand et al.. 2008; Houten and Sonnesso. 1980) and so were not evaluated further. The
seventeen studies represent the core studies evaluated by EPA, as described in more detail below.
Appendix B reviews the design, exposure-assessment approach, and statistical methodology for
each study. Most studies were of the inhalation route of exposure, of occupational exposure, and
lacked quantitative exposure information.
Thirteen studies reporting risk estimates for liver cancer examine occupational title as dry
cleaner, launderer, and presser as surrogate for tetrachloroethylene, given its widespread use
from 1960 onward in the United States and Europe (Andersen et al.. 1999; Blair et al.. 2003;
Calvert et al.. In Press; Ji and Hemminki. 2005c; Lindbohm et al.. 2009; Lynge et al.. 2006;
Lynge et al.. 1995; Lynge and Thygesen. 1990; Pukkala et al.. 2009; Selden and Ahlborg. 2011;
Stemhagen et al.. 1983; Suarez et al.. 1989; Travier et al.. 2002). Six studies conducted in
Nordic countries are either based on the entire Swedish population or on combined populations
of several Nordic countries; strengths of these studies are their use of job title as recorded in
census databases and ascertainment of cancer incidence using national cancer registries
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-127 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
(Andersen et al.. 1999; Lindbohm et al.. 2009; Lynge et al.. 2006; Lynge et al.. 1995; Lynge and
Thygesen. 1990; Pukkala et al.. 2009; Selden and Ahlborg. 2011). Lynge et al. (1995) is a
nested case-control study of subjects in Lynge and Thygsen (1990). Subjects in the multi-Nordic
country of Pukkala et al. (2009) overlapped those of Lynge and Thygesen (1990). Lynge et al.
(1995). Andersen et al. (1999). Lynge et al. (2006). and Selden and Ahlborg (2011). Studies
examining mortality among U.S. dry-cleaner and laundry workers (Blair et al.. 2003; Ruder et
al.. 2001) are of smaller cohorts than the Nordic studies, with fewer observed liver cancer events.
The exposure surrogate in studies of dry-cleaners and laundry workers is a broad
category containing jobs of differing potential for tetrachloroethylene exposure. Thus, these
studies have a greater potential for exposure misclassification bias compared to studies with
exposure potential to tetrachloroethylene assigned by exposure matrix approaches applied to
individual subjects. One dry-cleaning study included an analysis of subjects whose beginning
employment date was after 1960 (Calvert et al.. In Press), reducing the potential for coexposures
to other solvents in this setting. Lynge et al. (1995) classifies separately subjects in Lynge and
Thygsen (1990) as either dry cleaners or laundry workers using occupation and workplace
description from 1970 Census records. Lynge et al. (2006). using job title reported in the 1970
Census, identified subjects as dry cleaner (defined as dry cleaners and supporting staff if
employed in business of <10 workers), other job titles in dry cleaning (launderers and pressers),
unexposed (job title reported on 1970 Census was other than in dry cleaning), or unclassifiable
(information was lacking to identify job title of subject). Selden and Alhborg (2011) identified
subjects as either dry cleaners, assigned with potential for tetrachloroethylene exposure, or
laundry workers, assigned as unexposed, and presented risk estimates separately by job title.
Lindbohm et al. (2009) using a job exposure matrix approach based on job title and exposures
assigned a cumulative exposure index to chlorinated hydrocarbons to individual subjects.
Tetrachloroethylene is one of several chlorinated solvents included in the broad category, but
Lindbohm et al. (2009) do not present risk estimates for tetrachloroethylene-only subjects.
Three other cohorts with potential tetrachloroethylene exposure in industrial settings have
been examined. These studies include aerospace or aircraft maintenance workers in the United
States (Boice et al.. 1999). workers, electronic factory workers in Taiwan (Sung et al.. 2007). and
workers at a Dow plant in Michigan (Bond et al.. 1990). Boice et al. (1999) used an exposure
assessment based on a job-exposure matrix and Bond et al. (1990). a nested case-control study,
used company work history records to assign potential tetrachloroethylene exposure to individual
subjects. In contrast and less sensitive, the exposures in the Taiwan studies included multiple
solvents and tetrachlorethylene exposure was not linked to individual workers and cohorts
included white-collar workers, who had an expected lower potential for exposure (Sung et al..
2007).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-128 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Two geographical studies focused on residential proximity to drinking water sources
contaminated with tetrachloroethylene and other solvents. Vartiainen et al. (1993) examines
liver cancer incidence in two southern Finnish municipalities, with the exposure surrogate
assigned uniformly to all subjects. Lee et al. (2003) using a morality odds ratio approach
examined residence in two communities surrounding the factory whose workers were studied by
Chang et al. (2003; 2005) and Sung et al. (2007). One village upstream from the factory was
considered as unexposed and another village downstream from the factory identified as exposed
based on groundwater monitoring of drinking water wells during the period 1999-2000.
In summary, with respect to exposure-assessment methodologies, five studies with liver
cancer data assigned tetrachloroethylene exposure to individuals within the study using a job
exposure matrix (Boice et al.. 1999; Bond et al.. 1990). restricting the cohort to subjects who
started working after 1960 (Calvert et al.. In Press), or restricting analyses to subjects identified
as dry cleaners (Lynge et al.. 1995; Selden and Ahlborg. 2011). One other study sought
additional data using a questionnaire for use in refining potential exposure within dry-cleaning
settings (Lynge et al.. 2006). The relative specificity of these exposure-assessment approaches
strengthens their ability to identify cancer hazards compared to studies with broader and less
sensitive exposure-assessment approaches. The least sensitive exposure assessments are those
using very broad definitions such as working in a plant or factory (Chang et al.. 2003; Sung et
al.. 2007).
Four1 of the sixteen liver cancer studies evaluated by EPA with exposure-assessment to
tetrachloroethylene or employment as dry-cleaner or laundry worker reported estimated relative
risks based on 50 or more observed events (Ji and Hemminki. 2005c; Lynge et al.. 2006; Pukkala
et al.. 2009; Travier et al.. 2002). The observed number of liver cancer cases in these studies
ranged from 58 (Lynge et al.. 2006) to 113 (Pukkala et al.. 2009). The four large cohort studies
observed a standardized incidence ratio of 0.76 (95% CI: 0.38, 1.52), 1.02 (95% CI: 0.84, 1.24),
1.22 (95% CI: 1.03, 1.45), and 1.23 (95% CI: 1.02, 1.49) in Lynge et al. (2006). Travier et al.
(2002). Ji and Hemminki (2005c). and Pukkala et al. (2009). respectively, for the association
between liver cancer risk and ever having a job title of dry-cleaner or laundry worker (see
Table 4-14).
In addition to the evidence from the large cohort and case-control studies, eleven other
studies reported effect estimates for liver cancer based on fewer observed events and carry lesser
weight in the analysis. As expected, the magnitude of the point estimate of the association2
1 Lynge and Thygsen (19901 and Andersen et al. (19991 are not included in this summary of the data from the
individual studies because they were updated and expanded in the analysis by Lynge et al. (1995) and Pukkala et al.
(20091. respectively.
2In Lynge et al. (19951. all 17 primary liver cancer deaths occurred among laundry workers and a risk estimate and
associated 95% CIs were not presented for dry cleaners.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-129 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
reported in these studies is more variable than in the larger studies: 0.13 to 0.98 (Calvert et al.. In
Press; Suarez et al.. 1989; Sung et al.. 2007; Vartiainen et al.. 1993)(Blair et al., 2001), 1.2 to 1.8
(Bond et al.. 1990; Lindbohm et al.. 2009; Selden and Ahlborg. 2011) and 2.05 to 2.57 (Boice et
al.. 1999; Stemhagen etal.. 1983)(Lee et al., 2006). Only the 95% CIs of the risk estimate of Lee
et al. (2006) excluded 1.0.
Establishment of an exposure or concentration-response relationship can add to the
weight of evidence for identifying a cancer hazard, but only limited data pertaining to exposure-
response relationships for liver cancer and tetrachloroethylene exposure are available. Four
studies of liver cancer presented risk estimates for increasing exposure categories using exposure
duration, a proxy inferior to cumulative exposure due to inability to account for temporal
changes in exposure intensity (Boice et al.. 1999; Lynge et al.. 2006; Selden and Ahlborg. 2011;
Travier et al.. 2002). Boice et al. (1999) presents a statistical test for linear trend for subjects
with intermittent-routine tetrachloroethylene exposure, a broader category than that used to
examine overall tetrachloroethylene exposure (comprised of routine-exposed subjects only), and
reported a/>value of >0.20. In Travier et al. (2002). the standardized incidence ratio estimate
was 1.20 (95% CI: 073, 2.18) for dry-cleaners and laundry workers in both 1960 and 1970
Censuses, compared to 1.02 (95% CI: 0.84, 1.24) for only subjects in one of these census.
Standardized incidence ratio estimates for both males and females with tetrachloroethylene
exposure in Selden and Ahlborg (2011) appeared to decrease monotonically with increasing
employment duration.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-130 DRAFT-DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-14. Summary of human studies on tetrachloroethylene exposure and liver cancer
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Cohort Studies
Biologically monitored workers
Anttila et al. (1995)
All subjects
Not reported
849 Finnish men and women, blood PCE [0.4 (imol/L in females and 0.7
(imol/L in males (median)], follow-up 1974-1992, external referents (SIR)
Aerospace workers (Lockheed)
Boice et al. (1999)
Routine exposure to PCE
2.05 (0.83, 4.23)
7
77,965 (n = 2,631 with routine PCE exposure and n = 3,199 with
intermittent-routine PCE exposure), began work during or after 1960,
worked at least 1 yr, follow-up 1960-1996, job exposure matrix without
quantitative estimate of PCE intensity, 1987-1988 8-h TWA PCE
concentration (atmospheric monitoring) 3 ppm [mean] and 9.5 ppm
[median], external reference for routine exposure (SMR) and internal
references (workers with no chemical exposures) for routine-intermittent
PCE exposure (RR), liver and biliary tract (ICD-9, 155, 156)
Routine-Intermittent exposure duration to PCE
Not reported
0
1.0a
22
<1 yr
1.38 (0.40, 4.69)
3
1-4 yr
1.17 (0.39,3.47)
4
>5 yr
1.29 (0.46, 3.65)
5
p-valuc for trend
p > 0.20
Chemical workers
Bond et al. (1990)
PCE
1.8(0.8, 4.3)
6
Nested case-control study with cohort (n = 21,437 males), follow-up
1940-1982, 44, liver and biliary tract deaths, unmatched controls
randomly selected from cohort, PCE and 10 other potential exposures
assigned to individual subjects based on company records, Mantel-
Haenxzel x2 (OR)
Electronic factory workers (Taiwan)
Chang et al. (2003); Sung et al. (2007)
All Subjects
86,868 (n = 70,735 female), follow-up 1979-1997, multiple solvents
exposure, does not identify PCE exposure to individual subjects, cancer
mortality, external referents (SMR) (Chans et al.. 2003). Drimarv liver
cancer (A095)
63,982 females, follow-up 1979-2001, factory employment proxy for
exposure, multiple solvents exposures and PCE not identified to individual
subjects, cancer incidence, external referents, analyses lagged 10 yr (SIR),
liver and interheaotic bile ducts (Suns et al.. 2007)
Males
Not reported
0
0.69 exp
Females
Not reported
0
0.57 exp
Females
0.79 (0.55, 1.10)
36
a -r
>1
to 0\
2 ^
3
LtJ
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-14. Summary of human studies on tetrachloroethylene exposure and liver cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Aircraft maintenance workers from Hill Air Force Base
Radican et al. (2008)
Any PCE exposure
Not reported
10,461 men and 3,605 women (total n = 14,066, n = 10,256 ever exposed
to mixed solvents, 851 ever-exposed to PCE), employed at least 1 yr from
1952 to 1956, follow-up 1973-2000, job exposure matrix (intensity),
internal referent (workers with no chemical exposures (RR)
Dry-cleaner and laundry workers
Andersen et al. (1999)
All laundry worker and dry cleaners
1.30 (0.93, 1.78)
39
29,333 men and women identified in 1960 Census (Sweden) or 1970
Census (Denmark, Finland, Norway), follow-up 1971-1987 or 1991, PCE
not identified to individual subjects, external referents (SIR), Primary liver
cancer (ICD-7, 155.0)
Males
1.26 (0.69, 2.21)
11
Females
1.32 (0.88, 1.91)
28
Blair etal. (2003)
All subjects
0.8 (0.4, 1.5)
10
5,369 U.S. men and women laundry and dry-cleaning union members
(1945-1978), follow-up 1979-1993, semiquantitative cumulative
exposure surrogate to dry clean solvents, cancer mortality, external
referents (SMR), liver and gallbladder (ICDA-8, 155)
Semiquantitative exposure score
Not reported
Ji and Hemminki. (2005c)
Laundry workers and dry cleaners in 1960 Census
1.22 (1.03, 1.45)b
138
9,255 Swedish men and 14,974 Swedish women employed in 1960 (men)
or 1970 (women) as laundry worker or dry cleaner, follow-up
1961/1970-2000, PCE not identified to individual subjects, external
referent (SIR) and adjusted for age, period and socioeconomic status
Males
1.30 (0.97, 1.67)b
52
1.09 (0.70, 1.56)°
25
1.52 (0.83, 2.43)d
14
1.61 (0.88, 2.57)e
14
Females
1.18(0.94, 1.44)b
86
1.26 (0.82, 1.81)°
25
1.05 (0.75, 1.40)d
39
1.39 (0.87, 2.04)e
22
a -r
>1
to 0\
2 ^
3
U>
to
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
0
1
a -r
>1
to 0\
2 ^
U>
U>
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Lindbohm et al. (2009)
Launderers and dry cleaners
1.22 (0.56, 2.33)
9
Finnish population born 1906-1945 and participated in 1970 Census,
follow-up 1971-1995, Finnish cancer registry, 1,691 males and 783
female primary liver cancers, longest held occupation reported on 1970
Census, laundry and dry-cleaner exposure surrogate, external referent for
analyses examining job title (SIR) and all-other job titles for analyses for
chlorinated hydrocarbon (RR) adjusted for age, period, social class,
smoking and alcohol consumption
Males
2.91 (0.35, 4.26)
2
Females
1.05 (0.42,2.16)
7
Cumulative exposure chlorinated HCs
None
1.0a
1,618
<5 ppm-yr
1.25 (0.80, 1.95)b
1.23 (0.68, 2.24)°
20
11
5-49 ppm-yr
1.13 (0.84, 1.53)b
1.22 (0.83, 1.80)°
44
27
>50 ppm-yr
2.65 (1.38, 5.1 l)b
3.59 (1.71, 7.57)c
9
7
Lvnge and Thvgsen (1990); Lvnge et al. (1995)
All laundry worker and dry cleaners
2.19(0.88,4.51)
7
10,600 Danish men and women, 20-64 yr old, employed in 1970 as
laundry worker, dry cleaners and textile dye workers, follow-up
1970-1980, external referents (SIR), Primary liver cancer (ICD-7, 155)
(Lvnse and Thvsesen. 1990)
Nested case-control studv within Lvnse and Thvssen (19901 17 Drimarv
liver cancer cases in men and women, follow-up 1970-1987, 85 controls
randomly selected from within cohort, matched on sex, age, and
occupation, dry cleaner assigned using occupation and workplace on 1970
Census form, losistic regression (OR) (Lvnse et al.. 1995)
Males
0
1.1 exp
Females
3.33 (1.34,6.87)
7
Dry cleaner
Not reported
0 cases
Laundry worker
Not reported
17 cases
Pukkala et al. (2009)
Launderer and dry cleaner
1.23 (1.02, 1.49)
113
Men and women participating in national census on or before 1990, 5
Nordic countries (Denmark, Finland, Iceland, Norway, Sweden), 30-64
yr, follow-up 2005, occupational title of launderer and dry cleaner in any
census, external referents (SIR), Primary liver cancer (ICD-7, 155)
Male
1.13 (0.76, 1.63)
29
Female
1.27(1.01, 1.57)
84
-------
"^s
a
o
<
s?
cs
>!
o
^3
§•
rs
st
3
>!
sT
St
*3
o
>1
>!
s
Oq
o
>!
5
Si
Si
%•
s
v»
O
Sfc
I
§•
C>
^s
K) On
2
LtJ
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Calvert et al. (In Press)
All subjects
0.13 (0.00,0.73)
1
1,708 U.S. men and women dry-cleaning union member in CA, IL, MI,
Exposure duration/time since 1st employment
0.20 (0,00, 1.01)
1
NY follow-up 1940-2004, multiple solvent exposures (625 subjects
entered union after 1960 and identified as PCE-only exposure), cancer
mortality (SMR), liver and biliary tract (ICD-9, 155, 156)
PCE-only subjects
Not reported
0
Selden and Ahlbora (2011)
Dry-cleaners and laundry workers
1.12(0.73, 1.64)
26
9,440 Swedish men (n = 2,810) and women (« = 9,440) in 461 washing
Males
1.93 (0.97, 3.46)
11
and dry-cleaning establishments, identified by employer in mid-1980s,
employed 1973-1983, follow-up 1985-2000, exposure assigned using
company self-reported information on PCE usage—PCE (dry cleaners and
laundries with a proportion of PCE dry cleaning), laundry (no PCE use),
and other (mixed exposures to PCE, CFCs, TCE, etc.), external referents
Females
0.86 (0.48, 1.41)
15
PCE
1.21 (0.72, 1.92)
18
Males
2.14(0.92,4.21)
8
(SIR), liver and gallbladder (ICD-7, 155)
Duration of employment
<1 yr
(0.00, 9.71)
0
1-4 yr
3.19(0.66, 9.31)
3
5-11yr
2.06 (0.67, 4.80)
5
Females
0.90 (0.43, 1.65)
10
Duration of employment
<1 yr
1.66 (0.20, 6.01)
2
1-4 yr
1.50 (0.49, 3.50)
5
5-11yr
0.46 (0.09, 1.33)
3
Laundry
Males
1.74 (0.36, 5.09)
3
Females
0.67 (0.18, 1.70)
4
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
0
1
a -r
>1
to 0\
2 ^
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Travier et al. (2002)
All subjects, 1960 or 1970 Census in laundry and
dry cleaner or related occupation and industry
1.02 (0.84, 1.24)
105
Swedish men and women identified as laundry worker, dry cleaner, or
presser (occupational title), in the laundry, ironing, or dyeing industry or
related industry in 1960 or 1970 (543,036 person-years); or, as laundry
worker, dry cleaner, or presser (occupational and job title) (46,933 person-
years) in both censuses, follow-up 1971-1989, external referents (SIR),
liver and biliary passages
All subjects in 1960 and 1970 in laundry and dry
cleaner occupation and industry
1.26 (0.73,2.18)
13
Case-Control Studies
5 University Hospitals, United States (AL, FL, MA, NC, PA)
Austin etal. CI987)
Laundry and dry cleaning occupation
Not reported
0
80 histologically confirmed hepatocellular carcinoma cases, 18-84 yr,
years not identified, 161 hospital controls matched on sex, age, race, and
study center, unknown interview methods, exposure surrogate jobs held
>6 mo, OR from logistic regression
France
Ferrand et al. (2008)
Laundry and dry cleaning occupation
Not reported
125 hepatocellular carcinoma in men, lacking HBV and HCV infection,
identified from four hospitals, <75 yr, 2000-2003, 142 hospital controls in
other departments, face-to-face interview, job title >6 mo as exposure
surrogate, OR from logistic regression model and adjust for hospital, age,
and alcohol consumption
Houten and Sonnesso (1980)
Laundry and dry-cleaning operatives
Not reported
2
102 primary liver cancer cases in men and women, identified from
hospital records, 1956-1965, controls were all other hospitalized cancer
cases, self-reported occupation at time of hospitalization, x2 comparing
distribution of job titles
LtJ
o
-------
"^s
a
o
<
s?
cs
>!
o
^3
§•
rs
st
3
>!
sT
St
*3
o
>1
>!
s
Oq
o
>!
5
Si
Si
%•
s
v»
O
Sfc
I
§•
C>
^s
K) On
2
LtJ
On
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Nordic Countries (Denmark, Finland, Norway, Sweden)
Lvnge et al. (2006)
Unexposed
1.0a
58
Case-control study among 46,768 Danish, Finnish, Norwegian, and
Swedish men and women employed in 1960 as laundry worker or dry
cleaner, follow-up 1970-1971 to 1997-2001, 72 incident esophageal
cancer cases, 6 controls per case randomly selected from cohort matched
on country, sex, age, calendar period at diagnosis time, occupational task
at 1970 Census proxy for exposure, RR adjusted for matching criteria
Dry cleaner
0.76 (0.38, 1.52)
95
Other in dry cleaning
0.42 (0.09, 1.89)
22
Unclassifiable
1.11 (0.59,2.09)
121
Duration of employment in dry cleaning
<1 yr
Not reported
2-4 yr
Not reported
3-9 yr
1.21 (0.43, 3.44)
5
>10 yr
0.70 (0.26, 1.92)
5
Unknown
2.88 (0.21,38.81)
1
New Jersey (United States)
Stemhagenet al. (1983)
Laundering, cleaning, and other garment services
2.50(1.02, 6.14)f
10
265 histologically confirmed primary liver cancer cases and deaths,
1975-1980, New Jersey State Cancer Registry, 530 hospital controls
matched on age, race, sex, county of residence, vital status, in-person
interview, job title and industry coded to SIC/SOC, OR estimating using
Mantel-Haenszel with matched case-control set and not adjusted for
personal or lifestyle factors
Laundering, cleaning, and other garment services
2.29 (0.85, 6.13)8
8
Suarez et al. (1989)
Dry-cleaning services
0.98 (0.44, 2.20)
11
1,742 primary liver cancer deaths, 1969-1980, 1,742 dead controls,
frequency matched on age, sex, race, and year death, Texas vital records,
job tile on death certificate, OR from Mantel-Haenszel analyses for race
and sex separately and adjusted for age
Dry-cleaning operators
0.55 (0.17, 1.75)
4
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
0
1
a -r
>1
to 0\
2 ^
Exposure group
Relative risk
(95% CI)
No. obs.
events Reference
Geographic-Based Studies
Taoyuan, Taiwan
Lee et al., 2006
Residence in upstream village
1.0a
Population of two villages surrounding electronic factory
(Chans et al.. 2003; Chans et al.. 2005; Suns et al.. 20071
50 liver cancer deaths, primary, underlying, or underlying
condition as cause of death, 1966-1997, residence as
recorded on death certificate, MOR from logistic regression
adjusted for age and period
Residence in downstream village
2.57 (1.21,5.46)
30
Hausjarvi and Hattula, Finland
(Vartiainen et al.. 19931
Hausjarvi
0.7 (0.3, 1.4)
7
Lymphopoietic cancers, liver cancer and all cancers among
residents with PCE and other solvents in drinking water,
1953-1991, no subject-level exposure information, cancer
rates of Finnish population referent (SIR)
Hattula
0.6 (0.2, 1.3)
6
a Referent.
b SIR or RR for liver, biliary tract, and gallbladder cancers.
0 SIR or RR for hepatocellular carcinoma.
d SIR for gallbladder cancer.
e SIR for all other liver cancers .
f In Stemhagen et al. (19831. odds ratio for primary liver cancer and work in laundering, cleaning and other garment services industry.
g In Stemhagen et al. (19831. odds ratio for hepatocellular carcinoma and work in laundering, cleaning and other garment services industry.
HBV = hepatitis B virus, HCV = hepatitis C virus, ICD = International Classification of Disease, ICDA = International Classification of Disease, Amended,
ISCO = International Standard Classification of Occupation, ISIC = International Standard Industry Classification, JEM = job-exposure-matrix, MOR = mortality
odds ratio, PCE = tetrachloroethylene, TWA = time-weighted-average.
LtJ
^1
o
-------
1
2
3
4
5
6
7
8
9
10
11
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
Risk factors for liver cancer include alcohol and hepatitis B and C viruses, with diabetes
mellitus suggested based on recent epidemiologic studies (El-Serag. 2007). None of the cohort
or case-control studies on liver cancer and tetrachloroethylene controlled for these potential risk
factors.
In conclusion, studies carrying greater weight in the analysis based on a large number of
observed events or exposed cases or a strong exposure-assessment approach, show a mixed
pattern of results. The one case-control study with a large number of exposed liver cancer cases
and a relatively high quality exposure-assessment methodology reported an odds ratio estimate
of 0.76 (95% CI: 0.38, 1.72) for liver cancer and dry cleaning (Lynge et al.. 2006). A recent
multiple Nordic country cohort study and two cohort studies of Swedish subjects with broad
exposure-assessment approaches, and whose subjects overlapped with Lynge et al. (2006).
reported SIRs of 1.02 (95% CI: 0.84, 1.24), 1.22 (95% CI: 1.03, 1.45), and 1.23 (95% CI: 1.02,
1.49) for liver and biliary tract cancer and work as a dry-cleaner or laundry worker (Ji and
Hemminki. 2005: Pukkala et al.. 2009: Travier et al.. 2002). The study of Lindbohm et al.
(2009) of Finnish dry-cleaner and laundry workers whose subjects overlap the larger multiple-
country study of Pukkala et al. (2009) and that carries less weight in the analysis due to fewer
observed liver and biliary cancer cases supports observations in Swedish or the five Nordic
country dry-cleaner and laundry worker studies (Ji and Hemminki. 2005: Pukkala et al.. 2009).
Three other studies with strong exposure-assessment approaches specific to tetrachloroethylene
but whose risk estimates are based on fewer observed liver cancer cases or deaths provide
support for an association between liver cancer and tetrachloroethylene, risk estimates were 1.21
to 2.05 (Boice et al.. 1999: Bond et al.. 1990: Selden and Ahlborg. 2011). However, dry
cleaning or workers employed after 1960 when tetrachloroethylene use was more prevalent did
not have a higher liver cancer risk estimate than laundry workers (Selden and Ahlborg. 2011)
(Lynge et al.. 2006). An exposure-response relationship was not observed, and the SIR for
tetrachloroethylene-exposed subjects with the longest employment duration in Selden and
Ahlborg (2011) was lower than that for shorter employment duration. Potential confounding
may be an alternative explanation as no study adjusted for known and suspected risk factors for
liver cancer (Boice et al.. 1999: Bond et al.. 1990: Ji and Hemminki. 2005: Lynge et al.. 2006:
Pukkala et al.. 2009: Selden and Ahlborg. 2011: Travier et al.. 2002). Nine other cohort and
case-control studies with fewer observed events and/or a broad exposure-assessment
methodology carried less weight in the analysis; these studies also reported a mixed pattern of
results (Blair et al.. 2003: Lynge et al.. 1995: Ruder et al.. 2001: Stemhagen etal.. 1983: Suarez
et al.. 1989: Sung et al.. 2007: Vartiainen et al.. 1993). Lee et al. (2006) reported a risk estimate
of 2.57 (95% CI: 1.21, 5.46) for the association between liver cancer and residence in a village
with groundwater contamination, was in region with a high prevalence of HCV and did not
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-138 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
control for HCV status in the statistical analysis; potential confounding from HCV may be an
alternative explanation for the observed association.
4.3.2. Animal Studies
Liver toxicity and cancer has been observed in laboratory animal studies following
exposure to tetrachloroethylene through multiple routes of exposure. The sections below
describe studies of liver toxicity (see Section 4.3.2.1) and cancer (see Section 4.3.2.2). These
studies are summarized in Tables 4-15 and 4-16, respectively.
4.3.2.1. Liver Toxicity
Tetrachloroethylene causes hepatic toxicity in multiple species, including several strains
of rats and mice. Adverse effects on the liver have been observed in studies of animals exposed
to tetrachloroethylene by multiple routes of exposure, including inhalation and oral gavage.
Hepatic effects observed after subchronic or chronic inhalation exposure to tetrachloroethylene
include increased liver weight (Kiellstrand etal.. 1984; Kyrklund et al.. 1990); hypertrophy
(Odum et al.. 1988b); fatty degeneration (Kylin et al.. 1963; Odum et al.. 1988b); peroxisome
proliferation (Odum et al.. 1988b); other histological changes (Kiellstrand et al.. 1984; NTP.
1986b; Odum et al.. 1988b); and degeneration and necrosis (JISA. 1993; NTP. 1986b). When
administered by oral gavage, tetrachloroethylene also causes hepatic toxicity, including increased
liver enzymes, increased liver weights, histological changes, degeneration and necrosis,
regenerative repair, and polyploidy (Berman et al.. 1995; Buben and O'Flahertv. 1985; Ebrahim
et al.. 1996; Goldsworthy and Popp. 1987; Jonkeretal.. 1996; NCI. 1977; Philip et al.. 2007).
Table 4-15 presents a summary of inhalation and oral rodent liver toxicity studies, which are
briefly described below. This review focuses on studies that identify critical effects commonly
seen in tetrachloroethylene toxicity studies and could, accordingly, support oral and inhalation
reference values. The database of liver toxicity studies is more extensively reviewed in prior
assessments by EPA (1985a). AT SDR (1997b). NYSDOH (1997), and CalEPA (2001).
4.3.2.1.1. Inhalation
Hepatic toxicity was observed in chronic lifetime inhalation bioassays of tetrachloroethylene in
mice conducted by the National Toxicology Program (NTP. 1986b). and the Japan Industrial
Safety Association (JISA. 1993). The NTP study administered tetrachloroethylene to groups of
50 F344 rats of each sex (0, 200, or 400 ppm), or groups of 49 or 50 B6C3Fi mice (0, 100, or
200 ppm) for 6 hours/day 5 days/week for 103 weeks (NTP. 1986b). In addition to liver tumors
in mice of both sexes, liver degeneration was reported in 2/49, 8/49, and 14/50 of males and in
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-139 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-15. Summary of inhalation and oral rodent liver toxicity studies
Species/strain/sex/number
Exposure level/du ration
Effects
Reference
Mouse, B6C3F,. both sexes
mice (groups of 49 or 50
mice of each sex per dose
group, total of -300 mice)
0, 100, 200 ppmfor
104 wk, inhalation
Liver degeneration and necrosis at >100 ppm
in males and at 200 ppm in females
NTP
(1986b)
Mouse, Crj/BDFl mice
(both sexes, 50 animals per
sex per dose group, total of
400 mice)
0, 10, 50, 250 ppmfor
110 wk, inhalation
Focal necrosis in males at >50 ppm;
liver degeneration in males and females at
250 ppm
JISA (1993)
Rat, F344/DuCij (both
sexes, 50 animals per sex
per dose group, total of
400 rats)
0, 50, 200, 600 ppm for
110 wk, inhalation
Spongiosis hepatitis in males at 200 ppm and
higher; hyperplasia in males at 600 ppm
JISA (1993)
Mouse, NMRI, both sexes,
10 per group
0, 9, 37, 75, 150 ppm, 30 d,
inhalation, continuous
(24 h); and 225 (16 h/d),
450 (8 h/d), 900 (4 h/d),
1,800 (2 h/d), or 3,600
(1 h/d), inhalation
Increase in liver weight (>9 ppm);
morphological changes (>9 ppm); increased
plasma butylcholinesterase (>37 ppm)
Kjellstrand
etal. (1984)
Mouse, B6C3Fi, male; and
rat, Sprague-Dawley (both
sexes)
Radiolabeled PCE by
inhalation (10 or 600 ppm
for 6 h), or as a single oral
gavage dose (500 mg/kg)
Irreversible binding to hepatic
macromolecules at all exposures in male
mice and rats
Schumann et
al. (1980)
Rat, F344; and mouse,
B6C3Fi; both sexes
(5 animals per group)
0, 200 ppm (28 d only) and
400 ppm (14, 21, 28 d) for
6 h/d, inhalation
Increased palmitoyl Co A in mice (3.7-fold)
and rats (1.3-fold); increased peroxisome
proliferation in mouse liver in all sex, dose
and time groups; mitochondrial proliferation
in male mice at 400 ppm for 28 d;
increased relative liver weight, centrilobular
lipid accumulation in exposed mice of both
sexes
Odum et al.
(1988b)
Rat, Sprague-Dawley, male
only (8 animals per group)
0 or 320 ppm continuous
for 90 d; 0 or 320 ppm
continuous for 90 d
followed by a 30-d
recovery period, inhalation
Significantly increased relative liver weight
after exposure; this was decreased following
recovery; decreased cholesterol following
the recovery period
Kyrklund et
al. (1990)
Mice, albino (strain not
specified), female only (20
mice per group, 240 total)
0 or 200 ppm
4 h/d, 6 d/w for 1, 2, 4 or 8
wk, inhalation
Fatty degeneration after 1 wk; incidence
severity increased with longer exposure
Kylin et al.
(1965)
Mouse, Swiss-Cox, male
(4-6 mice per 1,500 and
2,000 mg/kg-day doses;
other doses, 12-15
mice/group)
0, 20, 100, 200, 500, 1,000,
1,500, 2,000 mg/kg-day for
6 wk, gavage
Increased liver/body weight ratio at
100 mg/kg-day; increased triglycerides at
100 mg/kg-day; no change at 20 mg/kg-day
Buben and
O'Flaherty
(1985)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-140 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-15. Summary of inhalation and oral rodent liver toxicity studies
(continued)
Species/strain/sex/number
Exposure level/du ration
Effects
Reference
Mouse, Swiss-Webster,
male (4/group)
0, 150, 500, and
1,000 mg/kg-day, aqueous
gavage for 30 d
Increased plasma ALT 24 hours to 14 d after
initial exposure; mild to moderate fatty
degeneration and necrosis, with focal
inflammatory cell infiltration; increased
mitotic figures and DNA synthesis (peaked
on 7 d, sustained at 14-30 d at all doses);
inhibition of PCE metabolism and TCA
production; no change in CYP2E1; CYP4A
increased at 7 but not 14 d, only at
1,000 mg/kg-day
Philip et al.
(2007)
Rat, Wistar, female only
(10 rats in each control
group; 5 rats in each
treatment group)
0, 600, and 2,400
mg/kg-day for 32 d, corn
oil gavage; alone or in
combination with other
compounds
(trichloroethylene,
hexachloro-l,2-butadiene,
1,1,2-trichloro-
3,3,3 -trifluoropropene)
Relative liver weight increases in animals
exposed to PCE alone or in combination;
hepatotoxicity at 600 mg/kg
Jonker et al.
(1996)
F344 rats (male only,
5/group) and B6C3Fi mice
(male only, 5/group)
0 or 1,000 mg/kg-day for
10 d, corn oil gavage
Increased relative liver weight in rats and
mice; 4.3-fold PCO increase in mice; modest
but not significant (1.4-fold) PCO increase
in rats
Goldsworthy
and Popp
(1987)
Mouse, Swiss, both sexes;
6 groups of 6 mice each
(Ebrahim et al.. 1996); male
only; 8 groups of 6 mice
each (Ebrahim et al.. 2001)
0 or 3,000 mg/kg-day for
15 d, sesame oil gavage
Significant increase in liver weight;
degeneration and necrosis of hepatocytes;
decreased blood glucose (glucose effects
mitigated by coexposures to 2-deoxy-
D slucose and vitamin E) (Ebrahim et al..
1996);
Decreased membrane-bound
Na K -ATPascs and Mg2+-ATPases activity
but increased Ca-ATPase activity; mitigated
by coexposure to 2-deoxy-D-glucose and
vitamin E, and taurine
Ebrahim et
al. (1996;
2001)
Rat, F344 female only
(8 rats per group)
0, 50, 150, 500, or 1,500
mg/kg-day, gavage, either
once or for 14 consecutive
days
Increased relative liver weight, elevated
ALT and hepatocellular hypertrophy at
1,500 mg/kg-day
Berman et
al. (1995)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-141 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-16. Incidence of hepatic tumors in rodents exposed to
tetrachloroethylene
Bioassay
Administered
dose/exposure
Continuous
equivalent
exposure
Sex
Hepatocellular
adenomas and
carcinomas
Hemangiomas or
hem angiosarcomas11
NCI (1977)
B6C3Fi miceb
Gavage:
5 d/wk,
78 wk
Vehicle
450 mg/kg-day
900 mg/kg-day
0
332 mg/kg-day
663 mg/kg-day
Male
2/20 (10)
32/48 (67)
27/45 (60)
None reported0
Vehicle
300 mg/kg-daya
600 mg/kg-day
0
239 mg/kg-day
478 mg/kg-day
Female
0/20 (0)
19/48 (40)
19/48 (40)
None reported
NCI (1977)d
Osborne-Mendel rats
Gavage:
5 d/wk,
78 wk
Vehicle
500 mg/kg-day
1,000 mg/kg-day
0
471 mg/kg-day
941 mg/kg-day
Male
1/20 (0)
1/49 (0)
0/50 (0)
None reported
Vehicle
500 mg/kg-day
1,000 mg/kg-day
0
474 mg/kg-day
974 mg/kg-day
Female
None reported
None reported
NTP (1986b)
B6C3Fi mice
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
100 ppm
200 ppm
0
18 ppm
36 ppm
Male
17/49 (35)
31/49 (70)
41/50 (82)
1/49 (2)
0/49 (0)
0/50 (0)
0 ppm
100 ppm
200 ppm
0
18 ppm
36 ppm
Female
4/50 (9)
17/42(40)
38/48 (79)
0/48 (0)
3/50 (6)
0/50 (0)
NTP (1986b)
F344/N rats
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
200 ppm
400 ppm
0
36 ppm
72 ppm
Male
0/50 (0)
1/50 (2)
1/50 (2)
0/50
0/50
0/50
0 ppm
200 ppm
400 ppm
0
36 ppm
72 ppm
Female
0/50
0/50
0/50
0/50
0/50
0/50
JISA (1993)
Cij:BDFl mice
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
10 ppm
50 ppm
250 ppm
0
1.8 ppm
9.0 ppm
45 ppm
Male
13/50 (28)
21/50 (43)
19/50 (40)
40/50 (82)
4/50 (4)
2/50 (2)
7/50 (13)
11/50 (18)
0 ppm
10 ppm
50 ppm
250 ppm
0
1.8 ppm
9.0 ppm
45 ppm
Female
3/50 (6)
3/47 (6)
7/49 (15)
33/49 (67)
1/50
0/47
2/49
3/49
JISA (1993)
F344/DuCij rats
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
50 ppm
200 ppm
600 ppm
0
9 ppm
36 ppm
108 ppm
Male
4/50
0/50
1/50
2/50
0/50
0/50
0/50
0/50
0 ppm
50 ppm
200 ppm
600 ppm
0
9 ppm
36 ppm
108 ppm
Female
1/50 (2)
0/50 (0)
1/50 (2)
0/50 (0)
1/50
0/50
0/50
0/50
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-142 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Table 4-16. Incidence of hepatic tumors in rodents exposed to
tetrachloroethylene (continued)
a These tumors were reported as hemangioendotheliomas in the JISA (1993) report. The term has been updated to
hemangioma (benign) or hemangiosarcoma (malignant). Note that these incidences do not match those tabulated
in Table 12 of the JISA report summary. The incidences reported here represent a tabulation of
hemangioendotheliomas from the individual animal data provided in the JISA report.
b Administered gavage doses listed were increased after 11 wk by 100 mg/kg-day in each low-dose group or by
200 mg/kg-day in each high-dose group. Mice received the listed TWA daily doses through Week 78, and
surviving mice were observed up to study termination in Week 90.
0 None reported: Individual animal data were not available, and summary data did not include a line item for this
tumor type.
d Gavage doses listed were adjusted several times during the course of the study. Male rats received the listed TWA
daily doses through Week 78, and surviving animals were observed up to study termination in Week 110.
1/49, 2/50, and 13/50 of females. Degeneration was characterized by a variety of histological
features, including cytoplasmic vacuolation, hepatocellular necrosis, inflammatory cell
infiltrates, pigment in cells, oval cell hyperplasia, and regenerative foci. Liver necrosis was seen
at increased incidence in dosed males (1/49, 6/49, and 15/50) and in females at 400 ppm (3/48,
5/50, and 9/50). Nuclear inclusions increased in male mice (2/49, 5/49, and 9/50). No dose-
related liver effects were reported in the rats.
In the 13-week NTP study, groups of ten rats and mice of each sex were exposed to air
containing tetrachloroethylene for 6 hours/day, 5 days/week for 13 weeks (0, 100, 200, 400, 800,
or 1,600 ppm). Some rats in the high-dose group died before the end of the studies (4/10 male,
7/10 female). In mice, 2/10 males and 4/10 females in the high-dose group died before the end
of studies. Tetrachloroethylene (200 ppm and above) increased the incidence of hepatic
congestion in male and female rats. In mice of both sexes, liver lesions (leukocytic infiltration,
centrilobular necrosis, and bile stasis) were observed at 400, 800, or 1,600 ppm. Mitotic
alterations were increased at 200 ppm in male mice. No hepatic effects were reported in the
single exposure or 14-day studies.
In the Japan Industrial Safety Association (1993) study (some results reported in Nagano
et al.. 1998). male and female Cij/BDFl mice were exposed to 0-, 10-, 50-, and 250-ppm
tetrachloroethylene for 104 weeks and sacrificed at 110 weeks. In addition to hepatocellular
carcinomas and adenomas in the mice, telangiectasis (vascular lesions formed by dilation of a
group of small blood vessels) and focal necrosis occurred in males at 50 ppm and above. Liver
degeneration was observed at 250 ppm in both sexes. Hemangiomas or hemangiosarcomas,
occurring primarily in the liver or spleen, were also reported in the male mice. This study also
examined effects in F344/DuCrj rats exposed to 0, 50, 200, and 600 ppm for 104 weeks and
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-143 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
sacrificed at 110 weeks. Male, but not female, rats had excess incidence of spongiosis hepatitis
at 200 ppm and 600 ppm.
The lowest reported level for liver effects by inhalation in laboratory animals is in female
NMRI mice exposed for 30 days at 9 ppm (61 mg/m3; Kiellstrand etal.. 1984). Significant
increases in liver weight as well as changes in liver morphology were observed in male and
female mice exposed continuously to 9 ppm and higher concentrations of tetrachloroethylene for
30 days. Livers were enlarged and vacuolization was evident. Reversible increases in levels of
the blood plasma enzyme butyrylcholinesterase were reported at all tetrachloroethylene
concentration levels at or above 37 ppm. The toxicological significance of the increased serum
cholinesterase is uncertain, and this effect of tetrachloroethylene has not been reported by other
investigators. After a recovery period, liver weight was still slightly elevated at 120 days after
cessation of tetrachloroethylene exposure for 30 days at 150 ppm. Total dose administered in the
continuous exposure experiment is not directly comparable to exposures during intermittent and
pulsed exposure experiments, which also found increased liver weight and increased serum
cholinesterase.
Schumann et al. (1980) administered radiolabeled tetrachloroethylene to male B6C3Fi
mice or Sprague-Dawley rats via inhalation (10 or 600 ppm for 6 hours). In mice, the percentage
metabolized based on recovery of the radiolabeled material was determined to be 88% for a
6-hour inhalation exposure of 10 ppm (as compared to only 17% for a single oral gavage dose of
500 mg/kg). At all dose levels in both rats and mice, irreversible binding of radioactivity to
hepatic macromolecules was observed. DNA binding was not seen. In mice, binding peaked at
the termination of the 6-hour inhalation exposure and 6 hours after the single oral dose. In
contrast, binding in the rat peaked 24 hours after either oral or inhalation exposure.
Odum et al. (1988b) exposed groups of male and female F344 rats and B6C3Fi mice by
inhalation for 6 hours/day to 200 ppm (28 days only) or 400 ppm (for 14, 21, or 28 days)
tetrachloroethylene. Five animals per group were exposed. In both sexes, hepatic palmitoyl
coenzyme A (PCO) activity was increased in mice (up to 3.6-fold) and, to a lesser extent, in rats
(up to 1.3-fold). Modest PCO increases were also seen in the kidney of male rats at 200 ppm at
28 days (1.3-fold) but not 400 ppm at 14, 21, or 28 days. In female rat kidney, PCO was
elevated (approximately 1.6-fold) at all doses and times. However, peroxisome proliferation was
not seen in rat kidney upon microscopy. In contrast, hepatic peroxisome proliferation was noted
in all mouse liver for all sexes, times and dose groups on electron microscopy, and the
percentage of cytoplasm occupied by peroxisomes also increased. Catalase, another peroxisomal
enzyme, was unaffected by tetrachloroethylene; male mice exposed at 400 ppm showed the only
moderate (1.4-fold) increase. Mitochondrial proliferation was seen at 28 days in 400-ppm male
mice. In addition, a time-dependent proliferation of smooth endoplasmic reticulum in the liver
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-144 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
of both sexes correlated well with centrilobular hypertrophy. Tetrachloroethylene caused
centrilobular lipid accumulation in male and female mice. Relative liver weight was increased in
mice of both sexes.
Kyrklund et al. (1990) exposed male Sprague-Dawley rats to 320-ppm
tetrachloroethylene continuously for 90 days, followed by a 30-day recovery period. Relative
liver weight was significantly increased in rats examined at the end of the exposure period. A
slight increase in relative liver weight was also observed in the recovered, solvent-treated group.
Cholesterol was also decreased, but this effect was only statistically significant in the
tetrachloroethylene-exposed group that also included a recovery period.
Kylin et al. (1965) exposed female albino mice (strain not specified) 200-ppm
tetrachloroethylene for four hours daily, 6 days a week for 1, 2, 4, or 8 weeks. Hepatic effects
were evaluated by histological examination and determination of extractable liver fat. The
incidence and severity of fatty degeneration increased with longer exposure. Neither liver cell
necrosis nor cirrhosis was observed.
4.3.2.1.2. Oral
In addition to studying the effects of inhalation and a single oral gavage dose (500
mg/kg), as described above, Schumann et al. (1980) also administered 100, 250, 500, or
1,000 mg/kg to male B6C3Fi mice or Sprague-Dawley rats as a daily oral dose for 11 days. At
all doses in mice, histopathological evidence of hepatocellular swelling in the centrilobular
region, a decrease in liver DNA content, and an increase in DNA synthesis was observed. At
>250 mg/kg, tetrachloroethylene increased the absolute or relative liver weights in mice. In rats,
no statistically significant treatment-related effects were seen at 100, 250, or 500 mg/kg;
however, increased liver DNA synthesis was seen in one rat in the 250 mg/kg-dose group,
resulting in a large variation in liver DNA synthesis at that exposure level.
Buben and O'Flaherty (1985) exposed male Swiss-Cox mice to tetrachloroethylene doses
of 0, 20, 100, 200, 500, 1,000, 1,500, or 2,000 mg/kg-day,5 days/week, for 6 weeks. Liver/body-
weight ratios and liver triglycerides were significantly increased at 100 mg/kg-day or more.
Enlarged hepatocytes, karyorrhexis (disintegration of the nucleus), necrosis, polyploidy in the
centrilobular region, and lipid accumulation was evident upon histopathological examination of
mice exposed to 200 or 1,000 mg/kg. Other indices of tetrachloroethylene hepatotoxicity
(decreased glucose-6-phosphatase activity, and increased serum glutamic pyruvic transaminase
activity) were significantly increased at 500 or more mg/kg-day. The liver response (percentage
increase in either liver weight/body weight ratios or G6P inhibition) was highly correlated with
the amount of tetrachloroethylene metabolized, and a plot of these measures against total urinary
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-145 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
metabolites was linear (r2 = 0.97 and 0.98 for increases in liver/body weight and G6P inhibition,
respectively). The LOAEL was 100 mg/kg-day.
Philip et al. (2007) exposed male 6-7 week old Swiss-Webster mice via aqueous gavage
to three dose levels (150, 500, and 1,000 mg/kg-day) for 30 days. At the highest exposure,
mortality was 10% due to apparent CNS toxicity (tremors and ataxia). Significant liver injury
(as assessed by increased plasma ALT) was evident 24 hours after the first, single exposure at all
doses. ALT levels decreased transiently to control levels by 30 days thereafter. Histopathology
was consistent with mild to moderate fatty degeneration and necrosis. Necrotic hepatocytes had
either pyknotic, karyorrhectic or karyolitic nuclei. Infiltration of neutrophils and macrophages
was present near necrotic foci. Regenerative repair was evident in the two higher dose groups by
30 days of exposure, with observed increases in mitotic figures, tritiated thymidine incorporation
with pulse-labeling, and PCNA immunostaining. At the two higher dose groups, a robust
increase DNA synthesis peaked on 7 days, was sustained at 14 days, and had returned to control
levels at 30 days of exposure. The amount of blood and liver TCA declined, while
tetrachloroethylene levels increased, from 1 to 30 days. This is consistent with an inhibition of
tetrachloroethylene metabolism. Because CYP2E1 levels and activity were unchanged, a
different CYP isoform is suggested to be critical for tetrachloroethylene metabolism. The study
found a transient increase in hepatic CYP4A expression, a marker of PPARa induction, which
was evident at 7 but not 14 days at the highest dose. This finding suggests that peroxisome
proliferation is not a sustained response in spite of continued tetrachloroethylene exposure.
In a study by Jonker et al. (1996). hepatotoxicity was observed in female Wistar rats
administered tetrachloroethylene (600 or 2,400 mg/kg-day) daily via corn oil oral gavage for
32 days. Relative liver weight was increased on exposure to tetrachloroethylene alone and in
combination with other hepatotoxicants (trichloroethylene, hexachloro-l,2-butadiene, and
l,l,2-trichloro-3,3,3-trifluoropropene). One high-dose animal died as a result of
tetrachloroethylene treatment, and one animal exposed to the high-dose combination also died as
a result of treatment. Hepatotoxic effects were noted at 600 mg/kg.
Goldsworthy and Popp (1987) administered tetrachloroethylene (1,000 mg/kg-day) by
corn oil gavage to 5 male F344 rats and 5 male B6C3Fi mice for 10 days. In
tetrachloroethylene-exposed rats, cyanide-insensitive palmitoyl CoA oxidation (PCO) was
modestly although not significantly elevated in the liver (1.4-fold increase) and kidney (1.7-fold
increase). In mice, tetrachloroethylene exposure increased PCO activity 4.3-fold in liver and by
2.3-fold in kidney. Relative liver weight was increased in rats and mice with tetrachloroethylene
exposure, but relative kidney weight was unaffected. A comparison of corn oil with methyl
cellulose revealed no effect of the gavage vehicle on tetrachloroethylene-induced PCO. A less-
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-146 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
than-additive effect of trichloroethylene (1,000 mg/kg) administered together was
tetrachloroethylene on PCO induction was seen.
Ebrahim et al. (1996) orally administered 3,000 mg/kg-day tetrachloroethylene in sesame
oil to male and female Swiss mice for 15 days and observed a significant increase in liver weight
and degeneration and necrosis of hepatocytes. These changes occurred simultaneously with a
decrease in blood glucose; elevated activities of enzymes hexokinase, aldolase, and
phosphoglucoisomerase; and decreased activities of gluconeogenic enzymes. Blood glucose
levels were significantly decreased, and this effect was mitigated by concomitant exposure to
2-deoxy-D-glucose (2DG) and vitamin E. A follow-up study by this group further examined the
potential protective properties of 2DG and vitamin E as well as taurine against
tetrachloroethylene-induced membrane damage (Ebrahim et al.. 2001). This study exposed male
albino Swiss mice to the same doses used in the previous study with the addition of a taurine
exposed group (tetrachloroethylene in sesame oil 3,000 mg/kg-day for 15 days orally by
intubation; tetrachloroethylene plus 2DG 500 mg/kg-day by i.p. injection once a day for 15 days;
tetrachloroethylene plus vitamin E 400 mg/kg-day by oral intubation once a day for 15 days; and
tetrachloroethylene plus taurine 100 mg/kg-day by oral intubation once a day for 15 days).
Compared to control cells in the liver, membrane bound Na+K+-ATPases and Mg2+-ATPases
activity was significantly decreased (p < 0.001), while Ca-ATPases activity was increased
(p < 0.001), following exposure to tetrachloroethylene alone. These levels remained near normal
in the animals exposed to tetrachloroethylene along with 2DG, vitamin E or taurine. This return
to normal levels following exposure to vitamin E and taurine may be due to their antioxidant
abilities, and reduced oxidative stress in exposed cells.
Berman et al. (1995) reported liver and kidney toxicity in a study of female F344 rats
exposed for 14 days by oral gavage to 0, 50, 150, 500, or 1,500 mg/kg-day tetrachloroethylene.
The reported LOAEL was 1,500 mg/kg-day. Hepatic effects included increased relative liver
weight, elevated ALT and hepatocellular hypertrophy.
4.3.2.1.3. Intraperitoneal injection
Binding of radiolabeled tetrachloroethylene to hepatic DNA was observed in mice
following i.p. injection (Mazzullo et al.. 1987) but not inhalation and oral exposure (Schumann et
al.. 1980. described above). Using a reportedly more sensitive assay, low levels of DNA binding
was observed in vivo in BALB/C mouse liver 22 hours after i.p. injection (1.4 mg/kg bw), with
10-fold lower levels observed in Wistar rat liver than mouse liver (Mazzullo et al.. 1987). Still
lower levels of DNA binding were observed in the kidney and stomach of mice and rats in this
study. Binding to RNA and protein was always higher than binding to DNA in both mice and
rats. Binding to calf thymus DNA in an in vitro study increased in the presence of microsomal
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-147 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
fractions from both mouse and rat liver, but not kidney, lung or stomach. Cytosolic fractions
from rat and mouse liver, kidney, lung and stomach all induced binding of tetrachloroethylene to
calf thymus DNA, with enzymes from both mouse and rat livers and mouse lung being the most
efficient. DNA binding in the presence of both cytosolic and microsomal fractions was similar
to cytosolic fraction alone. Phenobarbital pretreatment of animals increased cytosol-mediated
binding, but had only a slight effect on microsomal-mediated binding. Binding in the presence
of rat liver microsomal fraction was also increased (17-fold) with addition of GSH, but decreased
in the presence of superoxide dismutase or mannitol (Mazzullo etal.. 1987).
4.3.2.2. Liver Cancer
In carcinogenicity bioassays, tetrachloroethylene caused a statistically significant
increase in the incidence of hepatocellular carcinomas in both sexes of B6C3Fi mice following
either oral gavage administration or inhalation exposure (NCI. 1977; NTP. 1986b). Both sexes
of Crj:BDFl mice have also been shown to develop an increased incidence of hepatocellular
carcinomas when exposed to tetrachloroethylene by inhalation (JISA. 1993; Nagano et al.. 1998).
Additionally, in male Crj:BDFl mice, hemangiosarcomas (reported as malignant
hemangioendotheliomas) in the liver and both hemangiosarcomas and combined
hemangiosarcomas and hemangiomas (reported as benign hemangioendotheliomas) of the spleen
were increased. The studies are presented in Table 4-16 and are briefly summarized here.
4.3.2.2.1. Inhalation
The NTP (1986b) inhalation bioassay exposed groups of 50 B6C3Fi mice of each sex to
(epichlorohydrin free) tetrachloroethylene concentrations of 0, 100, or 200 ppm, 6 hours/day,
5 days/week, for 103 weeks. Tetrachloroethylene caused statistically significant dose-related
increases in the incidences of hepatocellular carcinoma and in combined hepatocellular adenoma
and carcinoma in both sexes. Hepatocellular neoplasms (adenomas and carcinomas combined)
were reported in 17/49, 31/49, and 41/50 males, and 4/48, 17/50, and 38/50 females. In male
mice, hepatocellular carcinomas metastasized to the lungs in 2/49, 7/49, and 1/50 animals.
Metastatic hepatocellular carcinomas were found in the lungs of 0/48, 2/50, and 7/50 female
mice.
A Japan bioassay exposed groups of 50 Crj:BDFl mice of each sex to 0-, 10-, 50-, and
250-ppm tetrachloroethylene, 6 hours/day, 5 days/week, for 104 weeks, and the terminal
sacrifice was performed at 110 weeks. Both males and females showed dose-related increased
incidences of liver carcinomas and combined liver adenomas and carcinomas. The incidence of
hepatocellular adenomas was 7/50, 13/50, 8/50, and 26/50 in males and 3/50, 3/47, 7/49, and
26/49 in females in control, 10-, 50-, and 250-ppm dose groups, respectively. Male
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-148 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
hepatocellular carcinomas also increased, with reported incidences of 7/50, 8/50, 12/50, and
25/50 in males and 0/50, 0/47, 0/49, and 14/49 in females in control, 10-, 50-, and 250-ppm dose
groups, respectively. Liver hemangiosarcomas (reported as malignant hemangioendotheliomas)
were also increased in males. In the spleen, both hemangiosarcomas and combined
hemangiosarcomas and hemangiomas (reported as benign hemangioendotheliomas) were
increased in males.
4.3.2.2.2. Oral
In the NCI (1977) tetrachloroethylene mouse gavage study, groups of 50 male mice
received TWA doses of 536- or 1,072-mg/kg tetrachloroethylene in corn oil by intragastric
gavage for 78 weeks (450 or 900 mg/kg for 11 weeks, then 550 or 1,100 mg/kg for 67 weeks).
Groups of 50 female mice received TWA doses of 386 or 772 mg/kg of tetrachloroethylene in
corn oil by gavage (300 or 600 mg/kg for 11 weeks, then 400 or 800 mg/kg for 67 weeks). Mice
were dosed 5 days/week. The tetrachloroethylene used in the study was greater than 99% pure,
but impurities were not identified (NCI. 1977; U.S. EPA. 1985b). The test sample was estimated
to contain epichlorohydrin concentrations of less than 500 ppm (U.S. EPA. 1985b). It was
considered unlikely, however, that the tumor response resulted from this low concentration of
epichlorohydrin. Tetrachloroethylene caused statistically significant increases (p < 0.001) in the
incidences of hepatocellular carcinoma in both sexes of mice in both treatment groups when
compared with untreated controls or vehicle controls. The time to tumor was significantly
decreased in treated mice.
4.3.3. Summary of Liver Effects in Humans and Animals
Two of four studies of occupationally exposed dry cleaners showed indications of liver
toxicity, namely sonographic changes of the liver and altered serum concentrations of liver
enzymes indicative of liver injury. Frank liver disease was not seen among these workers for a
number of possible reasons: individuals with frank liver disease may not have been included in
cross-sectional studies because they had left the workforce due to their conditions, the healthy
worker effect, and other selection biases. LOAELs in these human studies were between 12 and
16 ppm (TWA).
Liver toxicity has been reported in multiple animal species by inhalation and oral
exposures to tetrachloroethylene. The effects are characterized by increased liver weight, fatty
changes, necrosis, inflammatory cell infiltration, triglyceride increases, and proliferation. The
mouse has been shown to be more sensitive to hepatic toxicity than the rat in multiple subchronic
and chronic studies (e.g., JISA. 1993; NCI. 1977; NTP. 1986b; Schumann et al.. 1980). After
subchronic or chronic inhalation exposures in mice, liver toxicity is manifested by increased liver
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-149 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
weight (Kiellstrand et al.. 1984). liver enlargement (Kiellstrand etal.. 1984; Odum et al.. 1988b).
cytoplasmic vacuolation (fatty changes) (Kiellstrand et al.. 1984; NTP. 1986b; Odum et al..
1988b). centrolobular hepatocellular necrosis (JISA. 1993; NTP. 1986b). and inflammatory cell
infiltrates, pigment in cells, oval cell hyperplasia, and regenerative foci (NTP. 1986b). The
LOAEL for the inhalation studies, 9 ppm, is from a 30-day-exposure mouse study reporting
increased liver weight and morphological changes, and is supported by a finding of irreversible
macromolecular binding in mouse liver following a single, 6-hour exposure at 10 ppm. The
JISA (1993) chronic mouse inhalation bioassay reported liver necrotic foci at 50 ppm and higher.
In two lifetime inhalation cancer bioassays, increases in liver cancer occurred at 100 ppm and
above, and there was a significant dose-response trend in both studies.
With administration by oral gavage in mice, liver toxicity (increased liver weight,
hepatocellular swelling, necrosis, lipid accumulation, and increased DNA synthesis) has been
observed at 100 mg/kg-day (Buben and O'Flahertv. 1985; Schumann etal.. 1980) and above
(Berman et al.. 1995; Ebrahim et al.. 1996; Goldsworthy andPopp. 1987; Jonker et al.. 1996).
At 150 mg/kg-day administered for 30 days (Philip et al.. 2007). tetrachloroethylene increased
ALT levels transiently and stimulated fatty degeneration and necrosis, with ensuing regenerative
repair. These findings support a LOAEL of 100 mg/kg-day and a NOAEL of 20 mg/kg-day.
For liver cancer, epidemiologic studies carrying greater weight in the analysis, based on a
large number of observed events or exposed cases,or a strong exposure-assessment approach,
show a mixed pattern of results. The one case-control study with a large number of exposed
liver cancer cases and a relatively high quality exposure-assessment methodology reported an
odds ratio estimate of 0.76 (95% CI: 0.38, 1.72) for liver cancer and dry cleaning. A recent
multiple Nordic country cohort study and two cohort study of Swedish subjects with broad
exposure-assessment approaches and whose subjects overlapped with Lynge et al. (2006)
reported SIRs of 1.02, 1.22, and 1.23 for liver and biliary tract cancer and work as a dry cleaner
or laundry worker in Travier et al. (2002). Ji and Hemminki (2005c) and Pukkala et al. (2009).
respectively. Three other studies with strong exposure-assessment approaches specific to
tetrachloroethylene but whose risk estimates are based on fewer observed liver cancer cases or
deaths reported risk estimates of 1.21 to 2.05 for the association between liver cancer and
tetrachloroethylene, risk estimates were (Boice et al.. 1999; Bond et al.. 1990; S el den and
Ahlborg. 2011). However, dry cleaning or workers with employed after 1960 when
tetrachloroethylene use was more prevalent did not have higher liver cancer risk estimate than
laundry workers (Lynge et al.. 2006; Selden and Ahlborg. 2011). Exposure response was not
observed and the SIR for tetrachloroethylene-exposed subject with longest employment duration
in Selden and Ahlborg (2011) was lower than that for subjects with shorter employment
duration. Potential confounding may be an alternative explanation as no study adjusted for
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-150 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
known and suspected risk factors for liver cancer (Boice et al.. 1999; Bond et al.. 1990; Ji and
Hemminki. 2005c; Lynge et al.. 2006; Pukkala et al.. 2009; Selden and Ahlborg. 2011; Travier et
al.. 2002). Nine other cohort and case-control studies with fewer observed events and broad
exposure-assessment methodology carried less weight in the analysis and reported a pattern of
mixed results (Blair et al.. 2003; Calvert et al.. In Press; Lindbohm et al.. 2009; Lynge et al..
1995; Stemhagen et al.. 1983; Suarez et al.. 1989; Sung et al.. 2007; Vartiainen et al.. 1993)(Lee
et al., 2006). Lee et al. (2006) reported a risk estimate of 2.57 (95% CI: 1.21, 5.46) for the
association between liver cancer and residence in a village with groundwater contamination, but
subjects were from a region with a high prevalence of HCV infection and HCV status may
confound the observed association.
Tetrachloroethylene caused a statistically significant increase in the incidence of liver
tumors in both sexes of mice in multiple carcinogenicity bioassays. A statistically significant
increase in the incidence of hepatocellular carcinomas in both sexes of B6C3Fi mice was seen
following either oral gavage administration or inhalation exposure (NCI. 1977; NTP. 1986b).
Both sexes of Cij:BDFl mice also showed an increased incidence of hepatocellular carcinomas
and adenomas when exposed to tetrachloroethylene by inhalation (JISA. 1993; Nagano et al..
1998). Liver hemangiosarcomas were also increased in males. In the spleen, both
hemangiosarcomas and combined hemangiosarcomas and hemangiomas were increased in
males.
4.3.4. Mode of Action for Hemangiosarcomas or Hemangiomas in Mice
The incidence of hemangiomas or hemangiosarcomas occurring in the liver or spleen
(and to a lesser extent in fat, subcutaneous skin, and the heart) was significantly increased in
male Cij :BDF1 mice exposed to tetrachloroethylene by inhalation (JISA. 1993). This tumor type
is distinct from the hepatocellular adenomas and carcinomas induced by tetrachloroethylene in
male and female Crj:BDFl mice by inhalation exposure (JISA. 1993). and in male and female
B6C3Fi mice by inhalation (NTP. 1986b) or oral (NCI. 1977) exposure. No data are available
concerning either the metabolites or the mechanisms that may contribute to the induction of
hemangiosarcomas or hemangiomas occurring in the liver or spleen in male mice. It is
concluded that the mechanisms or modes of action by which tetrachloroethylene induces this
type of tumor is not known.
4.3.5. Mode of Action for Murine Hepatocellular Tumors
Multiple metabolites formed from tetrachloroethylene are toxic and carcinogenic in the
liver. In particular, it is likely that TCA and DCA, which are hepatocarcinogens in mice,
contribute to tetrachloroethylene-induced liver tumors. However, the mode of action through
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-151 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
which these (and potentially other) metabolites elicit the benign and malignant hepatocellular
tumors induced with oral or inhalation exposure to tetrachloroethylene in multiple strains and
both sexes of mice remains to be fully elucidated. As noted by NRC (2010). it is likely that key
events from several pathways, comprising several simultaneous mechanisms, operate in
tetrachloroethylene-induced liver cancer.
The discussion of mechanistic effects addresses the following topics: (1) contribution of
tetrachloroethylene metabolism to hepatocarcinogenicity (see Section 4.3.5.1); (2) genotoxicity
(see Section 4.3.5.2); (3) epigenetic effects, focusing on DNA hypomethylation (see
Section 4.3.5.3); (4) oxidative stress (see Section 4.3.5.4); and (2) receptor activation, focusing
on a hypothesized PPARa-activation mode of action (see Section 4.3.5.5). Because it has been
suggested that hepatocarcinogenesis caused through a PPARa-activation MOA is not relevant to
humans (e.g., Klaunig et al.. 2003). and such a conclusion would have significant implications
for hazard conclusions and dose-response analyses, this hypothesized MOA is discussed in
relatively more detail than other topics. In the NRC review of EPA's 2008 external review draft
of tetrachloroethylene, a dissenting opinion was put forth by one member that PPARa mediation
of tetrachloroethylene- induced hepatocarcinogenesis in mice is the plausible predominant MOA
and that this MOA lacks relevance to human hepatocarcinogenesis (see Appendix B. NRC.
2010). However, in their rebuttal (also presented in Appendix B. NRC. 2010). the committee as
a whole did not support these conclusions. Overall, the committee judged that many gaps in
knowledge remain with regard to the MOA of tetrachloroethylene. They stated that the
relevance of the peroxisome proliferator MOA to tetrachloroethylene-induced mouse hepatic
cancer and to tetrachloroethylene-induced human hepatic cancer remains hypothetical and
requires further rigorous testing. Hence, they concluded that it is premature to draw conclusions
on the relevance of the PPARa MOA to tetrachloroethylene-induced human hepatic
carcinogenesis (NRC. 2010). They encouraged an in-depth presentation of the relevant issues
and data, particularly with respect to tetrachloroethylene studies. The discussion below,
especially that in Section 4.3.5.4, follows these recommendations.
4.3.5.1. Contribution of Tetrachloroethylene Metabolites to Mode of Action and
Carcinogenicity
Several metabolites of tetrachloroethylene are carcinogenic in mice, and it is thought that
the hepatocarcinogenicity of the parent compound is mediated through the action of one or more
of its metabolites. Oxidative metabolism is thought to predominate in the liver, and TCA is the
major resultant urinary excretion product. As discussed in Section 3, TCA appears to be formed
from spontaneous decomposition of trichloroacetyl chloride, which is known to bind to
macromolecules. DCA may be formed from dechlorination of TCA, but DC A produced from
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-152 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
this pathway is likely to be rapidly metabolized in the liver and not detected in blood or urine.
DCA that has been detected in urine is thought to be the result of kidney-specific P-lyase
metabolism of the results of GSH conjugation of tetrachloroethylene, and DCA produced from
this pathway is presumed to not play a role in liver toxicity or cancer. The potential role of GST
conjugates of tetrachloroethylene in liver carcinogenicity, although unknown, is presumed to be
less important that the role of oxidative metabolites.
The focus of most hypotheses with respect to contributors to tetrachloroethylene
hepatocarcinogenicity has been on TCA and, to a lesser extent, DCA. Data supporting the
conclusion that TCA and DCA, alone and in combination, are hepatocarcinogenic in rodents is
summarized in Tables 4-17, 4-18, and 4-19. In mice, TCA significantly increased the incidence
of liver tumors in male and female B6C3Fi mice exposed via drinking water for 52-104 weeks
(Bull et al.. 2002; Bull et al.. 1990; Bull et al.. 2004; DeAngelo et al.. 2008; Herren-Freund et al..
1987; Pereira. 1996; Pereira and Phelps. 1996). Incidence of tumors increased with increasing
TCA concentrations (Bull et al.. 2002; Bull et al.. 1990; DeAngelo et al.. 2008; Pereira. 1996).
These results were obtained under conditions where the background incidence of tumors in
control animals was generally low. The development of tumors in animals exposed to TCA
progressed rapidly, as evidenced by significant numbers of tumors in less-than-lifetime studies of
82 weeks or less. Positive evidence for tumor promotion by TCA (following exposure to known
tumor initiators) has been reported for liver tumors in B6C3Fi mice (Pereira et al.. 2001; Pereira
et al.. 1997) and for GGT-positive foci in livers of partially hepatectomized Sprague-Dawley rats
(Parnell et al.. 1988). DCA also causes liver cancer in mice (Bull et al.. 1990; Bull et al.. 2004;
Daniel et al.. 1992; DeAngelo et al.. 1999; Herren-Freund et al.. 1987). DCA and TCA are also
hepatocarcinogenic in mice when coadministered in the drinking water for 52 weeks (Bull et al..
2004). Treatment-related liver tumors were observed in male F344/N rats exposed via drinking
water to DCA (DeAngelo et al.. 1996) but not TCA (DeAngelo etal.. 1997) for 60 or 104 weeks.
The carcinogenicity of TCA and DCA has not been evaluated in female rats or in other species
of experimental animals.
Data on tumor phenotype support the view that TCA may not be the sole tumorigenic metabolite
of tetrachloroethylene, but also do not provide definitive evidence testing any particular
hypothesis. For instance, liver tumor genotypes (e.g., with regard to H-ras codon 61 mutation)
and phenotypes (e.g., with regard to c-Jun staining) appear to differ among tumors induced by
TCA, DCA, the combination of TCA and DCA, and the structurally related compound
trichloroethylene (Bull et al.. 2002). Bull et al. (2002) suggest that for trichloroethylene, the data
are not consistent with the hypothesis that TCA is the sole active moiety, but a similar
experiment has not been conducted for tetrachloroethylene. However, by analogy, it is possible
that TCA and DCA, in combination with each other (and with other reactive intermediates
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-153 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-17. Hepatocarcinogenicity of TCA in rodent drinking water studies
Species (sex)
Exposure
Results
Authors
B6C3Fi mice (M)
0 and 5 g/L in drinking water for 61
wk
Carcinomas: 0/22, 7/22
Herren-Freund et al.
(1987)
B6C3Fi mice (M)
0, 1, and 2 g/L for 52 wk
Carcinomas: 0/35, 2/11, 4/24
Bull et al. ri990,l
B6C3FJ mice (M)
0, 0.05, 0.5, or 5 g/L TCA for 60 wk
Carcinomas: 7, 4, 21, 38%
DeAngelo et al. (2008)
B6C3FJ mice (M)
0, 0.5 and 2 g/L for 52 wk
Carcinomas: 1/20, 11/20, 9/20
Bull et al. (2002)
B6C3Fi mice (F)
0,0.35, 1.2, 3.5 g/L for 51 wk
0,0.35, 1.2, 3.5 g/L for 82 wk
Carcinomas (52 wk): 0/40, 0/40,
0/19, 5/20
Carcinomas (81 wk): 2/90, 0/53,
5/27, 5/18
Pereira et al. (1996)
F344/N rats (M)
0, 0.05, 0.5, 5 g/L for 104 wk
Carcinomas: 0, 0, 0, 0%
DeAneelo et al. (1997)
Adapted from NRC (2006).
Table 4-18. Hepatocarcinogenicity of DCA in rodent drinking water studies
Species (sex)
Exposure
Results
Authors
B6C3Fi mice (M)
0 and 5 g/L for 61 wk
Carcinomas: 0/22, 21/26
Herren-Freund et al.
(1987)
B6C3FJ mice (M)
0 and 2 g/L for 52 wk
Carcinomas: 0/35, 5/24
Bull et al. ri990,l
B6C3FJ mice (M)
0, 0.05, 0.5, 4.5 and 5 g/L for
60-95 wk
Carcinomas: 6.7-10, 22, 38, 98,
55%
DeAneelo et al. (1991)
B6C3Fi mice (M)
0, 0.05 g/L for 60 wk
0,0.5, 1,2, 3.5 g/L for 100 wk
Carcinomas (60 wk): 8/12, 25/30
Carcinomas (100 wk): 5/50, 5/24,
16/32, 6/14, 4/8
DeAneelo et al. (1999)
B6C3Fi mice (M)
0, 0.05 for 60 wk
Carcinomas: 2/20, 15/24
Daniel et al. (1992)
B6C3Fi mice (F)
0, 0.28, 0.93, and 2.8 g/L for
52 wk
0, 0.28, 0.93, and 2.8 g/L for
81 wk
Carcinomas (52 wk): 0/40, 0/40,
0/20, 1/20
Carcinomas (81 wk): 2/90, 0/50,
1/28, 5/19
Pereira et al. (1996)
F344 rats (M)
0, 0.05, 0.5, 2.4 g/L for 60 wk
0, 0.05, 0.5 g/L for 104 wk
Carcinomas (60 wk): 0/7, 0/7, 0/7,
1/27
Carcinomas (104 wk): 0/23, 0/26,
2/29
DeAneelo et al. (1996)
Adapted from NRC (2006).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-154 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Table 4-19. Incidence of mouse liver tumors with drinking water
administration of TCA and DCA, alone and in combination
Species (sex)
Exposure (52 wk)
Liver tumor incidence
Authors
B6C3Fi mice (M)
0 (drinking water vehicle)
1/20
Bull et al. (2004)
0.5 g/L TCA
11/20
2 g/L TCA
9/20
0.1 g/L DCA
2/20
0.5 g/L DCA
5/20
2 g/L DCA
12/19
0.1 g/L DCA + 0.5 g/L TCA
9/20
0.5 g/L DCA + 0.5 g/L TCA
13/19
0.1 g/L DCA+ 2 g/L TCA
15/20
0.5 g/L DCA + 2 g/L TCA
13/20
Adapted from Bull et al. (20041.
produced during the oxidative metabolism of tetrachloroethylene) may contribute to the
production of liver tumors. This appears to be the case for noncancer effects, as the spectrum of
endpoints caused by tetrachloroethylene includes effects broader than that produced by TCA,
and including fatty degeneration, focal necrosis and regenerative repair, some of which may play
a role in liver carcinogenesis (see below).
The hepatocarcinogenic potencies of TCA and tetrachloroethylene have not been directly
compared in a single rodent bioassay. Appendix D presents a comparative quantitative analysis
of the carcinogenicity of TCA (including that predicted using PBPK modeling to be produced
from tetrachloroethylene) with the carcinogenicity of tetrachloroethylene. This analysis suggests
that TCA might explain the incidence of carcinomas observed in the available
tetrachloroethylene bioassays, but that a wide range of possible contributions cannot be ruled out
by the available data. Specifically, a contribution of TCA from as little as 12% up to 100%
cannot be ruled out, under the assumptions that the tetrachloroethylene NTP and JISA bioassay
data can be combined, and using the Chiu and Ginsberg PBPK model for tetrachloroethylene
and the Chiu PBPK model for TCA and TCA bioavailability. If either of these assumptions is
relaxed—i.e., given that residual uncertainties of about twofold exist in the PBPK model
predictions for TCA internal dose and that there may be some underlying differences between
the NTP and JISA bioassays—then the CIs will be greater. Furthermore, the high control tumor
incidence reported in the TCA bioassay of DeAngelo et al. (2008) raises questions as to the
representativeness of that bioassay for comparison with tetrachloroethylene bioassays. Overall,
as discussed in Chiu with regards to the contribution of TCA to TCE-induced hepatomegaly,
factors such as study-to-study experimental variability in kinetics (e.g., metabolism,
bioavailability) or in dynamics (e.g., background tumor rates), different analytical methods used
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-155 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
to quantify TCA in blood and tissues, and uncertainty in TCA dosing patterns in drinking water
studies further limit the ability to discern the quantitative contribution of TCA. A more precise
quantitative measure of the relative contribution of TCA to tetrachloroethylene-induced liver
tumors requires an appropriately designed experiment to better control for these factors.
4.3.5.2. Genotoxicity
A hypothesized mutagenic MOA entails the following key events leading to
tetrachloroethylene-induced liver tumor formation: following metabolism of tetrachloroethylene
to one or more mutagenic intermediates, the genetic material is altered in a manner that permits
changes to be transmitted during cell division through one or more mechanisms (gene mutations,
deletions, translocations, or amplification); the resulting mutations advance acquisition of the
multiple critical traits contributing to carcinogenesis. This MOA may apply to multiple cancer
types.
The genotoxic potential of tetrachloroethylene is addressed in Section 4.8. To
summarize, the results of a large number of in vitro genotoxicity tests in which
tetrachloroethylene was the test agent support the conclusion that tetrachloroethylene does not
exhibit direct mutagenic activity in the absence or presence of the standard S9 fraction (Bartsch
et al.. 1979; Connor et al.. 1985; DeMarini etal.. 1994; Greim et al.. 1975; Hardin et al.. 1981;
Ha worth et al.. 1983; Kringstad et al.. 1981; Milman et al.. 1988; NTP. 1986b; Roldan-Ariona et
al.. 1991; Shimada et al.. 1985; Warner et al.. 1988; Watanabe et al.. 1998). However, the few in
vitro mutagenicity studies of tetrachloroethylene under conditions that would generate the GSH
conjugate were positive ("Vamvakas et al.. 1989b; Vamvakas et al.. 1989c). Several other known
(DCA) and putative (tetrachloroethylene oxide) P450 metabolites also exhibit in vitro
mutagenicity. Studies of chromosomal aberrations following exposure to tetrachloroethylene are
mostly negative, but positive results have been reported from in vitro studies with enhanced
metabolic activation (Dohertv et al.. 1996).
TCA, the primary oxidative metabolite of tetrachloroethylene, exhibits little, if any,
genotoxic activity in vitro. TCA did not induce mutations in S. typhimurium strains in the
absence of metabolic activation or in an alternative protocol using a closed system (DeMarini et
al.. 1994; Giller et al.. 1997; Kargalioglu et al.. 2002; Nelson et al.. 2001; Rapson et al.. 1980;
Waskell. 1978). but a mutagenic response was induced in TA100 in the Ames fluctuation test
(Giller et al.. 1997). 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 (Mackav et al..
1995). Measures of DNA-repair responses in bacterial systems have shown induction of DNA
repair reported in S. typhimurium but not in E. coli. Mutagenicity in mouse lymphoma cells was
only induced at cytotoxic concentrations (Harrington-Brock et al.. 1998). TCA was positive in
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-156 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
some genotoxicity studies in vivo mouse, newt, and chick test systems (Bhunya and Behera.
1987; Bhunya and Jena. 1996; Birner et al.. 1994; Gilleretal.. 1997). DNA unwinding assays
have either shown TCA to be much less potent than DCA (Nelson and Bull 1988) or negative
(Nelson et al.. 1989; Styles et al.. 1991). Due to limitations in the genotoxicity database, the
possible contribution of TCA to tetrachloroethylene genotoxicity is unclear.
The limited in vivo studies of tetrachloroethylene are inconsistent, with only negative
(Bronzetti et al.. 1983; NTP. 1986b) or equivocal (Beliles et al.. 1980; Cederberg et al.. 2010)
genotoxicity assay results demonstrated following inhalation or oral exposure. These include
findings that tetrachloroethylene at higher concentrations induces at most modest increases in
DNA damage and DNA binding in liver tissue (Cederberg et al.. 2010; Murakami and Horikawa.
1995). Intraperitoneal injection assays have demonstrated both negative (NTP. 1986b) as well as
positive results for different genotoxicity endpoints (Walles. 1986). Assays of clastogenic
effects following inhalation exposure in humans have shown inconsistent results, and are
suggested to be related to coexposures (Ikedaetal.. 1980; Seiii et al.. 1990).
Thus, although tetrachloroethylene has largely yielded negative in standard genotoxicity
assays, uncertainties remain with respect to the possibility that genotoxicity contributes to
hepatocarcinogenesis. Not all metabolites have been identified or characterized, but several
known metabolites including those derived from P450 as well as GSH pathways are clearly
mutagenic in the standard battery of tests. Tetrachloroethylene is mutagenic in bacterial assays
in the presence of GST and GSH whereas the standard S9 fraction has typically yielded negative
results. Tetrachloroethylene at higher concentrations also induces modest increases in DNA
damage and DNA binding in liver tissue (Cederberg et al.. 2010; Murakami and Horikawa.
1995). The metabolite DCA is the most potent mutagen of the P450-derived metabolites,
exhibiting mutagenic activity in a number of assays. A putative P450 derived metabolite,
1,1,2,2-tetrachloroethylene oxide, is also mutagenic; the mutagenicity of this epoxide would be
predicted from structure-activity relationships. Given the demonstrated mutagenicity of several
tetrachloroethylene metabolites, the hypothesis that mutagenicity contributes to the MOA for
tetrachloroethylene carcinogenesis cannot be ruled out, although the specific metabolic species
or mechanistic effects are not known.
4.3.5.3. Altered DNA Methylation
Another hypothesis is that tetrachloroethylene induces hepatocarcinogenesis via the
induction of epigenetic changes, particularly DNA methylation. This MOA entails the following
key events leading to tetrachloroethylene-induced liver tumor formation: following metabolism
of tetrachloroethylene to one or more reactive intermediates, particularly TCA, DCA, and other
reactive species, epigenetic changes ensue; the resulting alterations advance acquisition of the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-157 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
multiple critical traits contributing to carcinogenesis. This MOA may apply to multiple cancer
types.
No tetrachloroethylene-specific data are available regarding a role of alteration in DNA
methylation in tumorigenesis. However, experimental evidence supports the hypothesis that
hypomethylation of DNA may be related to the carcinogenicity of TCA and DC A in mice. In
female B6C3Fi mice that received an i.p. injection of A-methyl-A-nitrosourea (MNU) and were
then administered TCA or DCA in drinking water, DNA methylation in the resulting
hepatocellular adenomas and carcinomas was about half that seen in noninvolved tissue from the
same animal or from animals given only MNU (Tao et al.. 1998). Drinking water exposure of
female B6C3Fi mice to TCA or DCA for 11 days also decreased total liver DNA methylation by
60% (Tao et al.. 1998). The same investigators (Tao et al.. 2004) also demonstrated
hypomethylation of a region of the IGF-II gene in liver and tumors from mice initiated with
MNU and subsequently exposed to TCA or DCA. An association between hypomethylation and
cell proliferation in liver of TCA- or DCA-exposed mice was demonstrated by Ge et al. (2001a).
An increase in DNA replication (evidenced by increased proliferating cell nuclear antigen
labeling index and mitotic labeling index) was observed 72 hours and 96 hours after the first
daily gavage dose of either TCA or DCA. Hypomethylation of the internal cytosine of CCGG
sites in the promoter region of the c-myc gene began between 48 and 72 hours from the initiation
of TCA or DCA exposure and continued to 96 hours. These observed effects of TCA and DCA,
together with the fact that methylation changes represent common early molecular event in most
tumors (Bavlin et al.. 1998; Zingg and Jones. 1997). support the plausibility of a hypothesis that
dysregulation of gene methylation plays a role in tetrachloroethylene-induced tumorigenesis.
However, no data are available specifically testing this hypothesis for tetrachloroethylene.
4.3.5.4. Cytotoxicity and Secondary Oxidative Stress
Another hypothesis is that oxidative stress produced secondary to tetrachloroethylene-
induced cytotoxicity plays a critical role in hepatocarcinogenesis. This MOA entails the
following key events leading to tetrachloroethylene-induced liver tumor formation: following
metabolism of tetrachloroethylene to one or more reactive intermediates, toxicity to the liver
ensues; oxidative stress is produced during hepatocyte injury, from infiltrating inflammatory
cells, and/or as part of the intracellular/extracellular repair processes; the resultant oxidative
stress, via a variety of potential mechanisms (damage to and alteration of macromolecules, cell
signaling alterations, etc.), advances acquisition of the multiple critical traits contributing to
carcinogenesis. This MOA may apply to multiple cancer types.
Numerous studies, including chronic bioassays, have demonstrated that
tetrachloroethylene is hepatotoxic. Reported characteristics of the hepatic injury induced by
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-158 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
tetrachloroethylene and the ensuing tissue repair include increased liver weight, fatty changes,
necrosis, inflammatory cell infiltration, triglyceride increases, and proliferation. The NTP
chronic bioassay reported a variety of histological changes, including cytoplasmic vacuolation,
hepatocellular necrosis, inflammatory cell infiltrates, pigment in cells, oval cell hyperplasia, and
regenerative foci. Liver tissue repair is a complex process involving cell division, angiogenesis,
ductulogenesis, cell mobility, and extracellular matrix repair, all in a coordinated manner
(Mehendale. 2005). Reactive oxygen species can play a role in mediating many of these
processes, and are produced during hepatocyte injury, from infiltrating inflammatory cells,
and/or as part of the intracellular/extracellular repair processes.
A limited database of studies is available on tetrachloroethylene-induced hepatic
oxidative stress. Two studies by Ebrahim et al. (1996; 2001) have examined the ability of
2-deoxy-glucose (2DG), vitamin E or taurine to modulate hepatic effects following short-term
exposure. Ebrahim (1996) orally administered 3,000 mg/kg-day tetrachloroethylene in sesame
oil to male and female Swiss mice for 15 days and observed a significant increase in liver weight
and degeneration and necrosis of hepatocytes. These changes occurred simultaneously with a
decrease in blood glucose; elevated activities of enzymes hexokinase, aldolase, and
phosphoglucoisomerase; and decreased activities of gluconeogenic enzymes. Blood glucose
levels were significantly decreased, and this effect was mitigated by concomitant exposure to
2-deoxy-D-glucose and vitamin E.
In a follow-up study, Ebrahim et al. (2001) further examined the potential protective
properties of 2DG and vitamin E as well as taurine against membrane damage induced with a
similar exposure paradigm. This study exposed male albino Swiss mice to the same doses used
in the previous study with the addition of a taurine exposed group (tetrachloroethylene in sesame
oil 3,000 mg/kg-day for 15 days by oral gavage; tetrachloroethylene plus 2DG 500 mg/kg-day by
i.p. injection once a day for 15 days; tetrachloroethylene plus vitamin E 400 mg/kg-day by oral
gavage once a day for 15 days; and tetrachloroethylene plus taurine 100 mg/kg-day by oral
gavage once a day for 15 days). Compared to control cells in the liver, membrane bound
Na+K+-ATPases and Mg2+-ATPases activity was significantly decreased (p < 0.001), while
Ca-ATPases activity was increased (p < 0.001), following exposure to tetrachloroethylene alone.
These levels remained near normal in the animals exposed to tetrachloroethylene along with
2DG, vitamin E or taurine. This return to normal levels following exposure to vitamin E and
taurine may be due to their antioxidant abilities, and reduced oxidative stress in exposed cells.
A recent in vitro investigation examined tetrachloroethylene-induced gene expression
changes in the HepG2 cultured human hepatoma cell line using an Affymetrix platform (Kawata
et al.. 2009). HepG2 cells retain Phase 1 and Phase 2 metabolic enzymes. Tetrachloroethylene
(2 mM) altered the expression of 445 genes, of which 367 were annotated in Gene Ontology
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-159 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
terms to represent 261 biologic processes. The major processes included cell death, regulation of
metabolic processes, phosphorylation, lipid biosynthesis, steroid metabolism, intracellular
transport, DNA repair, and regulation of cell cycle. Based on KEGG pathway mapping, —dd
cycle" and MAPK signaling" pathways comprised were prominent; a similar finding was
reported for other chemicals (dimethyl nitrosamine and the phorbol ester
12-O-tetradecanoylphorbol-13-acetate) and metals (nickel, cadmium and arsenic). The authors
noted that this pathway has been shown to be activated by reactive oxygen species and metals in
earlier studies (Guvton et al.. 1996; Liu et al.. 1996) and demonstrated that metal-induced gene
changes associated with this pathway could be inhibited by vitamin C. Upregulation of the
oncogene PTT1G was noted in all exposures. This hypothesis-generating in vitro experiment
may aid in elucidating molecular pathway-based biomarkers of tetrachloroethylene.
4.3.5.5. Peroxisome Proliferator-Activated Receptor (PPAR) Activation Mode of Action
4.3.5.5.1. Description of hypothesized MOA
Another hypothesis is that tetrachloroethylene acts by a PPARa-agonism MOA in
inducing mouse hepatocarcinogenesis. According to this hypothesis, the key events leading to
tetrachloroethylene-induced liver tumor formation constitute the following: tetrachloroethylene
metabolites (primarily the oxidative metabolite, TCA), after being produced in the liver,
activates the PPARa receptor, which then causes alterations in cell proliferation and apoptosis,
followed by clonal expansion of initiated cells. This MOA is assumed to apply only to the liver.
This corresponds to the widely cited version of the hypothesized MOA for hepatocarcinogenesis
induced by PPARa agonists posited by Klaunig et al. (2003). in which three key causal events
were proposed: activation of the receptor, perturbation of hepatocellular apoptosis and
proliferation, and selective clonal expansion. A number of intermediary events were considered
associative (e.g., expression of peroxisomal and nonperoxisomal genes, peroxisome
proliferation, inhibition of gap junction intracellular communication, hepatocyte oxidative stress
and Kupffer cell-mediated events). The data requirements suggested by Klaunig et al. (2003) for
demonstrating that the PPARa-activation MOA is operative did not comprise all purportedly
causal events; instead, these requirements included PPARa-agonism combined with microscopic
evidence for peroxisome proliferation (or, in lieu of evidence of peroxisome proliferation,
increased liver weight together with in vivo markers such as increases in peroxisomal
P-oxidation, CYP4A or acyl CoA oxidase). Alterations in proliferation and apoptosis were
considered corroborative evidence.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-160 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
4.3.5.5.2. Induction of hypothesized key events by tetrachloroethylene and metabolites
4.3.5.5.2.1. Activation of PPARa and associated markers
As summarized in Table 4-20, several in vivo studies have examined the effect of
tetrachloroethylene on peroxisome proliferation or its markers (Goldsworthy and Popp. 1987;
Odum et al.. 1988b; Philip et al.. 2007). Odum et al. (1988b) exposed groups of male and female
F344 rats and B6C3Fi mice by inhalation for 6 hours/day to 200-ppm (28 days only) or 400-ppm
(for 14, 21, or 28 days) tetrachloroethylene. Five animals per group were exposed. In both
sexes, hepatic PCO activity was increased in mice (up to 3.6-fold) and, to a lesser extent, in rats
(up to 1.3-fold). Modest PCO increases were also seen in the kidney of male rats at 200 ppm at
28 days (1.3-fold) but not 400 ppm at 14, 21, or 28 days. In female rat kidney, PCO was
elevated (approximately 1.6-fold) at all doses and times. However, peroxisome proliferation was
not seen in rat kidney upon microscopy. In contrast, hepatic peroxisome proliferation was noted
in all exposed mice on electron microscopy, and the percentage of cytoplasm occupied by
peroxisomes also increased in mice. In rats, variable increases in peroxisome volume were noted
at 200 ppm, but results lacked statistical significance. Catalase, another peroxisomal enzyme,
was unaffected by tetrachloroethylene; male mice exposed at 400 ppm showed the only moderate
(1.4-fold) increase. Mitochondrial proliferation was seen at 28 days in 400-ppm male mice. In
addition, a time-dependent proliferation of smooth endoplasmic reticulum in the liver of both
sexes correlated well with centrilobular hypertrophy. Tetrachloroethylene caused centrilobular
lipid accumulation in male and female mice. Relative liver weight was increased in mice of both
sexes.
Goldsworthy and Popp (1987) administered tetrachloroethylene (1,000 mg/kg-day) by
corn oil gavage to 5 male F344 rats and 5 male B6C3Fi mice for 10 days. In
tetrachloroethylene-exposed rats, PCO was modestly although not significantly elevated in the
liver (1.4-fold increase) and kidney (1.7-fold increase). In mice, tetrachloroethylene exposure
increased PCO activity 4.3-fold in liver and by 2.3-fold in kidney. Relative liver weight was
increased in rats and mice with tetrachloroethylene exposure, but relative kidney weight was
unaffected. A comparison of corn oil with methyl cellulose revealed no effect of the gavage
vehicle on tetrachloroethylene-induced PCO. A less-than-additive effect of trichloroethylene
(1,000 mg/kg) administered together was tetrachloroethylene on PCO induction was seen.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-161 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Table 4-20. Rodent studies of induction of peroxisome proliferation or its
markers by tetrachloroethylene
Species/strain/sex/number
Effect
Dose
Time
Rat, F344; and mouse,
B6C3FJ; both sexes
(5/group)
Odumetal. (1988b)
Mice of both sexes: increased relative liver
weight, centrilobular lipid accumulation and
peroxisome proliferation; increased PCO (up
to 3.7-fold)
200 and 400 ppm,
inhalation
14, 21,28 d
Male mice: mitochondrial proliferation
400 ppm, inhalation
28 d
Rats of both sexes: increased PCO (up to
1.3-fold)
200 and 400 ppm,
inhalation
14, 21,28 d
Rat, F344 (male only,
5/group) and B6C3Fi mice
(male only, 5/group)
Goldsworthy and Popp
(1987)
Mice: Increased relative liver weight; 4.3-fold
PCO increase
1,000 mg/kg-day for
10 d, corn oil gavage
10 d
Rats: Increased relative liver weight; modest
but not significant (1.4-fold) PCO increase
1,000 mg/kg-day for
10 d, corn oil gavage
10 d
Mouse, Swiss-Webster,
male (4 mice/group)
Philip et al. (2007)
Increased plasma ALT
150, 500, and
1,000 mg/kg-day,
aqueous gavage
24 hours to 14 d
after initial
exposure
Mild to moderate fatty degeneration and
necrosis, with focal inflammatory cell
infiltration
150, 500, and
1,000 mg/kg-day,
aqueous gavage
24 hours to 30 d
after initial
exposure
Increased mitotic figures and DNA synthesis
150, 500, and
1,000 mg/kg-day,
aqueous gavage
Peaked on 7 d,
sustained at
14-30 d
CYP4A increased at 7 but not 14 d, only at
1,000 mg/kg-day
1,000 mg/kg-day,
aqueous gavage
7 but not 14 d
The peroxisome-related effects of tetrachloroethylene are most likely mediated primarily
through TCA based on tetrachloroethylene metabolism producing more TCA than DCA, and the
lower doses of TCA required to elicit a response relative to DCA. 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 TCA and DCA. They report that a direct comparison of
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 tetrachloroethylene itself is relatively inactive in
the in vitro system, at least with mouse PPARa (Malonev and Waxman. 1999; Zhou and
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-162 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Waxman. 1998). In addition, Laughter et al. (2004) reported that the responses of ACO, PCO,
and CYP4A induction by TCA and DCA were substantially diminished in PPARa null mice.
Therefore, evidence suggests that tetrachloroethylene activates PPARa in vivo, and that the role
of TCA in activating PPARa is likely to predominate at doses relevant to tetrachloroethylene-
induced hepatocarcinogenesis.
4.3.5.5.2.2. Alterations of cell proliferation and apoptosis and clonal expansion of initiated
cells
As discussed above, increased cell proliferation in mice has been reported following
exposure to tetrachloroethylene. However, few data are available to inform the hypothesis that
activation of PPARa after tetrachloroethylene exposure causes alterations in cell proliferation
and apoptosis, followed by clonal expansion of initiated cells. Moreover, available data suggest
that PPARa-activation may not be the predominant cause of the observed cell proliferative
response. For example, transient increases in DNA synthesis and PCNA staining in the liver
were reported by Philip et al. (2007). similar to that observed with other PPARa agonists (with
the exception of WY-14,643, which induces sustained proliferation) (see Section 4.3.5.2.4.2).
However, Philip et al. (2007) suggest that PPARa-activation is not required for the observed cell
proliferative response, and rather that this is a regenerative response following cytotoxicity. This
is based on evidence of significantly increased CYP4A expression at only the highest dose
(1,000 mg/kg-day) and at the earliest time point (7 days), in contrast to the robust dose-
dependent proliferative response of a more prolonged nature (lasting for 14-30 days post
exposure) observed at the same and lower (150, 500 and 1,000 mg/kg-day) levels of
tetrachloroethylene. The authors concluded that their findings suggest peroxisome proliferation
is not a sustained response in spite of continued tetrachloroethylene exposure and, therefore, are
not supportive of a close mechanistic relationship of carcinogenicity and PPARa induction for
tetrachloroethylene-derived TCA. This interpretation is limited by the possible lack of
sensitivity of CYP4A protein expression as a marker of peroxisome proliferation, and the lack of
other supporting data for the observed absence of sustained peroxisome proliferation in the
context of a robust regenerative proliferative response. Additionally, the sensitivity of the SW
mouse to tetrachloroethylene hepatocarcinogenicity is unknown, somewhat limiting the
significance of these findings for the interpretation of hepatocellular tumor findings in other
mouse strains. However, other studies of the toxicity of tetrachloroethylene in the B6C3Fi strain
discussed above (e.g.. Schumann et al.. 1980) have reported liver toxicity and repair at
100 mg/kg-day, whereas Odum et al. (1988b) reported only modest increases in peroxisomal
markers in B6C3Fi mice with repeated exposures to 1,000 mg/kg-day. Another noteworthy
finding in the Odum et al. (1988b) was the modest increases in peroxisome proliferation
observed in rats.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-163 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Data on TCA are also informative of the extent to which tetrachloroethylene alters cell
proliferation and apoptosis through PPARa-activation, as it was concluded above that the
PPARa-agonism following tetrachloroethylene is mostly likely caused by its metabolism to
TCA. Data that inform the hypothesis that activation of PPARa after TCA exposure causes
alterations in cell proliferation and apoptosis, followed by clonal expansion of initiated cells, are
discussed in the EPA Toxicological Review of TCA (U.S. EPA. 2009). To summarize, several
studies have observed hepatocyte proliferation in response to TCA in mice (DeAngelo et al..
2008; Dees and Travis. 1994; Pereira. 1996; Sanchez and Bull. 1990; Stauber and Bull. 1997).
For instance, Dees and Travis (1994) observed relatively small (two- to threefold) but
statistically significant increases in [3H]thymidine incorporation in hepatic DNA in mice exposed
for 11 days at TCA doses (100-1,000 mg/kg) that increased relative liver weight. Increased
hepatic DNA labeling was seen at doses lower than those associated with evidence of necrosis,
suggesting that TCA-induced cell proliferation is not due to regenerative hyperplasia.
PPARa-null mice exposed to 2-g/L TCA in drinking water for 7 days do not show the
characteristic responses of ACO, PCO, and CYP4A induction associated with PPARa-activation
and peroxisome proliferation in wild-type mice (Laughter et al.. 2004). In addition, the livers
from wild-type but not PPARa-null mice exposed to TCA developed centrilobular hepatocyte
hypertrophy, although no significant increase in relative liver weight was observed. Therefore,
while there are data associating TCA exposure, PPARa-activation, and cell proliferation, it is not
clear the extent to which PPARa-activation is the cause of the observed cell proliferation.
Data informing the hypothesis that PPARa-activation following tetrachloroethylene
exposure causes clonal expansion of initiated cells, are limited to studies of its metabolite TCA.
Mechanistic studies reveal that the mode of action for TCA hepatocarcinogenesis is complex and
that TCA may induce tumors by multiple modes of action that may not be mutually exclusive
(U.S. EPA. 2009). In particular, tumor induction by TCA appears to involve perturbation of cell
growth, reduced intercellular communication (Benane et al.. 1996). release of cytokines and
oxidants by activated Kupffer cells and hypomethylation of DNA.
4.3.5.5.2.3. Conclusions regarding induction of hypothesized key events by
tetrachloroethylene and metabolites
The available evidence from tetrachloroethylene and its metabolites supports the
conclusion that tetrachloroethylene exposure leads to PPARa-activation predominantly through
its metabolite TCA. There is more limited evidence supporting the hypothesis that PPARa-
activation is the cause of the cell proliferative responses observed, and some evidence suggesting
that PPARa-activation is not the cause of these responses. Data informing the hypothesis that
PPARa-activation following tetrachloroethylene exposure causes clonal expansion of initiated
cells are even more limited.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-164 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
4.3.5.5.3. Are activation of PPARa and its sequelae key events in tetrachloroethylene-
induced hepatocarcinogenesis?
No tetrachloroethylene-specific data have directly tested the hypothesis that
tetrachloroethylene-induced PPARa-activation, along with its sequelae, are key or causative
events in tetrachloroethylene-induced hepatocarcinogenesis (e.g., bioassays with knockout mice
or involving the blocking of hypothesized key events). With respect to more associative data,
Philip et al. (2007) found increases in CYP4A, a marker for PPARa-activation, to be transient
(only increased at 7 days) rather than sustained, and only occurring at the highest dose (1,000
mg/kg-day). These data are not supportive of PPARa-activation as a key event in
tetrachloroethylene-induced hepatocarcinogenesis for two reasons: (1) chronic activation would
be needed to sustain changes in cell proliferation, apoptosis, and clonal expansion, and
(2) statistically significant increases in liver tumors have been reported at doses around 500
mg/kg-day (NCI. 1977). at which no increased CYP4A activity was reported. However, the SW
strain of mouse used by Philip et al. (2007) may differ in tumor responsiveness from those used
in the cancer bioassays discussed above.
Support for this MOA is based primarily on the hypothesis that TCA induces tumors
through PPARa-activation, and the fact that TCA is formed after in vivo exposure to
tetrachloroethylene. The experimental evidence related to the hypothesis that TCA induces
tumors through PPARa-activation is discussed extensively in the EPA ToxicologicalReview of
TCA (U.S. EPA. 2009). TCA activates PPARa, and induces peroxisome proliferation and
hepatocyte proliferation. However, a number of inconsistencies and data gaps reduce the
confidence in the conclusion that TCA induces hepatocarcinogenesis solely through a PPARa-
activation MOA. First, while TCA induces peroxisome proliferation (a marker for PPARa-
agonism) in both rats and mice, to date, TCA has been shown to be tumorigenic in B6C3Fi mice
but not F344 rats (DeAngelo et al.. 1997) (the only strains tested for carcinogenicity). In
addition, the tumor phenotype of TCA-induced mouse liver tumors has been reported to have a
different pattern of H-ras mutation frequency from DCA and other peroxisome proliferators
(Bull et al.. 2002)(Stanelv et al., 1994; Fox et al., 1990; Hegi et al., 1993). Other effects of TCA,
including increased c-myc expression and hypomethylation of DNA, are not specific to the
PPARa-activation MOA, and other data (discussed below in Section 4.3.4.2.4) also contribute
uncertainty as to whether PPARa independent mechanisms may be involved in TCA-induced
tumors in mice.
To summarize, based on data from tetrachloroethylene and its metabolites alone, there is
only limited evidence that activation of PPARa and its sequelae are key events in
tetrachloroethylene-induced hepatocarcinogenesis. In all, the modest peroxisome proliferation
observed in response to tetrachloroethylene may lack specificity and consistency with respect to
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-165 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
tissue, species, and dose, and studies of the temporal sequence of events are limited. Given the
limitations in the database of tetrachloroethylene-specific studies, it can be concluded that the
few studies demonstrating activation of PPARa and related markers by tetrachloroethylene are
insufficient to demonstrate a causative role of this effect in the induction of other key events
posited for the PPARa mode of action hypothesis, and for hepatocarcinogenesis by
tetrachl oroethy lene.
4.3.5.5.4. Other experimental evidence for the hypothesized MOA
4.3.5.5.4.1. Evidence from PPARa-null mouse bioassays
An apparent reduction was seen in tumor response to an 11-month exposure to the
prototypical agonist 4-chloro-6-(2,3-xylidino)-2-pyrimidyl-thio]acetic acid (Wy-14,643) in
PPARa-null mice in comparison to wild-type mice (Peters et al.. 1997). Peters et al. reported the
absence of tumors in nine PPARa-null mice exposed to Wy-14,643 at 11 months, whereas each
of the six similarly exposed wild-type mice had multiple hepatocellular neoplasms.
As has also has been shown for Wy-14,643, the monoester metabolite
(mono-2-ethylhexylphthalate, MEHP) of DEHP activates PPARa in vitro (Issemann and Green.
1990; Malonev and Waxman. 1999). Other evidence for DEHP includes induction of
peroxisome proliferation (or an increase in peroxisomal enzyme activity), an associative event in
the MOA, by tumorigenic doses of DEHP in the liver of mice and rats and of MEHP in rat
hepatocytes (David et al.. 1999; Gray et al.. 1982; Gray et al.. 1983; Hasmall et al.. 1999;
Mitchell et al.. 1984; Mitchell et al.. 1985; Reddv et al.. 1986). Additionally, an absence of
peroxisomal enzyme induction and peroxisome proliferation in PPARa-null mice exposed to
DEHP for 24 weeks was demonstrated (Ward et al.. 1998).
However, as reviewed recently by Guyton et al. (2009). a 2-year bioassay found that
DEHP (100 or 500 ppm) induces liver tumors in PPARa-null mice (Ito et al.. 2007). Ito et al.
reported a significant trend for the observed increase in total liver tumors with DEHP in
PPARa-null male mice with Sv/129 genetic background generated as described in Lee et al.
(1995). Guyton et al. (2009) performed additional statistical analyses to compare the Ito et al.
results with those of a prior DEHP bioassay in B6C3Fi wild-type mice (David et al.. 1999). A
pair-wise analysis found that DEHP (500 ppm) significantly increased adenomas in PPARa-null,
but not in companion wild-type, mice compared to their respective controls (see Figure 4-3,
single asterisks). In the David et al. study of B6C3Fi mice, DEHP (500 ppm) also significantly
increased adenomas and adenomas plus carcinomas (see Figure 4-3B, single asterisks).
Moreover, a significant dose-response trend for adenomas and for adenomas plus carcinomas
was seen in both the Ito et al. PPARa-null mice and the David et al. B6C3Fi mice after exposure
to DEHP (see Figure 4-3B, double asterisks). Additionally, Guyton et al. (2009) found no
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-166 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
statistically significant differences between groups at the same dose, including controls,
consistent with mouse strain and PPARa genotype having no influence on carcinogenicity under
the study conditions.
The observed lack of difference in reported control incidences across groups lends
support to the approach of basing comparative analyses on concurrent controls. Historical data
on spontaneous liver tumor incidences in PPARa-null mice are limited; Ito et al. (2007) is the
largest published 2-year bioassay in PPARa-null mice, reporting findings for 24/25 surviving
unexposed animals at 23 months of age. A different laboratory that had established a distinct
breeding colony reported mouse liver tumor incidences in 12 PPARa-null Svl29/C57BL/6 mice
~2 years of age (Howroyd et al.. 2004). Adenomas and carcinomas were reported in 6/12 and
2/12 PPARa-null mice, respectively, compared with adenomas in 5/22 wild-type animals. As
Howroyd et al. note, —The relatiely small number of animals available made it difficult to draw
robust conclusions concerning enhancement of spontaneous findings in PPARa-null mice." In
addition, cross-laboratory differences (particularly the low survival of PPARa-null mice in the
Howroyd et al. relative to the Ito et al. study) limit statistical comparisons based on this data set.
In summary, the Ito et al. (2007) study indicates that DEHP carcinogenesis can occur
independently of PPARa-activation. As noted in a recent National Research Council report on
risk assessment (NRC. 2008). this finding -ealls into question" the IARC conclusions regarding
the carcinogenic risks of DEHP (IARC. 2000). Indeed, PPARa-activation and the subsequent
key events in the hypothesized MO A do not represent the sole cause of DEHP liver
tumorigenesis. Although new hypotheses are being generated based on more detailed
comparisons between wild-type and PPARa-null mice (Eveillard et al.. 2009; Ito et al.. 2007;
Takashima et al.. 2008). the mechanisms by which DEHP induces hepatocarcinogenesis remain
unknown.
4.3.5.5.4.2. Quantitative analyses of hypothesized key events and carcinogenic potency
If potency for PPARa-activation or its attendant sequelae is quantitatively associated with
carcinogenic activity or potency, then it might be possible to predict differences in sensitivity for
carcinogenesis (such as may occur across species) for environmental contaminants that activate
PPARa (e.g., certain phthalates and chloroacetic acids) using quantitative information about the
key events alone. It is, thus, of interest to assess whether potency for inducing these events is
quantitatively related to hepatocarcinogenic potential by these and other compounds that also
activate PPARa. However, there are limitations in the dose-response data available for such
analyses, specifically for precursor events in the proposed PPARa-activation MOA as well as for
liver tumor induction. Most tumor data, including for the best characterized PPARa agonists, are
for exposure concentrations inducing well above 50% tumor incidence with less-than-lifetime
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-167 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
administration. Precursor events have typically been studied at a single dose, often eliciting a
near maximal response, thus, precluding benchmark-based comparisons across studies. This is
especially true for Wy-14,643, which has been administered most often at only one exposure
concentration (1,000 ppm) that elicits a 100% tumor incidence after 1 year or less (Peters et al..
1997) and that also appears to be necrogenic (Woods et al.. 2007). On the other hand,
hypothesized precursor events such as hepatomegaly, peroxisome proliferation, and increased
DNA synthesis appear to have reached their maximal responses at 50 ppm Wy-14,643, with
some statistically significant responses as low as 5 ppm (Marsman et al.. 1992; Wada et al..
1992). Potencies across compounds have rarely been compared in a single study using the same
experimental paradigm. These deficits in the database notwithstanding, provided below is an
assessment of the quantitative predictive power of the potency for four proposed data elements
for establishing the hypothesized MOA for hepatocarcinogenesis: PPARa-activation in mice;
and hepatomegaly, DNA synthesis, and increased peroxisome proliferation in rats.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-168 DRAFT-DO NOT CITE OR QUOTE
-------
-/- (Ito et al.) +/+ (Ito et al.) +/+ (David et al.)
Adenomas
Adenomas+Carcinomas
0 100 500
0 100 500
ppm DEHP
0 100 500
Figure 4-3. Incidences of hepatocellular adenomas (A) and hepatocellular
adenomas and carcinomas (B) in mice exposed to DEHP. Ito et al. (2007)
exposed PPARa null [-/-] and wild-type [+/+] Sv/129 mice for 22 months;
David et al. (1999) exposed B6C3Fi wild-type [+/+] mice for up to 104 weeks.
Data are presented as incidence +/- SD assuming a binomial distribution for
each group. Single asterisks (*) indicate a significant difference from
controls of the same genotype in the same study (Fisher exact test,p < 0.05).
Double asterisks (**) indicate a significant trend with dose in the study
(Cochran Armitage test,p < 0.05). All pair-wise cross-study comparisons
between like dose groups (e.g., Ito et al. [-/-] 500 ppm compared with David et
al. [+/+] 500 ppm) were not significant (Fisher exact test p > 0.05). Because
David et al. (1999) reported only adenomas and carcinomas, the
cholangiocellular carcinoma reported by Ito et al. (2007) in DEHP-exposed
PPARa null mice was excluded from analyses. Adapted from Guyton et al.
(2009).
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/2011 4-169 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
4.3.5.5.4.2.1. PPARa-activation in mice
Table 4-21 presents data for four peroxisome proliferators in order of decreasing potency
for inducing mouse liver tumors. These compounds were selected because of their importance to
environmental human health risk assessments and because data to derive receptor activation
potency indicators were available from a single study (Malonev and Waxman. 1999). The
transactivation potencies of MEHP, Wy-14,643, dichloroacetic acid (DCA), and TCA for the
mouse PPARa were monitored using a luciferase reporter gene containing multiple PPAR
response elements derived from the rat hydratase/dehyrogenase promoter in transiently
transfected COS-1 monkey kidney cells. The derived potency indicators were compared to the
TD5o (i.e., the daily dose inducing tumors in half of the mice that would otherwise have remained
tumor-free) from the Carcinogenic Potency Database (CPDB) of Gold et al. (2005). Note that
for Wy-14,643, the dose listed yielded a maximal response and, thus, represents an upper limit to
the TD50 (indicated by Two estimates of PPARa transactivation potency are given, the first
based on 50% of the maximal response (i.e., EC50) and the second based on the effective
concentration required for a twofold increase in activity (i.e., EC2-f0id) (Malonev and Waxman.
1999). Because unmetabolized DEHP does not exhibit PPARa activity, the transactivation
activity of its metabolite MEHP is given but compared to the hepatocarcinogenic potency
indicator for DEHP. In addition, unmetabolized tetrachloroethylene does not exhibit PPARa
activity, so is not included in the table. No data on the potency for transactivation of rat PPARa
by chemicals in the CPDB were located to enable a similar comparison in rats.
These data clearly show a lack of correlation between the potencies for in vitro PPARa
transactivation and in vivo tumorigenesis across different PPARa agonists. Especially notable is
that MEHP exhibited orders of magnitude more potency for transactivating mouse PPARa than
DCA, but DEHP was sixfold less potent as a mouse hepatocarcinogen. TCA was more similar in
potency to DCA for both outcomes, i.e., was also dramatically less active at transactivating
PPARa than DEHP despite exhibiting comparable hepatocarcinogenic potency. Wy-14,643 and
MEHP activate PPARa at comparable concentrations when directly compared in the
transactivation assay, but the carcinogenic potency of Wy-14,643 was estimated to be at least
70-fold higher than DEHP. This difference cannot be explained by pharmacokinetics (Kessler et
al.. 2004; Pollack et al.. 1985). Possible explanations for these results include one or more of the
following: (1) the transactivation assay is not an accurate quantitative indicator of in vivo
receptor activation, (2) the rate and nature of effects downstream of PPARa-activation depends
on the ligand or, (3) there are rate-limiting events independent of PPARa-agonism that
contribute to mouse hepatocarcinogenesis by the agonists examined.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-170 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Table 4-21. Potency indicators for mouse hepatocarcinogenicity and in vitro
transactivation of mouse PPARa for four PPARa agonists3
Chemical
Carcinogenic potency
indicators (mg/kg-day)
Transactivation potency
indicators (jiM)
TDS0
EC50
EC2-fold
Hepatocarcinogens
Wy-14,643
<10.8
0.63
-0.4
DCA
119
-300
-300
TCA
584
-300
-300
DEHP/MEHP
700
-0.7
-0.7
a TD50, the daily dose inducing tumors in half of the mice that would otherwise have remained
tumor-free, estimated from the Carcinogenic Potency Database (Gold et al. 2005). EC50, the
effective concentration yielding 50% of the maximal response; EC2.f0id, the effective
concentration required for a twofold increase in activity. Transactivation potencies were
estimated from Maloney and Waxman (19991. The "<" symbol denotes an upper limit due to
maximal response. A symbol indicates that the transactivation potency was approximated
from figures in Maloney and Waxman (19991.
Adapted from Guy ton et al. (2009).
4.3.5.5.4.2.2. Hepatomegaly, DNA synthesis, and peroxisome proliferation in rats
Table 4-22 compares potency indicators for various precursor effects at the TD50 for four
PPARa agonists and rat hepatocarcinogens. The analysis of whether there are consistent levels
of in vivo precursor effect induction across peroxisome proliferators at the TD50 does not include
all of the data from a similar, prior analysis by Ashby et al. (1994) for several reasons. First,
unlike the CPDB, Ashby et al. did not adjust carcinogenicity data for less-than-lifetime dosing,
which is relevant for most compounds. Second, for those mouse carcinogens reported in the
CPDB, only acute data are available regarding DNA synthesis effects from Ashby et al.
Therefore, this analysis was restricted to rat precursor and potency data for the four compounds
Wy-14,643, nafenopin, clofibrate, and DEHP and included both 1-week and 13-week data to
separately address transient and sustained changes in DNA synthesis. Even for this small set of
compounds, several limitations in the rat database were apparent. Because no single study
provided comparative data for the precursor endpoints of interest, four separate reports were
used. In the Wada et al. (1992) and Tanaka et al. (1992) studies of Wy-14,643 and clofibrate,
respectively, administered doses were within 10% of the TD50. However, nafenopin data were
only available at a single dose of 500 ppm (Lake et al.. 1993). which was linearly interpolated to
the TD50. The highest administered dose of DEHP was 12,500 ppm (David et al.. 1999). a dose
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-171 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
notably below the TD50, and, thus, a lower limit based on the assumption of monotonicity with
dose is shown. A further data limitation is that in the CPDB, only the TD50 for one of the four
compounds, DEHP, incorporates data from studies administering more than one dose for 2 years.
The results shown in Table 4-22 indicate that potency for the occurrence of short-term in
vivo markers of PPARa-agonism varies widely in magnitude and lacks any apparent correlation
with carcinogenic potency. Such differences have been noted previously. Similar to the results
presented in Table 4-22, Marsman et al. (1988) noted that although DEHP (12,000 ppm) and
Wy-14,643 (1,000 ppm) induced a similar extent of hepatomegaly and peroxisome proliferation
(measured either morphologically or biochemically) after 1 year, the frequency of hepatocellular
lesions was over 100-fold higher in Wy-14,643 relative to DEHP-exposed rats. In addition, a
higher labeling index was reported for 12,500 ppm DEHP than the maximal level attained after
50 to 1,000 ppm Wy-14,643 (David et al.. 1999; Tanaka et al.. 1992; Wada et al.. 1992). Such
differences in response with dose and time seen among PPARa agonists, which are prominent
enough to prevent displaying dose-response data on a common scale. For instance, labeling
differences in maximal responses were not examined in this analysis. Also not addressed are
Table 4-22. Potency indicators for rat hepatocarcinogenicity and common
short-term markers of PPARa-agonism for four PPARa agonists3
Chemical
Tumor TD50
(ppm in diet)
Fold-increase over control at tumor TD50
1 wk
13 wk
RLW
LI
PCO
RLW
LI
PCO
Wy-14,643
109
1.8
12
13
2.6
6.8
39
Nafenopin
275
1.4
3.6
7.6
1.5
1.12
6.7
Clofibrate
4,225
1.4
4.4
4.2
1.4
0.95
3.7
DEHP
17,900
>1.4
>19
>3.6
>1.9
>1.25
>4.9
Tor ease of comparison with precursor effect studies, administered doses for the tumor TD50s in the
Carcinogenic Potency Database were back-converted to equivalent ppm in diet using the formula of
Gold et al. (2005), i.e., TD50 (mg/kg-day) = TD50 (ppm in diet) * 0.04 (for male rats). Administered
doses for precursor data on Wy-14,643 (Wadaetal.. 1992) and clofibrate (Tanaka et al.. 1992) were
within 10% of the TD50. Because nafenopin precursor data were only available at 0 and 500 ppm (Lake
et al.. 1993). these doses were linearly interpolated to the TD50. Because the highest administered dose
of DEHP in precursor effect studies was 12,500 ppm (David et al.. 1999). a lower limit is shown, based
on the assumption of monotonicity with dose. RLW = relative liver weight, LI = labeling index, PCO =
cyanide insensitive palmitoyl CoA oxidation.
Adapted from Guyton et al. (2009).
index is increased in a dose-dependent manner at 1 week by clofibrate (1,500, 4,500, and 9,000
ppm) but is decreased compared with controls at 13 weeks at the two higher doses (Tanaka et al..
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-172 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
1992). Together, these findings underscore the significant chemical-specific quantitative
differences in these markers that limit their utility for predicting carcinogenic dose-response
relationships.
4.3.5.5.4.3. Evidence from transgenic animals
Data from transgenic animals suggest the key events in the hypothesized MO A—PPARa-
activation, hepatocellular proliferation, and clonal expansion—are not sufficient to cause tumors.
This suggests that other events not mediated by PPARa-activation, either independently or in
combination with PPARa-activation, are necessary to induce tumors. The discussion below is
based on the review by Guyton et al. (2009).
Yang et al. (2007b) raises questions regarding whether PPARa-activation in hepatocytes
is causally linked to hepatocarcinogenesis as a sole operant MO A. The experimental approach
entailed fusing the mouse PPARa to the potent viral transcriptional activator VP 16 under control
of the liver enriched activator protein (LAP) promoter, resulting in targeted constitutive
expression of activated PPARa in hepatocytes. In LAP-VP16PPARa transgenic mice, ligand-
independent hepatocyte PPARa-activation evoked many of the same hepatic responses (in type
and magnitude) as seen with PPARa ligand treatment of companion wild-type 129/Sv mice. For
instance, DNA synthesis was increased in LAP-VP16PPARa transgenic mice; the effect was
persistent and still evident at 11 months of age. In addition, increases were reported in markers
of peroxisome proliferation (including increases in expression of peroxisomal membrane protein
70, acyl CoA oxidase and CYP4A family genes, and enhanced cyanide insensitive palmitoyl
CoA oxidation). Other effects included an increase in cell-cycle genes (cyclin D1 and cyclin-
dependent kinases 1 and 4) and a decrease in serum triglycerides and free fatty acids. Together,
these results are consistent with the view that PPARa-activation and its sequelae are alone
sufficient to induce increased hepatocyte DNA synthesis and peroxisome proliferation.
However, constitutive PPARa-activation in hepatocytes in the LAP-VP16PPARa
transgenic mouse model was not sufficient to induce several important hepatic responses
stimulated by PPARa ligand treatment of wild-type mice. Notably, no preneoplastic hepatic
lesions or hepatocellular neoplasia were found in —>20 LAIVP16PPARa mice at the age of
over 1 year" (Yang et al.. 2007b). In sharp contrast, wild-type mice exposed to the PPARa
agonist Wy-14,643 for 11 months developed grossly visible lesions consistent with previous
reports of its hepatocarcinogenicity (e.g.. Peters et al.. 1997). Interestingly, nonparenchymal cell
proliferation was seen with Wy-14,643 exposure of wild-type mice but was absent in the
LAP-VP16PPARa transgenic mice. In addition, although liver weight was increased in
LAP-VP16PPARa transgenic mice, the extent of hepatomegaly was reduced in comparison to
Wy-14,643-exposed wild-type mice and hepatocellular hypertrophy was absent.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-173 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Thus, the Yang et al. (2007b) study provides evidence that, by itself, PPARa-activation
(and its sequelae) is not sufficient to induce hepatocarcinogenesis. These data are, therefore,
inconsistent with the hypothesis that effects mediated through PPARa-activation constitute a
complete MOA for carcinogenesis. Notably, key events in the proposed MOA such as the robust
and sustained elevation in hepatocyte proliferation (evidenced by enhanced DNA synthesis),
accompanied by enzyme changes commonly associated with peroxisome proliferation, did not
evoke hepatocarcinogenesis. In fact, a comparable extent of sustained increases in hepatocyte
DNA synthesis was seen with constitutive PPARa-activation in the LAP-VP16PPARa
transgenic mouse model and Wy-14,643 exposure in wild-type mice, but only the latter
developed liver tumors under comparable experimental paradigms.
4.3.5.5.5. Rationale for species differences
Toxicodynamic differences across species, including in the absolute or allometrically
scaled amount or activity of the receptor, may contribute to differences in sensitivity of response
to PPARa agonists. Absolute levels of PPARa are generally thought to be lower in human
compared with rodent liver. However, PPARa amount varies by an order of magnitude among
individuals (Palmer et al.. 1998; Tugwood et al.. 1996). e.g., 1 of the 6 human samples examined
expressed levels comparable to the mouse in one study (Walgren et al.. 2000). The pattern of
PPARa expression across tissues also differs across species (Melnick. 2001; Tugwood et al..
1996). e.g., human levels are higher in kidney and skeletal muscle than in liver, while the highest
rodent levels are in liver and kidney. In addition, considerable interindividual variation in
PPARa structure and function among humans has been reported (Tugwood et al.. 1996). and
polymorphisms have been shown to increase or decrease receptor levels and to modulate
baseline lipid and apolipoprotein levels, atherosclerotic progression, and the presence of diabetes
mellitus and insulin resistance (Flavell et al. 2002, 2005)(Foucher et al.. 2004; Jamshidi et al..
2002; Tai et al.. 2006; Tanaka et al.. 2007). An impact of PPARa polymorphisms on preexisting
disease status and response to PPARa agonists is also suggested from bezafibrate
[2-(4-(2-[(4-chlorophenyl)formamido]ethyl)phenoxy)-2-methylpropanoic acid] and gemfibrozil
[5-(2,5-dimethylphenoxy)-2,2-dimethyl-pentanoic acid] trials (Jamshidi et al.. 2002; Tai et al..
2006).
The human PPARa is functional in in vitro transactivation assays and is responsive to a
number of PPARa agonists (e.g., nafenopin, clofibrate, and WY-14,643) (Malonev and
Waxman. 1999; Mukheriee et al.. 1994; Sher et al.. 1993). Compared with the mouse PPARa,
human PPARa is suggested to be 10- to 20-fold less responsive to Wy-14,643 (Malonev and
Waxman. 1999; Mukheriee et al.. 1994; Palmer et al.. 1998). However, this magnitude of
interspecies difference has not been demonstrated for other compounds. Hurst and Waxman
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-174 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
(2003) reported a fivefold lower sensitivity to the DEHP metabolite MEHP of human compared
with mouse PPARa (EC50 = 3.2 [xM vs. 0.6 (xM) in transfected COS-1 monkey kidney cells, but
acknowledged that they could not quantify the relative amount of each receptor. Using a similar
experimental paradigm, Wolf et al. (2008) found an approximately twofold lower slope of the
dose-response curve for activation of human compared with mouse PPARa for perfluorooctanoic
acid and other perfluoroalkyl acids. For other PPARa agonists, including TCA and DCA, little
(
-------
1
2
3
4
5
6
7
8
9
10
11
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
(e.g., Kupffer cells), may limit the in vitro hepatocyte proliferative response, as observed for
other species (e.g. Parzefall et al.. 2001). The extent of peroxisome proliferation in human liver
following exposure to fibrate drugs (e.g., with clofibrate, gemfibrozil, or fenofibrate) or dialysis
treatment (possibly due to DEHP exposure) is reported to be generally less than the rodent
response (Blumcke et al.. 1983; De La Iglesia et al.. 1982; Ganning et al.. 1984; Hanefeld et al..
1983)(Ganning. 1987)(Gariot et al. 1987)(Hanefeld et al. 1980). However, the ability to
quantitatively characterize human sensitivity to this effect is limited (e.g., by the small number of
subjects studied).
In sum, despite notable qualitative similarities, quantitative differences in receptor
activation and the subsequent events in the hypothesized MOA are evident across species. The
magnitude of these differences has been best characterized for Wy-14,643, to which rodents
appear to have 10-fold or more greater sensitivity for response (Cheung et al.. 2004; Malonev
andWaxman. 1999; Morimura et al.. 2006; Mukheriee et al.. 1994; Palmer etal.. 1998; Yu et al..
2001). Although more limited, studies of other agonists suggest a smaller magnitude of
difference in sensitivity for response across species than is seen for Wy-14,643 (Hurst and
Waxman. 2003; Malonev and Waxman. 1999; Yu et al.. 2001). Considerable interindividual
variation in PPARa amount, structure and function has been reported among humans (Tugwood
et al.. 1996). and some studies have suggested variability in human response to PPARa agonists
(Jamshidi et al.. 2002; Tai et al.. 2006). However, few studies have examined directly how these
factors may affect sensitivity—as well as the potential for heterogeneity of response—to
hepatocarcinogenesis induced by PPARa agonists in humans.
Another consideration is whether human epidemiologic data on fibrates offer an indirect
test of the PPARa-activation MOA hypothesis. Human exposures to exogenous and endogenous
PPARa agonists encompass a broad group of chemicals, including environmental contaminants
known to activate the receptor, as well as a number of therapeutic agents whose molecular target
is one or more receptors in the PPAR family. Indeed, fibrate drugs were developed using rodent
models to treat hyperlipidemia in humans before the receptor was identified. These agents have
varying degrees of affinity for PPARa (Shearer and Hoekstra. 2003). and some have multiple
mechanisms of action. Drugs that have PPARa agonist activity include fibrates or fibric acid
derivatives (which are primarily PPARa agonists), bezafibrate (which also shows PPARy
activity), dual PPARa/y agonists currently under development, the glitazones, and nonsteroid
anti-inflammatory drugs (e.g., ibuprofen) (Sertznig et al.. 2007).
Some human data on PPARa agonist effects are available from fibrate clinical trials and
population case-control studies of site-specific cancer (BIP Study Group. 2000; Frick et al..
1987; Frick et al.. 1997; Huttunen et al.. 1994; Keech et al.. 2005; Tenkanen et al.. 2006;
Fortuney et al., 2006; Keech et al., 2005; Meade, 2001; Rubins et al., 1993; Rubins et al., 1999;
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-176 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Rubins et al., 1992; Canner et al., 1986; Committee of Principal Investigators 1978, 1980, 1984;
Coronary Drug Research Group 1975, 1977; De Faire et al. 1995; Diabetes Atherosclerosis
Intervention Study Investigators 2001; Freeman et al., 2006). These studies examined a range of
human responses to PPARa agonists, which included atherosclerosis, cardiovascular disease,
serum biomarkers of fatty acid metabolism, acute toxicity, and, more limitedly, organ-specific
chronic toxicity, including cancer. However, examination of hepatotoxicity in the fibrate clinical
trials has been limited to alterations in hepatic metabolic pathways and changes in liver enzymes
as assessments of drug tolerance, because the primary focus of these trials was cardiovascular
events.
Reviews of the PPARa-activation MOA hypothesis have generally focused on liver
cancer response in two fibrate clinical trials, the Helsinki Heart Study (Frick et al.. 1987;
Huttunen et al.. 1994; Tenkanen et al.. 2006) and the World Health Organization's Cooperative
Trial on Primary Prevention of Ischemic Heart Disease (Committee of Principal Investigators
1978, 1980, 1984), and have concluded that, while limited, those data did not provide evidence
of an increased liver cancer risk from fibrate exposure (Ashbv et al.. 1994; Klaunig et al.. 2003).
However, the available studies have low power to detect statistical differences in the risk of liver
cancer; an estimated five or fewer liver cancer deaths would have been expected in these studies
using data from the National Cancer Institute's Surveillance, Epidemiology, and End Results
database (Ries et al., 2008). This low statistical power, in addition to the studies' exclusion or
removal of subjects showing signs of liver (or other) toxicity from treatment, precludes a strong
conclusion about the presence or lack of liver cancer risk. These studies and the other fibrate
trials did not examine site-specific causes of mortality or morbidity and did not follow subjects
for a sufficient period to adequately consider cancer latency; in addition, placebo subjects were
offered fibrate therapy at the end of the clinical trials, making analyses after further follow-up
difficult to interpret. For example, the three trials that did assess mortality after a follow-up
period longer than 10 years included liver cancers in a larger category of contiguous sites or in
the category of all cancers, introducing disease misclassification and a downward bias for any
site-specific treatment-related cancers (Canner et al., 1986; Committee of Principal Investigators
1978, 1980, 1984)(Huttunen et al.. 1994; Tenkanen et al.. 2006). In voluntary postmarketing
safety reports to the U.S. Food and Drug Administration (FDA), rates of liver adverse event
reports for gemfibrozil and fenofibrate (2.6 and 6.9 per 1,000,000 prescriptions, respectively)
were similar to that of statins (Holoshitz et al.. 2008). However, an examination of liver cancer
is precluded by the general under-reporting of chronic toxicities to FDA, and the lack of specific
FDA reporting requirements for cancer, even premarketing. Because of these inadequacies, the
available epidemiologic data for fibrate drugs cannot inform conclusions about the relevance of
PPARa-activation to human cancer.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-177 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
4.3.5.6. Mode of Action Conclusions for Hepatocellular Tumors
There is only limited experimental support for the position that tetrachloroethylene-
induced hepatocarcinogenesis is mediated solely by the hypothesized PPARa-activation MOA.
Chemical-specific data for PPARa-activation support the view that this is not the primary MOA
for hepatocarcinogenesis. Philip et al. (2007) suggest that PPARa-activation is not required for
the observed cell proliferative response. This is based on evidence of significantly increased
CYP4A expression at only the highest dose (1,000 mg/kg-day) and at the earliest time point
(7-days), in contrast to the robust dose-dependent proliferative response of a more prolonged
nature (lasting for 14-30 days post exposure) observed at the same and lower (150, 500, and
1,000 mg/kg-day) levels of tetrachloroethylene. The authors concluded that their findings
suggest peroxisome proliferation is not a sustained response in spite of continued
tetrachloroethylene exposure and, therefore, are not supportive of a close mechanistic
relationship of carcinogenicity and PPARa induction for tetrachloroethylene-derived TCA.
Limitations of this interpretation include the possible lack of sensitivity of CYP4A protein
expression as a marker of peroxisome proliferation, and the unknown sensitivity of the SW
mouse to tetrachloroethylene hepatocarcinogenicity. However, other investigators (e.g.,
Schumann et al.. 1980) have reported liver toxicity and repair at 100 mg/kg-day in the B6C3Fi
strain, whereas repeated exposures to 1,000 mg/kg-day were reported by Philip et al. (2007) and
Odum et al. (1988b) to only modestly increase peroxisomal markers in SW and B6C3Fi mice,
respectively. Odum et al. (1988b) also observed moderate increases in peroxisome proliferation
in rats, a species insensitive to tetrachloroethylene hepatocarcinogenicity. In all, these findings
indicate that the modest peroxisome proliferation observed in response to tetrachloroethylene
may lack specificity with respect to species, tissue and dose. Studies of the temporal sequence of
events are limited. Given the limitations in the database of tetrachloroethylene-specific studies,
it can be concluded that the few studies demonstrating peroxisome proliferation by
tetrachloroethylene are insufficient to demonstrate a causative role of this effect in the induction
of other key events posited for the PPARa mode of action hypothesis, and for
hepatocarcinogenesis by tetrachloroethylene.
Other data and analyses more generally support the view that the hypothesized PPARa-
activation MOA is not a sole causative factor in rodent hepatocarcinogenesis. PPARa-agonism
may play a significant role in mouse liver tumor induction by some compounds, such as
Wy-14,643. However, recent studies suggest that DEHP can induce tumors in a PPARa
independent manner without any loss of potency (Ito et al.. 2007). and that PPARa-agonism in
hepatocytes is itself insufficient to cause tumorigenesis (Yang et al.. 2007b). Additional analyses
presented above demonstrate that peroxisome proliferation and associated markers are poor
quantitative predictors of hepatocarcinogenesis in rats or mice. These data and analyses raise
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-178 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
serious concerns about basing human health risk assessment conclusions exclusively on evidence
of key events in the hypothesized PPARa-activation MO A, given that other modes, mechanisms,
toxicity pathways and molecular targets may contribute to or be required for the observed
adverse effects. Indeed, for most PPARa agonists, chemical-specific data to define the range of
effects that may contribute to human carcinogenesis are insufficient. Similarly, the
epidemiologic data are inadequate to inform conclusions of human relevance.
A recent review (Rusvn et al.. 2006) addressed other mechanistic effects of the PPARa
agonist DEHP and proposed that tumors arise from a combination of molecular signals and
pathways, rather than from a single event such as PPARa-activation. Indeed, the PPARa
agonists are pleiotropic and have been reported to exhibit a diversity of responses in addition to
the hallmark effect of peroxisome proliferation, including genotoxicity (reviewed by Melnick.
2001). epigenetic alterations (e.g., hypomethylation) (Pogribny et al.. 2007). oxidative stress
(reviewed in O'Brien et al., 2005), and effects on other receptors (e.g., Guo et al., 2007) and
other organelles (e.g., mitochondria) within parenchymal cells (Lundgren et al. 1987)(Scatena et
al.. 2003; Youssef and Badr. 1998; Zhou and Wallace. 1999). As reviewed above, the
metabolites of tetrachloroethylene have been shown to induce a number of effects that may
contribute to carcinogenicity, including mutagenicity, alterations in DNA methylation, and
oxidative stress. Given the demonstrated mutagenicity of several tetrachloroethylene
metabolites, the hypothesis that mutagenicity contributes to the MOA for tetrachloroethylene
carcinogenesis cannot be ruled out, although the specific metabolic species or mechanistic
effects are not known. Epigenetic effects and oxidative stress, including that produced
secondary to cytotoxicity, may also contribute. Currently, the available database of
tetrachloroethylene-specific studies addressing these mechanisms are very limited, and merit
further exploration.
Cancer is a complex, multicausal process that is characterized by the acquisition and/or
activation of multiple critical traits. As described by Hanahan and Weinberg (2000). these traits
or hallmarks comprise six essential features: self-sufficiency in growth signals, insensitivity to
growth-inhibitory (antigrowth) signals, evasion of programmed cell death (apoptosis), limitless
replicative potential, sustained angiogenesis, and tissue invasion and metastasis. Epigenetic
changes (e.g., in the expression of microRNAs that negatively regulate gene expression by
targeting mRNA for translational repression or cleavage) appear to contribute to many of the
observed phenotypic alterations. The acquisition of these six capabilities can also be facilitated
by genomic instability, another feature of the cancer phenotype. A number of factors, such as
inflammation (Grivennikov et al.. 2010). diet and physiological factors (e.g., obesity (e.g..
obesity Park et al.. 2010)). can affect the tumor microenvironment in ways that advance these
features of tumor development. Studies of human hepatocarcinogenesis reveal significant
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-179 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
heterogeneity, with evidence of aberrant signaling in multiple, overlapping pathways involved in
cellular proliferation (e.g., EGF, HGF, RAS/mitogen-activated protein kinase), survival,
differentiation (e.g., Wnt, Hedgehog), and angiogenesis (e.g., VEGF, PDFG, FGF) [see recent
review by Hoshida et al. (2010)1. Other studies have provided support for a hypothesized role of
stem cells in hepatocarcinogenesis (Marquardt and Thorgeirsson. 2010). In contrast to the
stochastic cancer model, the cancer stem cell hypothesis posits a hierarchical model in which a
minor cell population possessing sternness undergoes epigenetic changes to generate
heterogeneous tumors (see review by Reva et al.. 2001). The potential cell types of origin of
liver cancer stem cells include mature hepatocytes possessing stem-like characteristics, as well as
circulating cells (Kim et al.. 2009) including bone-marrow derived stem cells (Marquardt and
Thorgeirsson. 2010). Such stem cells have been posited to play a role in liver development and
regeneration in addition to carcinogenesis (see review by Kung et al.. 2010). Thus, although
significant knowledge gaps remain, particularly with respect to the particular pathways and
processes necessary and sufficient for the disease to originate and develop, the etiology of
hepatocarcinogenesis appears complex.
Given the multiple metabolites and mechanisms that may contribute, and the known
complexity and heterogeneity in liver cancer development in general, it is unlikely that a single
causative metabolite, mechanism, pathway, or mode of action will be identified for
tetrachloroethylene-induced hepatocarcinogenesis. A single, linear sequence of key events does
not seem likely to explain the observed hepatocarcinogenicity, given the multiple cell types and
processes involved. Instead, a plausible hypothesis may be posited of multiple, contributing
mechanistic effects that may, in turn, be affected by multiple modifying factors. Accordingly,
the mechanisms described in this review are not intended to be interpreted as being mutually
exclusive. Altogether, the described mechanistic effects may aid in identifying sources of human
vulnerability, as well as informing the likelihood of other outcomes influenced by same
mechanisms, pathways, and biological processes. They may be informative of future analysis
integrating data on human —uptsream" biomarkers of hepatocarcinogenesis with chemically
induced perturbations. In this manner, the mechanistic data may be informative for addressing
the issues of cumulative assessment across exposures as well as overall population risk.
In summary, as noted by NRC (2010). there are significant gaps in the scientific
knowledge of mechanisms contributing to tetrachloroethylene-induced mouse liver cancer.
Multiple metabolites formed from tetrachloroethylene are toxic and carcinogenic in the liver.
Given this knowledge, and the known complexity and heterogeneity in liver cancer development
in general, the available evidence supports a hypothesis of multiple, contributing mechanistic
effects that may, in turn, be affected by multiple modifying factors.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-180 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
4 4 ESOPHAGEAL CANCER
Twelve epidemiologic studies reporting data on esophageal cancer and
tetrachloroethylene exposure were identified. This set of publications includes 10 cohort studies
(Andersen et al.. 1999; Blair et al.. 2003; Boice et al.. 1999; Calvert et al.. In Press; Chang et al..
2003; Lynge and Thygesen. 1990; Pukkala et al.. 2009; Selden and Ahlborg. 2011; Sung et al..
2007; Travier et al.. 2002) and three case-control studies of occupational exposures (Lynge et al..
2006; Siemiatvcki. 1991; Vaughan et al.. 1997). No studies of residential exposure through
contaminated drinking water were identified in the literature review. These 12 studies represent
the core studies evaluated by EPA, as described in more detail below. Two other cohort studies
included information on tetrachloroethylene but did not report risk estimates for esophageal
cancer (Anttila et al.. 1995; Radican et al.. 2008). and one case-control study did not observe any
cases exposed as a dry cleaner (Siemiatvcki. 1991). and so were not evaluated further. There is
some overlap in the study populations among these studies: Travier et al. (2002) used
occupational data from the Swedish national census and Lynge and Thygsen (1990) used a
similar design in Denmark; Andersen et al. (1999) and Lynge et al. (2006) expanded these
studies to include Denmark, Finland, Norway in addition to Sweden, and Pukkala et al. (2009)
added Iceland to this set. Appendix B reviews the design, exposure-assessment approach, and
statistical methodology for each study. All studies were of the inhalation route, of occupational
exposure, and, except for the case-control study of Vaughan et al. (1997). unable to quantify
tetrachloroethylene exposure.
4.4.1. Consideration of Exposure-Assessment Methodology
Many studies examine occupational title as dry cleaner, launderer, and presser as
surrogate for tetrachloroethylene, given its widespread use from 1960 onward in the United
States and Europe (Andersen et al.. 1999; Blair et al.. 2003; Calvert et al.. In Press; Lynge et al..
2006; Lynge and Thygesen. 1990; Pukkala et al.. 2009; Travier et al.. 2002). Six studies
conducted in Nordic countries are based on either the entire Swedish population or on combined
populations of several Nordic countries; strengths of these studies are their use of job title as
recorded in census databases and ascertainment of cancer incidence using national cancer
registries (Andersen et al.. 1999; Lynge et al.. 2006; Lynge and Thygesen. 1990; Pukkala et al..
2009; Travier et al.. 2002). Studies examining mortality among U.S. dry-cleaner and laundry
workers (Blair et al.. 2003; Calvert et al.. In Press) are of smaller cohorts than most Nordic
studies, with fewer observed esophageal cancer events.
The exposure surrogate in studies of dry-cleaners and laundry workers is a broad
category containing jobs of differing potential for tetrachloroethylene exposure. Thus, these
studies have a greater potential for exposure misclassification bias compared to studies with
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-181 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
exposure potential to tetrachloroethylene assigned by exposure matrix approaches. Three studies
used additional information pertaining to work environment to refine the exposure classification
(Calvert et al.. In Press; Lynge et al.. 2006; Selden and Ahlborg. 2011). Selden and Ahlborg
(2011) obtained information about the dry-cleaning establishment (e.g., washing techniques,
chemicals used, number of employees, and work history of individual employees) in a
questionnaire sent to businesses in Sweden in the 1980s. Lynge et al. (2006). using job title
reported in the 1970 Census, identified subjects based on occupational code of—launri/ and dry-
cleaning worker" or industry code of—handry and dry cleaning." Additional information to
refine this occupational classification was sought for incident cancer cases, including esophageal
cancer, within this defined cohort. Five controls, matched to the cases by country, sex, age, and
calendar period, were also included in the study. The additional information included
handwritten task information from the census forms from Denmark and Norway, pension
databases in Denmark and Finland, and next-of-kin interviews in Norway and Sweden.
Exposure classification categories were dry cleaner (defined as dry cleaners and supporting staff
if employed in business of <10 workers), other job titles in dry cleaning (launderers and
pressers), unexposed (job title reported on 1970 Census was other than in dry cleaning), or
unclassifiable (information was lacking to identify job title of subject). The unclassifiable
category represented 18 of 72 esophageal cancer cases (25%) and 108 out of 567 controls (19%).
The study by Calvert et al. of unionized dry cleaners in the United States included an analysis of
subjects who worked for one or more years before 1960 in a shop known to use
tetrachloroethylene as the primary solvent (Calvert et al.. In Press; Ruder et al.. 1994. 2001).
The cohort was stratified into two groups based on the level of certainty that the worker was
employed only in facilities using tetrachloroethylene as the primary solvent; tetrachloroethylene-
only and tetrachloroethylene plus. There were 6 esophageal cancer deaths among this subset
(n = 618) of the study subjects. Calvert et al. also presented risk estimates by exposure duration
and by latent periods for the full set of study subjects. Two additional studies used an exposure
metric for semiquantitative or quantitative exposure within a dry-cleaning setting. Blair et al.
(2003) used an exposure metric for semiquantitative cumulative exposure, and the case-control
study of Vaughan et al. (1997) used a job exposure matrix (JEM) with quantitative exposure
assessment for dry-cleaning and laundry jobs.
Two other cohorts with potential tetrachloroethylene exposure in manufacturing settings
have been examined. These studies include aerospace workers in the United States (Boice et al..
1999) and electronic factory workers in Taiwan (Chang et al.. 2003; Sung et al.. 2007). Boice et
al. (1999) used an exposure assessment based on a job-exposure matrix to classify exposures. In
contrast, the exposures in the Taiwan studies included multiple solvents, tetrachloroethylene
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-182 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
exposure was not linked to individual workers, and cohorts included both white- and blue-collar
workers (Chang et al.. 2003; Sung et al.. 2007).
In summary, with respect to exposure-assessment methodologies, five studies with
esophageal cancer data assigned tetrachloroethylene exposure to individuals using a
semiquantitative surrogate or a job exposure matrix (Blair et al.. 2003; Boice et al.. 1999;
Vaughan et al.. 1997). information about working conditions obtained through a questionnaire
(Selden and Ahlborg. 2011). or a classification of the cohort by certainty of tetrachloroethylene
exposure(Calvert et al.. In Press). One other study based on occupational census data sought
additional data for use in refining potential exposure within dry-cleaning settings (Lynge et al..
2006). The relative specificity of these exposure-assessment approaches strengthens their ability
to identify cancer hazards compared to studies with broader and less sensitive exposure-
assessment approaches.
4.4.2. Summary of Results
All studies evaluated by EPA reported estimated relative risks based on a small number
of observed events; 35 or fewer deaths/incident cases in cohort studies (Andersen et al.. 1999;
Blair et al.. 2003; Boice et al.. 1999; Calvert et al.. In Press; Chang et al.. 2003; Lynge et al..
2006; Lynge and Thygesen. 1990; Sung et al.. 2007; Travier et al.. 2002). except Pukkala et al.
(2009) whose esophageal cancer findings are based on 95 exposed subjects. The few esophageal
cancers in cohort studies and exposed cases in case-control studies, contribute to reduced
statistical power and limited ability to inform an evaluation of tetrachloroethylene exposure,
particularly for esophageal cancer whose estimated incidence is lower than for other cancer sites
discussed in Section 4 (Edwards et al.. 2010).
The largest cohort study observed an SIR estimate of 1.18 (95% CI: 0.96, 1.46) (Pukkala
et al.. 2009). Some evidence for an association between esophageal cancer risk and ever having
a job title of dry cleaner or laundry worker or routine exposure to tetrachloroethylene is also
found in cohort studies1 whose effect estimates are based on fewer observed events and that
carry lesser weight in the analysis. As expected, the magnitude of the point estimate of the
association reported in these studies is more variable than in the larger study. The smaller cohort
studies reported risks of 0.74 (95% CI: 0.41, 1.25), 1.16 (95% CI: 0.14, 4.20), 1.32 (95% CI:
0.94, 1.85), 1.47 (95% CI: 0.54, 3.21), 2.2 (95% CI: 1.15, 3.3) and 2.44 (95% CI: 1.4, 3.97) in
Lynge and Thygsen (1990). Sung et al. (2007). Travier et al. (2002). Boice et al. (1999). Blair et
al. (2003). and Calvert et al. , respectively (see Table 4-23). The 10-year follow-up period in
Lynge and Thygsen (1990) may represent an insufficient latent period with respect to the
1 Andersen et al. (19991 is not included in this summary of the data from the individual studies because it was
updated and expanded in the analysis by Pukkala et al. (20091.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-183 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
development of cancer, reducing the study's sensitivity compared to Pukkala et al. (2009). whose
follow-up was >15 years.
The case-control study of Lynge et al. (2006) reported an odds ratio of 0.76
(95% CI: 0.34, 1.69) for dry cleaners, with 8 exposed cases, compared to no exposure. In this
study, job title could not be classified for 25% of the cases and 19% of the controls. The odds
ratio for risk cancer in this -ttnclassifiable" group was 2.04 (95% CI: 0.91, 4.62). Lynge et al.
(2006) carried out sensitivity analyses using different assumptions regarding the true
classification for these subjects. In these analyses, the odds ratio for the association between dry
cleaner and esophageal cancer was 0.66 (95% CI: 0.30, 1.45) assuming all unclassified subjects
were unexposed and 1.19 (95% CI: 0.67, 2.21) assuming all unclassified subjects were dry
cleaners. One other case-control study that adopted a JEM approach to assign exposure reported
an odds ratio of 6.5 (95% CI: 0.6, 68.9) and 0.9 (0.1, 10.0) for overall exposure to
tetrachloroethylene, based on two and one exposed case, respectively, for squamous cell
carcinoma and adenocarcinoma of the esophagus (Vaughan et al.. 1997).
Several studies had been previously identified based on the relative strengths of their
exposure-assessment methodology. The results from these studies are mixed. Lynge et al.
(2006) reported no evidence of an increased risk among individuals classified as dry cleaners,
with relative risks of 0.76, but a higher risk was seen in the —unclassifible" group (RR: 2.04).
Selden and Ahlborg (2011) reported similar but slightly higher relative risks for laundry workers
(SIR: 1.56) compared with dry cleaners (SIR: 1.25). In contrast, data from other studies with
relatively strong exposure-assessment methods provide more evidence of an effect, with relative
risks of 1.47 (Boice et al.. 1999: routine exposure). 2.2 (Blair et al.. 2003). and 2.68 (Calvert et
al.. In Press): tetrachloroethylene-only workers), and 6.4 (Vaughan et al.. 1997).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-184 DRAFT-DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-23. Summary of human studies on tetrachloroethylene exposure and esophageal cancer
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Cohort Studies
Biologically monitored workers
Anttila et al. (1995)
All subjects
Not reported
849 Finnish men and women, blood PCE [0.4 |imol/L in females and 0.7
(imol/L in males (median)], follow-up 1974-1992, external referents (SIR)
Aerospace workers (Lockheed)
Boice et al. (1999)
Routine exposure to PCE
1.47 (0.54, 3.21)
6
77,965 (ri = 2,631 with routine PCE exposure and n = 3.199 with
intermittent-routine PCE exposure), began work during or after 1960,
worked at least 1 yr, follow-up 1960-1996, job exposure matrix without
quantitative estimate of PCE intensity, 1987-1988 8-h TWA PCE
concentration (atmospheric monitoring) 3 ppm [mean] and 9.5 ppm
[median], external reference for routine exposure (SMR) and internal
references (workers with no chemical exposures) for routine-intermittent
PCE exposure (RR)
Routine-Intermittent exposure to PCEa
Duration of exposure
Never exposed
1.0b
28
<1 yr
1.0 (0.30,3.34)
3
1-4 yr
0.79 (0.27, 2.50)
4
>5 yr
0.91 (0.13, 1.60)
3
p for trend
p = 0.07
Electronic factory workers (Taiwan)
Chang et al. (2003); Sung et al. (2007)
All Subjects
86,868 (n = 70,735 female), follow-up 1985-1997, multiple solvents
exposure, does not identify PCE exposure to individual subjects, cancer
mortality, external referents (SMR) (Chans et al.. 2003);
Males
0
Females
0
63,982 females, follow-up 1979-2001, factory employment proxy for
exposure, multiple solvents exposures and PCE not identified to individual
subjects, cancer incidence, external referents, analyses lagged 10 yr (SIR)
(Suns et al.. 2007)
Females
1.16(0.14, 4.20)
2
Aircraft maintenance workers from Hill Air Force Base
Radican et al. (2008)
Any PCE exposure
Not reported
10,461 men and 3,605 women (total n = 14,066, n = 10,256 ever exposed to
mixed solvents, 851 ever-exposed to PCE)), employed at least 1 yr from
1952 to 1956, follow-up 1973-2000, job exposure matrix (intensity),
internal referent (workers with no chemical exposures) (RR)
a -r
>1
to 0\
2 ^
3
00
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-23. Summary of human studies on tetrachloroethylene exposure and esophageal cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Dry cleaner and laundry workers
Andersen et al. C1999)
All laundry worker and dry cleaners
0.91 (0.57, 1.40)
21
29,333 men and women identified in 1960 Census (Sweden) or 1970
Census (Denmark, Finland, Norway), follow-up 1971-1987 or 1991, PCE
not identified to individual subjects, external referents (SIR)
Males
0.82 (0.33, 1.70)
7
Females
0.97 (0.53, 1.62)
14
Blair et al. (^2003^
All subjects
2.2(1.15,3.3)
26
5,369 U.S. men and women laundry and dry-cleaning union members
(1945-1978), follow-up 1979-1993, semiquantitative cumulative exposure
surrogate to dry clean solvents, cancer mortality, external referents (SMR)
Semiquantitative exposure score
Little to no exposure
2.1 (0.9,4.4)
7
Medium to high exposure
2.2(1.2,3.5)
16
Lvnge and Thvgsen C1990)
All laundry worker and dry cleaners
0.74 (0.41, 1.25)
14
10,600 Danish men and women, 20-64 yr old, employed in 1970 as
laundry worker, dry cleaners and textile dye workers, follow-up
1970-1980, external referents (SIR)
Males
0.62 (0.23, 1.35)
6
Females
0.88 (0.38, 1.73)
8
Pukkala et al. (2009)
Launderer and dry cleaner
1.18(0.96, 1.46)
95
Men and women participating in national census on or before 1990, 5
Nordic countries (Denmark, Finland, Iceland, Norway, Sweden), 30-64 yr,
follow-up 2005, occupational title of launderer and dry cleaner in any
census, external referents (SIR)
Male
0.99 (0.66, 1.44)
28
Female
1.29 (1.00, 1.64)
67
a -r
>1
to 0\
2 ^
3
00
On
~n
H
O
O
-------
Table 4-23. Summary of human studies on tetrachloroethylene exposure and esophageal cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Calvert et al. (In Press)
All subjects
2.44 (1.4, 3.97)
16
1,704 U.S. men and women dry-cleaning union member in CA, IL, MI, NY
follow-up 1940-2004 (618 subjects worked for one or more years prior to
1960 only at shops where PCE was the primary cleaning solvent, identified
as PCE-only exposure), cancer mortality (SMR)
Exposure duration/time since 1st employment
<5 yr/<20 yr
0
<5 yr/>20 yr
2.16 (0.85,4.54)
5
>5 yr/<20 yr
0
>5 yr/>20 yr
4.78 (2.68, 7.91)
11
PCE-only subjects
2.68 (0.98, 5.83)
6
Selden and Ahlborg (2011)
Dry-cleaners and laundry workers (females)
1.33 (0.43, 3.10)
5
9,440 Swedish men (n = 2,810) and women (n = 9,440) in 461 washing and
dry-cleaning establishments, identified by employer in mid-1980s,
employed 1973-1983, follow-up 1985-2000, exposure assigned using
company self-reported information on PCE usage—PCE (dry cleaners and
laundries with a proportion of PCE dry cleaning), laundry (no PCE use),
and other (mixed exposures to PCE, CFCs, TCE, etc.), external referents
(SIR). No observed cases in males
PCE (females)
1.25 (0.26, 3.25)
3
Laundry (females)
1.56 (0.19, 5.65)
2
Travier et al. (2002)
All subjects, 1960 or 1970 Census in laundry
and dry cleaner occupation and industry
1.32 (0.94, 1.85)
34
Swedish men and women identified in 1960, 1970, or both Censuses as
laundry worker, dry cleaner, or presser (occupational title) or in the
laundry, ironing, or dyeing industry, follow-up 1971-1989, separates
launders and dry cleaners form pressers, external referents (SIR)
All subjects in 1960 and 1970 in laundry and dry
cleaner occupation and industry
0.34 (0.05, 2.39)
1
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-23. Summary of human studies on tetrachloroethylene exposure and esophageal cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Case-Control Studies
Nordic Countries (Denmark, Finland, Norway, Sweden)
Lv nac et al. (2006)
Unexposed
1.00
41
Case-control study among 46,768 Danish, Finnish, Norwegian, and
Swedish men and women employed in 1960 as laundry worker or dry
cleaner, follow-up 1970-1971 to 1997-2001, 72 incident esophageal
cancer cases, 6 controls per case randomly selected from cohort matched on
country, sex, age, calendar period at diagnosis time, occupational task at
1970 Census proxy for exposure, RR adjusted for matching criteria
Dry cleaner
0.76 (0.34, 1.69)b
8
Assume unclassifiable exposed as dry cleaner
1.19(0.67, 2.21)b
26
Assume unclassifiable unexposed
0.66 (0.30, 1.45)b
59
Other in dry-cleaning
1.22 (0.41, 3.63)b
5
Unclassifiable
2.04 (0.91, 4.62)b
18
Dry cleaner, employment duration, 1964-1979
Unexposed
1.0
41
<1 yr
0
2-4 yr
1.20 (0.19, 2.29)b
1
5-9 yr
0.66 (0.19, 2.29)b
3
>10 yr
0.70 (0.20, 2.49)b
3
Unknown
1.65 (0.18, 14.98)b
1
Montreal, Canada
Siemiatvcki (1991)
Launderers and dry cleaners
Histologically confirmed esophageal cancers (n = 99), 1979-1985, 35-70
yr, population control group and cancer control group, in-person interviews,
occupational title, OR adjusted age, family income, and cigarette index,
90% CI
Any exposure
(0.0, 2.4)
0
Substantial exposure
(0.0,4.3)
0
a -r
>1
to 0\
2 ^
3
00
00
~n
H
o
O
-------
a
o
<
s?
cs
>!
o
^3
§•
rs
st
3
>!
sT
St
o
>1
>!
09
o
>!
Si
Si
s
O
Sfc
I
§•
K) On
2
Table 4-23. Summary of human studies on tetrachloroethylene exposure and esophageal cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Washington State (United States)
Vaughan et al. (1997)
Squamous cell carcinoma
Esophageal cancer cases (404 cases), 1983-1987, 20-74 yr, 724 population
controls, in-person interview, occupational title and JEM for PCE, blinded
exposure assessment, OR adjusted for age, sex, education, study period,
alcohol consumption and cigarette smoking
Ever exposed to PCE (probable exposure)
6.4 (0.6, 68.9)
2
Cumulative PCE exposure (possible exposure)
1-29 ppm-yr
11.9(1.1, 124.0)
2
30+ ppm-yr
0
Adenocarcinoma
Ever exposed to PCE (Probable exposure)
0.9(0.1, 10.0)
1
Tor Boice et al. (1999). relative risks for employment duration from Poisson regression with internal referents of factory workers not exposed to any solvent and
with adjustment for date of birth, date first employed, date of finishing employment, race and sex.
bIn Lynge et al. (20061. odds ratio from logistic regression adjusted for country, sex, age, calendar period at time of diagnosis.
JEM = job-exposure matrix, PCE = tetrachloroethylene.
oo
VO
~n
H
O
O
-------
Establishment of an exposure or concentration-response relationship can add to the
weight of evidence for identifying cancer hazard, but only limited data pertaining to exposure-
response relationships for esophageal cancer and tetrachloroethylene exposure are available.
Five studies reported risk by exposure categories using exposure duration (Boice et al.. 1999;
Calvert et al.. In Press; Lynge et al.. 2006) or a semiquantitative or quantitative surrogate (Blair
et al.. 2003; Vaughan et al.. 1997). However, Boice et al. (1999) and Vaughan et al. (1997) were
based on relatively few observed cases, with <5 cases in individual exposure categories, greatly
limiting the usefulness of these exposure-response examinations. Boice et al. (1999) presented a
formal statistical test of linear trend (p = 0.07) for exposure duration and esophageal cancer
deaths among workers with routine or intermittent exposure; three of the 10 esophageal cancer
deaths in this group had exposure durations 5 years or longer (RR: 0.91, 95% CI: 0.13, 1.60, with
an internal comparison group of factory workers not exposed to any solvents as the referent).
This analysis included subjects whose exposure was infrequent and likely of lesser certainty than
subjects identified as having routine exposure. The overall SMR for any tetrachloroethylene
exposure in this study was 1.47 (95% CI:0.54, 3.21). Both exposed cases in Vaughan et al.
(1997) were identified with lower cumulative exposure, 1-29 ppm-years (OR: 11.9, 95% 1.1,
124.0) compared to no cases with 30+ ppm-years. Effect estimates in one of the two larger
studies that examined exposure duration was not suggestive of a trend (Lynge et al.. 2006) (see
Table 4-23). However, all 16 exposed esophageal deaths in Calvert et al. had >20 years since
first employment, with effect estimates of 2.16 (95% CI: 0.85, 4.54) and 4.78 (95% CI: 2.68,
7.91) for <5 years and >5 years exposure duration, respectively. Sixteen of the 26 esophageal
cancer deaths in Blair et al. (2003) had medium-to-high cumulative exposure to dry-cleaning
solvents with an effect estimate of 2.2 (95% CI: 1.2, 3.5).
Only Vaughan et al. (1997) directly evaluated possible effects due to smoking or alcohol,
which are risk factors for the squamous cell histologic type of esophageal cancer; all other
studies lacked control for these potential confounders. Both Calvert et al. and Blair et al.
(2003) noted that the magnitude of the risks for esophageal cancer was greater than could be
explained by smoking alone; any smoking effect was estimated to contribute to no more than a
20% increase in risk. This suggests a further contribution from another risk factor, such as
occupational exposure. The incidence of esophageal cancer is generally higher for non-
Caucasian males than for Caucasian males (Blot and McLaughlin. 1999; Brown et al.. 2001). In
contrast, Calvert et al. observed similar SMRs for esophageal cancer across all race-sex
groupings (supplementary table at http://www.cdc.gov/niosh/dc-mort.html). suggesting the
contribution of another factor such as occupational exposure. However, the inability to adjust for
potential effects of alcohol use in cohort studies is an uncertainty.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-190 DRAFT-DO NOT CITE OR QUOTE
-------
In conclusion, the epidemiologic data provide suggestive but limited evidence pertaining
to tetrachloroethylene exposure and esophageal cancer risk. The SIR in the only large cohort
study (n = 95 cases) was 1.18 (95% CI: 0.96, 1.46) (Pukkala et al.. 2009). The point estimates of
the association in seven of eight smaller studies, four studies with specific exposure assessments
and four other studies with less precise assessments, were between 1.16 and 2.44 (Blair et al..
2003; Boice et al.. 1999; Calvert et al.. In Press; Lynge and Thygesen. 1990; Pukkala et al..
2009; Selden and Ahlborg. 2011; Sung et al.. 2007; Travier et al.. 2002). Two small case-control
studies with relatively high quality exposure-assessment approaches, Lynge et al. (2006) and
Vaughanetal. {, 1997, 631120} reported an odds ratio of 0.76 (95% CI: 0.34, 1.69) and of 6.4
(95% CI: 0.6, 68.9), respectively. Some uncertainties in these estimate arise from the lack of job
title information for 25% of the cases and 19% of the controls, and the variability in the results
from the sensitivity analysis using different assumptions regarding the correct classification of
individuals in this group in Lynge et al. (2006) and the small numbers of exposed cases in
Vaughan et al. (1997). One of the two larger studies examining exposure-response suggested a
positive relationship, with SMRs of 2.16 (95% CI: 0.85, 4.54) and 4.78 (95% CI: 2.68, 7.91) for
durations of <5 years and >5 years, respectively (Calvert et al.. In Press), but no dose-response
trend, or overall suggestion of an increased risk, was seen in Lynge et al. (2006). An
approximate twofold risk was seen in the little-to-no and in the medium-to-high-exposure groups
in Blair et al. (2003). None of the cohort studies can exclude possible confounding from alcohol
and smoking—risk factors for squamous cell carcinoma of the esophagus. Based on smoking
rates in blue-collar workers, the twofold risk estimate reported in Calvert et al. (In Press) and
Blair et al. (2003) was higher than that attributable to smoking.
4 5 LUNG AND RESPIRATORY CANCER
Nineteen epidemiologic studies reporting data on lung cancer and tetrachloroethylene
exposure were identified. This set of studies includes 12 cohort or nested case-control studies
within a cohort (Andersen et al.. 1999; Anttila et al.. 1995; Blair et al.. 2003; Boice et al.. 1999;
Calvert et al.. In Press; Chang et al.. 2003; Lynge and Thygesen. 1990; Pukkala et al.. 2009;
Selden and Ahlborg. 2011; Sung et al.. 2007; Travier et al.. 2002; Ji et al., 2005), 6 case-control
studies of occupational exposures (Brownson et al.. 1993; Consonni et al.. 2010; MacArthur et
al.. 2009; Pohlabeln et al.. 2000; Richiardi et al.. 2004; Siemiatvcki. 1991). and one case-control
study of residential exposure through contaminated drinking water (Paulu et al.. 1999). Some of
these studies represent overlapping populations. For example, Travier et al. (2002) and Lynge
and Thygsen (1990) used occupational data from Sweden and Denmark, respectively; Andersen
et al. (1999) included Denmark, Finland, and Norway in addition to Sweden, and Pukkala et al.
(2009) added Iceland to the study population. Additionally, nonsmoking cases in Richiardi et al.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-191 DRAFT-DO NOT CITE OR QUOTE
-------
(2004). whose lung cancer cases included both smokers and nonsmokers, were included in the
International Agency for Research on Cancer (IARC) multicenter study of lung cancer among
nonsmokers (Pohlabeln et al.. 2000). These studies represent the core studies evaluated by EPA,
as described in more detail below. One other cohort study included information on
tetrachloroethylene but did not report risk estimates for lung cancer (Radican et al.. 2008). Also,
one other lung cancer case-control study did not identify any cases as a dry cleaner or launderer
(Zeka et al.. 2006) and was not evaluated further. Appendix B reviews the design, exposure-
assessment approach, and statistical methodology for each study. Most studies were of the
inhalation route, of occupational exposure, and unable to quantify tetrachloroethylene exposure.
4.5.1. Consideration of Exposure-Assessment Methodology
Most of these studies examine occupational titles such as dry cleaner, launderer, and
presser as surrogates for tetrachloroethylene, given its widespread use from 1960 onward in the
United States and Europe (Andersen et al.. 1999; Blair et al.. 2003; Brownson et al.. 1993;
Calvert et al.. In Press; Consonni et al.. 2010; Ji et al.. 2005a. b; Ji and Hemminki. 2005a. b, c;
Lynge and Thygesen. 1990; Mac Arthur et al.. 2009; Pohlabeln et al.. 2000; Pukkala et al.. 2009;
Richiardi et al.. 2004; Selden and Ahlborg. 2011; Siemiatvcki. 1991; Travier et al.. 2002; Zeka et
al.. 2006). Seven studies conducted in Nordic countries are based on either the entire Swedish
population or on combined populations of several Nordic countries; the strengths of these studies
are their use of job titles as recorded in census databases and ascertainment of cancer incidence
using national cancer registries (Andersen et al.. 1999; Ji et al.. 2005a. b; Ji and Hemminki.
2005a. b, c; Lynge et al.. 2006; Lynge and Thygesen. 1990; Pukkala et al.. 2009; Selden and
Ahlborg. 2011; Travier et al.. 2002). Studies examining mortality among U.S. dry-cleaner and
laundry workers (Blair et al.. 2003; Calvert et al.. In Press) are of smaller cohorts than the Nordic
studies, with fewer observed lung cancer events.
The exposure surrogate in studies of dry-cleaners and laundry workers is a broad
category containing jobs of differing potential for tetrachloroethylene exposure. Thus, these
studies have a greater potential for exposure misclassification bias compared to studies with
exposure potential to tetrachloroethylene assigned by exposure matrix approaches applied to
individual subjects. Three studies used additional information pertaining to work environment to
refine the exposure classification. Selden and Ahlborg (2011) obtained information about the
dry-cleaning establishment (e.g., washing techniques, chemicals used, number of employees, and
work history of individual employees) in a questionnaire sent to businesses in Sweden in the
1980s. Blair et al. (2003) used an exposure metric for semiquantitative cumulative exposure
within the dry-cleaning setting. The study by Calvert et al. of unionized dry cleaners in the
United States included an analysis of subjects who worked for one or more years before 1960 in
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-192 DRAFT-DO NOT CITE OR QUOTE
-------
a shop known to use tetrachloroethylene as the primary solvent (Calvert et al.. In Press; Ruder et
al.. 1994. 2001). The cohort was stratified into two groups based on the level of certainty that
the worker was employed only in facilities using tetrachloroethylene as the primary solvent;
tetrachloroethylene-only and tetrachloroethylene plus. Twenty-six of the 77 observed lung
cancer deaths were among this subset (n = 618) of the study subjects.
Four other cohorts with potential tetrachloroethylene exposure in manufacturing settings
have been examined. These studies include aerospace workers in the United States (Boice et al..
1999). workers, primarily in the metal industry, in Finland (Anttila et al.. 1995) and electronic
factory workers in Taiwan (Chang et al.. 2005; Sung et al.. 2007). Boice et al. (1999) used an
exposure assessment based on a job-exposure matrix, and Anttila et al. (1995) used biological
monitoring of tetrachloroethylene in blood to assign potential tetrachloroethylene exposure to
individual subjects. In contrast, the exposures in the Taiwan studies included multiple solvents,
and tetrachloroethylene exposure was not linked to individual workers. These cohorts also
included white-collar workers, who had an expected lower potential for exposure (Chang et al..
2003; Sung et al.. 2007).
Paulu et al. (1999) is a case-control study that examined residential proximity to drinking
water sources contaminated with tetrachloroethylene in Cape Cod, MA. This study used an
exposure model incorporating leaching and characteristics of the community water distribution
system to assign a household relative dose of tetrachloroethylene.
In summary, with respect to exposure-assessment methodologies, six studies with lung
cancer data assigned tetrachloroethylene exposure to individuals within the study using
biological monitoring data (Anttila et al.. 1995). a job exposure matrix (Boice et al.. 1999). a
semiquantitative metric (Blair et al.. 2003). an exposure model (Paulu et al.. 1999). additional
details pertaining to work environment (Selden and Ahlborg. 2011). or a classification of the
cohort by certainty of tetrachloroethylene exposure (Calvert et al.. In Press). The relative
specificity of these exposure-assessment approaches strengthens their ability to identify cancer
hazards compared to studies with broader and less sensitive exposure-assessment approaches.
The least sensitive exposure assessments are those using very broad definitions such as working
in a plant or factory (Chang et al.. 2003; Sung et al.. 2007).
4.5.2. Summary of Results
Lung cancer is a relatively common cancer, and six of the cohort studies of dry-cleaners
and laundry workers evaluated by EPA reported estimated relative risks based on 100 or more
deaths/incident cases (Andersen et al.. 1999; Blair et al.. 2003; Ji and Hemminki. 2005a; Pukkala
et al.. 2009; Selden and Ahlborg. 2011; Travier et al.. 2002); Pukkala et al. (2009) was the
largest study, with 965 incident lung cancers. Two other cohort studies, Lynge and Thygsen
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-193 DRAFT-DO NOT CITE OR QUOTE
-------
(1990) and Calvert et al., observed 60 and 77 lung cancers, respectively. In contrast, the number
of exposed cases in the case-control studies ranged from 3 cases each of small cell and
adenocarcinoma histological subtypes in MacArthur et al. (2009) to 30 (all histological types) in
Brownson et al. (1993). The three cohort studies with exposure assessment specific to
tetrachloroethylene observed 5 incident cancer cases, 46 lung cancer deaths, and 125 lung cancer
deaths in Anttila et al. (1995). Boice et al. (1999). and Blair et al. (2003). respectively. The
geographic-based case-control study of Paulu et al. (1999) observed 33 of the 326 lung cancer
cases living in a residence receiving tetrachloroethylene contaminated water, and only 5 of these
cases were identified as highly exposed.
The seven1 cohort studies with findings based on 50 or more events observed a
standardized incidence ratio estimate between 1.15 and 1.4 for the association between lung
cancer risk and ever having a job title of dry-cleaner or laundry worker, each with relatively tight
95% CIs (see Table 4-24). These estimates by study were 1.15 (95% CI: 1.02, 1.31) in Travier et
al. (2002). 1.2 (0.9, 1.5) in Lynge and Thygsen (1990). 1.26 (95% CI: 1.18, 1.34) in Pukkala et
al. (2009). 1.32 (1.07, 1.60) in Ji et al. (2005a. b; 2005a. b, c), 1.32 (95% CI: 1.20, 1.45) in
Selden and Ahlborg (2011). 1.31 (1.04, 1.64) in Calvert et al. , and 1.4 (1.1, 1.6) in Blair et al.
(2003). respectively. Selden and Alhborg (2011) examined separately subjects working in a dry
cleaner using tetrachloroethylene (potential tetrachloroethylene exposure) and laundry workers,
subjects without potential tetrachloroethylene exposure. The standardized incidence ratios were
1.16 (95% CI: 0.89, 1.51) and 1.62 (95% CI: 1.15, 2.19) for dry cleaners and for laundry
workers, respectively.
In addition to the large cohort studies, evidence also comes from cohort and case-control
studies whose effect estimates are based on fewer observed events. Smaller studies that do not
also have a more sensitive or specific exposure metric carry lesser weight in the analysis. As
expected, the magnitude of the point estimate of the association reported in these studies is more
variable than in the larger studies: one study reported an odds ratio estimate below 1.0
(Siemiatvcki. 1991). four studies reported a relative risk estimate between 1.0 and 1.3 (Boice et
al.. 1999: Consonni et al.. 2010: MacArthur et al.. 2009: Paulu et al.. 1999), three studies
reported relative risks between 1.8 and 2.0 (Anttila et al.. 1995: Brownson et al.. 1993: Pohlabeln
et al.. 2000). and two studies reported odds ratios estimates over 2.0 (MacArthur et al.. 2009:
Richiardi et al.. 2004). Except for the estimate from Brownson et al. (1993) (OR: 1.8, 95% CI:
1.1, 3.0) and McArthur et al. (2009) (small cell carcinoma, OR: 3.55, 95% CI: 1.13, 11.17), all of
the 95% CIs of these estimates overlap 1.0.
1 Andersen et al. (19991 is not included in this summary of the data from the individual studies because it was
updated and expanded in the analysis by Pukkala et al. (20091.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-194 DRAFT-DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
°S o
§ £.
3 ^ a-
O
s:
>!
§•
rs
st
5
>!
Si
Table 4-24. Summary of human studies on tetrachloroethylene exposure and lung cancer
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Cohort studies
Biologically monitored workers
Anttila et al. (1995)
All subjects
1.92 (0.62, 4.48)
5
849 Finnish men and women, blood PCE [0.4 (imol/L in females and 0.7
|imol/L in males (median)], follow-up 1974-1992, external referents
(SIR)
Aerospace workers (Lockheed)
Boice et al. (1999)
Routine exposure to PCE
1.08 (0.79, 1.44)
46
77,965 (n = 2,631 with routine PCE exposure and n = 3.199 with
intermittent-routine PCE exposure), began work during or after 1960,
worked at least 1 yr, follow-up 1960-1996, job exposure matrix without
quantitative estimate of PCE intensity, 1987-1988 8-h TWA PCE
concentration (atmospheric monitoring) 3 ppm [mean] and 9.5 ppm
[median], external reference for routine exposure (SMR) and internal
references (workers with no chemical exposures) for routine-intermittent
PCE exposure (RR)
Routine-Intermittent exposure duration to PCE
0
1.0a
288
<1 yr
1.15 (0.80, 1.66)
33
1-4 yr
1.09 (0.80, 1.48)
51
>5 yr
0.71 (0.49, 1.02)
36
p-valuc for linear trend
p = 0.02
Electronic factory workers (Taiwan)
Chang et al. (2003); Sung et al. (2007)
All Subjects
0.97 (0.69, 1.33)
38
86,868 (n = 70,735 female), follow-up 1979-1997, multiple solvents
exposure, does not identify PCE exposure to individual subjects, cancer
mortalitv. external referents (SMR) (Chans et al.. 2003);
Males
0.90 (0.48, 1.53)
13
Females
1.01 (0.65, 1.49)
25
63,982 females, follow-up 1979-2001, factory employment proxy for
exposure, multiple solvents exposures and PCE not identified to
individual subjects, cancer incidence, external referents, analyses lagged
10 vr rSIR1) rSune et al.. 20071
Females
0.92 (0.67, 1.23)
46
Aircraft maintenance workers from Hill Air Force Base
Radican et al. (2008)
Any PCE exposure
Not reported
10,461 men and 3,605 women (total n = 14,066, n = 10,256 ever exposed
to mixed solvents, 851 ever-exposed to PCE)), employed at least 1 yr
from 1952 to 1956, follow-up 1973-2000, job exposure matrix
(intensity), internal referent (workers with no chemical exposures) (RR)
a -r
>1
to 0\
2 S3
3
vo
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
°S o
§ £.
3 ^ a-
O
s:
>!
§•
rs
st
5
>!
Si
Table 4-24. Summary of human studies on tetrachloroethylene exposure and lung cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Dry-cleaner and laundry workers
Andersen et al. (1999)
All laundry worker and dry cleaners
1.19 (1.07, 1.34)
313
29,333 men and women identified in 1960 Census (Sweden) or 1970
Census (Denmark, Finland, Norway), follow-up 1971-1987 or 1991,
PCE not identified to individual subjects, external referents (SIR)
Males
1.24 (1.05, 1.46)
141
Females
1.16 (1.00, 1.35)
172
Blair et al. (2003)
All subjects
1.4(1.1, 1.6)
125
5,369 U.S. men and women laundry and dry-cleaning union members
(1945-1978), follow-up 1979-1993, semiquantitative cumulative
exposure surrogate to dry clean solvents, cancer mortality, external
referents (SMR)
Semiquantitative exposure score
Little to no exposure
1.0 (0.7, 1.4)
34
Medium to high exposure
1.5 (1.2, 1.9)
78
Ji et al. C2005a. b): Ji and Hemminki (2005a. b. c)
Laundry workers and dry cleaners in 1960 Census
1.32(1.20, 1.46)
403
9,255 Swedish men and 14,974 Swedish women employed in 1960 (men)
or 1970 (women) as laundry worker or dry cleaner, follow-up
1961/1970-2000, PCE not identified to individual subjects, external
referent (SIR) and adjusted for age, period and socioeconomic status
Males
1.36 (1.20, 1.54)
247
Females
1.26 (1.07, 1.47)
156
Laundry workers and dry cleaners in both 1960 and 1970 Censuses
Males
Not reported
Females
Not reported
Laundry workers and dry cleaners in 1960, 1970, and 1980 Censuses
Males
Not reported
Females
Not reported
a -r
>1
to 0\
2 S3
3
vo
On
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
°S o
§ £.
3 ^ a-
O
s:
>!
§•
rs
st
5
>!
Si
Table 4-24. Summary of human studies on tetrachloroethylene exposure and lung cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Lvnac and Thvascn (1990)
All laundry worker and dry cleaners
1.2(0.9, 1.5)
60
10,600 Danish men and women, 20-64 yr old, employed in 1970 as
laundry worker, dry cleaners and textile dye workers, follow-up
1970-1980, external referents (SIR)
Males
1.1 (0.8, 1.7)
28
Females
0.3 (0.9, 1.8)
32
Pukkala et al. (2009)
Launderer and dry cleaner
1.26 (1.18, 1.34)
965
Men and women participating in national census on or before 1990, 5
Nordic countries (Denmark, Finland, Iceland, Norway, Sweden), 30-64
yr, follow-up 2005, occupational title of launderer and dry cleaner in any
census, external referents (SIR)
Male
1.28 (1.15, 1.42)
353
Female
1.25 (1.15, 1.35)
612
Calvert et al. (In Press)
All subjects
1.31 (1.04, 1.64)
77
1,704 U.S. men and women dry-cleaning union member in CA, IL, MI,
NY follow-up 1940-2004 (618 subjects worked for one or more years
prior to 1960 only at shops where PCE was the primary cleaning solvent,
identified as PCE-only exposure), cancer mortality (SMR)
Exposure duration/time since 1st employment
<5 yr/<20 yr
0.63 (0.21, 1.44)
4
<5 yr/>20 yr
1.75 (1.33,2.26)
32
>5 yr/<20 yr
1.27 (0.55, 2.50)
6
>5 yr/>20 yr
1.08 (0.75, 1.51)
26
PCE-only subjects
1.25 (0.82, 1.83)
26
Selden and Ahlborg (2011)
Dry-cleaners and laundry workers
1.32(1.07, 1.60)
100
9,440 Swedish men (n = 2,810) and women (n = 9,440) in 461 washing
and dry-cleaning establishments, identified by employer in mid-1980s,
employed 1973-1983, follow-up 1985-2000, exposure assigned using
company self-reported information on PCE usage—PCE (dry cleaners
and laundries with a proportion of PCE dry cleaning), laundry (no PCE
use), and other (mixed exposures to PCE, CFCs, TCE, etc.), external
referents (SIR)
PCE
1.16(0.89 1.51)
58
Males
1.30 (0.82, 1.94)
23
Females
1.09 (0.76, 1.51)
35
Laundry
1.62(1.15,2.21)
Males
1.60 (0.85, 2.74)
13
Females
1.63 (1.06, 2.39)
26
a -r
>1
to 0\
2 S3
3
vo
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
°S o
§ £.
3 ^ a-
O
s:
>!
§•
rs
st
5
>!
Si
Table 4-24. Summary of human studies on tetrachloroethylene exposure and lung cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Travier et al. (2002)
All subjects, 1960 or 1970 Census in laundry and
dry cleaner or related occupation and industry
1.15 (1.02, 1.31)
248
Swedish men and women identified as laundry worker, dry cleaner, or
presser (occupational title), in the laundry, ironing, or dyeing industry or
related industry in 1960 or 1970 (543,036 person-years); or, as laundry
worker, dry cleaner, or presser (occupational and job title) (46,933
person-years) in both censuses, follow-up 1971-1989, external referents
(SIR)
All subjects in 1960 and 1970 in laundry and dry
cleaner occupation and industry
1.20 (0.84, 1.72)
30
Case-Control Studies
Missouri, United States
Brownson et al. (1993)
Dry-cleaning industry
429 female primary lung cancer cases, 30-84 yr, 1986-1991, never
smokers or ex-smokers (>15 yr prior to diagnosis), identified from
Missouri Cancer Registry, 1,021 female population controls matched on
age, identified from state driver's licenses (<65 yr) or HFCA roles
(65-84 yr), telephone and in-person interview using questionnaire, dry
cleaner occupation or job title exposure surrogate, OR adjusted for age,
smoking, and history of previous lung disease
All subjects
1.8(1.1,3.0)
30
Lifetime nonsmokers
2.1 (1.2, 3.7)
23
Former smokers
1.1 (not reported)
7
Exposure duration
<1.125 yr
0.8 0.2, 1.7)
Not reported
>1.125 yr
2.9(1.5,5.4)
Not reported
Lombardy, Italy (EAGLE study)
Consonni et al. (2010)
Dry-cleaning industry
1,943 histologically or cytologically confirmed hospital lung cancer cases
in men and women, 35-79 yr, 2002-2005, and 2,116 population controls
matched on residence, sex, and age, in-person and self-administered
Males
Not reported
3
Females
1.26 (0.46, 3.41)
12
questionnaire, job title and industry coded to ISCO and ISIC surrogate
for exposure, dry-cleaning industry identified a priori suspected lung
hazard, OR adjusted for residential area, age, smoking and number of
jobs held
a -r
>1
to 0\
2 S3
3
vo
00
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
°S o
§ £.
3 ^ a-
O
s:
>!
§•
rs
st
5
>!
Si
Table 4-24. Summary of human studies on tetrachloroethylene exposure and lung cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
British Columbia, Canada
MacArthur et al. (2009)
Dry cleaner and launderer occupation
2,998 male histologically confirmed lung cancer cases, >20 yr,
1983-1990, 10,233 all-other sites-cancer controls matched on age and
diagnosis year, identified from British Columbia Cancer Registry, self-
administered questionnaire, job title and industry coded to Canadian SOC
and Canadian SIC as exposure surrogate, OR adjusted for smoking
duration, respondent status, and education
Squamous cell carcinoma
1.25 (0.47, 3.35)
4
Adenocarcinoma
1.28 (0.44, 3.70)
3
Small cell
3.55 (1.13, 11.17)
3
Large cell
0
International Lung Cancer Study (IARC Study) (France, Germany, Italy, Portugal, Spain,
Sweden, United Kingdom)
Pohlabeln et al. (2000)
All Centers
660 nonsmoking lung cancer cases, <75 yr, 1988-1994, 1,542
nonsmoking controls, 12 study centers in 7 countries, various sources of
nonsmoking controls (community based in 6 centers, hospital-based in 1
center, both sources in 5 centers), hospital controls with diseases not
related to smoking, in-person interview, job title and industry coded to
ISCO and ISIC exposure surrogate, dry-cleaning industry identified a
priori suspected lung hazard, OR adjusted for age and center
Dry-cleaning industry
Males
Not reported
1
Females
1.83 (0.98, 3.40)
19
Turin and Veneto Regions, Italy
Richiardi et al. (2004)
Dry Cleaners and Launderers
1,132 histologically or cytologically confirmed lung cancer cases, <75 yr,
1990-1991 or 1991-1992, population controls identified from population
registries and matched on sex and age, in-person interview, job title and
industry >6 mo duration coded to ISCO and ISIC exposure surrogate,
dry-cleaning industry identified a priori suspected lung hazard, OR
adjusted for age, study area, cigarette smoking, other tobacco product
use, and number of jobs. Cases and controls included in international
multicenter studv of Pohlaban et al. (2000)
Males
1.6 (0.2, 12)
3
Females
2.1 (0.8, 5.6)
9
a -r
>1
to 0\
2 S3
3
vo
vo
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
°S o
§ £.
3 ^ a-
O
s:
>!
§•
rs
st
5
>!
Si
Table 4-24. Summary of human studies on tetrachloroethylene exposure and lung cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Montreal, Canada
Siemiatvcki (1991)
Launderers and dry cleaners
857 histologically confirmed lung cell carcinoma cancer, 1979-1985,
35-70 yr, 533 population control group and 1,900 cancer control group,
in-person interviews, occupational title, OR adjusted age, family income,
Any exposure
0.8 (0.5, 1.5)b
12
Substantial exposure
0.6 (0.2, 1.4)b
5
ethnic origin, respondent status, cigarette smoking, and alcohol
consumption, 90% CI
International Lung Cancer Study (IARC Study) (Czech Republic, Hungary, Poland,
Romania, Russia, Slovakia, United Kingdom)
Zeka et al., 2006
Launderers and dry cleaners
223 hospital lung cancer cases, 20-74 yr, 1998-2002, lifetime
nonsmokers, identified from 16 hospitals or clinics in 7 countries,
hospital (14 centers) or population controls (2 centers) frequency-
matched on sex and age, in-person interview, industry and job title
exposure surrogate, dry-cleaning industry identified a priori suspected
lung hazard, OR adjusted for age, sex, and study center, with ETS
exposure included as additional covariate in some analyses
Males
Not reported
0
Females
Not reported
0
Geographic-Based Studies
Cape Cod, MA
Paulu et al. (1999)
Overall PCE exposure
1.1 (0.7, 1.7)
33
326 histologically confirmed lung cancer cases in males and females,
1983-1986, MA Cancer Registry, 2,236 population controls identified by
random digit dialing, vital records for deceased controls, and HCFA
records if >65 yr, telephone interview, algorithm of Webler and Brown
(1993) to estimate mass of PCE in drinkine water enterine residence was
surrogate exposure metric, OR adjusted for age of diagnosis or index
year, vital status at interview, sex, occupation exposure to PCE, other
solvents, and exposures associated with lung cancer, usual number of
cigarettes smoked, history of cigar/pipe use, living with a smoker
PCE RDD >90lh percentile
2.7(1.0, 11.7)
5
a -r
>1
to 0\
2 S3
3
to
o
o
~n
H
O
O
aReferent.
bIn Siemiatycki (19911. 90% CI.
CFC = chloroflourocarbon, HCFA = Health Care Financing Administration, ISCO = International Standard Classification of Occupation, ISIC = International
Standard Industry Classification, JEM = job-exposure-matrix, PCE = tetrachloroethylene, RDD = relative delivered dose, TCE = trichloroethylene, TWA = time-
weighted-average.
-------
1
2
3
4
5
6
7
8
9
10
11
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
Five occupational studies were identified as having a relatively strong exposure-
assessment methodology. The results from four of these studies provide support for an increased
risk in the dry-cleaning cohorts with a relative risk of 1.4 (95% CI: 1.1, 1.6) in Blair et al. (2003).
1.31 (95% CI: 1.04, 1.64) in Calvert et al. , and in other settings, a relative risk of 1.08 (95%
CI: 0.79, 1.44) in Boice et al. (1999) and 1.92 (95% CI: 0.62, 4.48) in Anttila et al., 1997. In
contrast, (Selden and Ahlborg. 2011) reported similar, but slightly higher, relative risks for
laundry workers compared with dry-cleaning workers in their study. Two studies of an
electronics factory using relatively weak exposure-assessment approaches (i.e., no classification
of individuals within the study) observed relative risks or SMRs of 0.97 (95% CI: 0.69, 1.33)
(Chang et al.. 2003) and 0.92 (95% CI: 0.67, 1.23) (Sung et al.. 2008).
Establishment of an exposure or concentration-response relationship can add to the
weight of evidence for identifying a cancer hazard, but only limited data pertaining to exposure-
response relationships for lung cancer and tetrachloroethylene exposure are available. Seven
studies presented risk estimates for increasing exposure categories: three studies using exposure
duration as a proxy (Boice et al.. 1999: Calvert et al.. In Press: Travier et al.. 2002) and four
studies with a semiquantitative exposure surrogate (Blair et al.. 2003: Brownson et al.. 1993:
Paulu et al.. 1999: Siemiatvcki. 1991). Boice et al. (1999) was the only study to present a formal
statistical test for trend and reported a statistically significant decreasing trend between lung
cancer risk estimates and duration among subjects with routine and intermittent
tetrachloroethylene exposure (p = 0.02). A monotonic increasing trend in risk estimates and
exposure surrogate was apparent in four studies (Blair et al.. 2003: Brownson et al.. 1993: Paulu
et al.. 1999: Travier et al.. 2002).
A know risk factor for lung cancer is cigarette smoking (NTP. 2005). Subjects in both
(Brownson et al.. 1993) and Pohlabeln et al. (2000) were either lifetime nonsmokers or ex-
smokers who had terminated smoking 15 years before cancer diagnosis, reducing any potential
role of confounding from smoking. Furthermore, in the case of Pohlabeln et al. (2000). the
inclusion of occasional smoking (ever smoked occasionally but fewer than 400 cigarettes total)
and exposure to tobacco smoking as possible confounders did not significantly affect the odds
ratio estimate and were not included in the final model. Statistical analyses in all other case-
control studies controlled for cigarette smoking (Consonni et al.. 2010: MacArthur et al.. 2009:
Paulu et al.. 1999: Richiardi et al.. 2004: Siemiatvcki. 1991). However, both (Brownson et al..
1993) and MacArthur et al. (2009) had a high percentage of surrogate or proxy respondents, 58
and 27%), respectively. Proxy respondents may have motivations to report or not report specific
exposures leading to differential information bias that could result in the relative risk estimate
towards or away from the null depending on whether controls were more or less likely to recall
or report such exposure than cases (Pearce et al.. 2007).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-201 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Direct examination of possible confounders is less common or feasible in cohort studies
relying on company-supplied or census work history data compared to case-control studies
where information is obtained from study subjects or their proxies. In cohort studies, however,
use of internal controls rather than an external referent group (e.g., national mortality rates) can
minimize effects of potential confounding due to smoking or socioeconomic status, because
exposed and referent subjects are drawn from the same target population. Only one of the
available cohort studies included an analysis using internal controls and reported a decreasing
trend between lung cancer and tetrachloroethylene exposure duration, p = 0.02(Boice et al..
1999). Blair et al. (2003) considered the potential effect of differences in the prevalence of
smoking in their study of laundry and dry-cleaning workers. Surveys from 1970 to 1990
indicated that smoking rates among dry cleaners were 5-10% higher than the general population.
With this level of difference, confounding from smoking is unlikely to result in a relative risk
greater than 1.2 but may explain most of the observed 40% excess in lung cancer. The
magnitude of relative risk estimates in cohort studies of dry-cleaners and laundry workers
(Calvert et al.. In Press; Ji and Hemminki. 2005b; Lynge and Thygesen. 1990; Pukkala et al..
2009; Selden and Ahlborg. 2011; Travier et al.. 2002) is similar to or less than that of Blair et al.
(2003) and suggests smoking may contribute to the observed association.
In conclusion, the epidemiologic data provide limited evidence pertaining to
tetrachloroethylene exposure and lung cancer risk. The results from seven large cohort studies of
dry cleaners are consistent with an elevated lung cancer risk of 10-40%). Similar results were
seen in four of the five occupational studies that were identified as having a relatively strong
exposure-assessment methodology (Blair et al.. 2003; Calvert et al.. In Press)(Anttila et al., 1997;
Boice et al., 2003). However, (Selden and Ahlborg. 2011) observed similar, but slightly higher,
relative risks for laundry workers compared with dry-cleaning workers in their study. These
studies were unable to control for potential confounding from cigarette smoking; however, and
the magnitude of the association in these studies is consistent with that expected, assuming the
prevalence of smoking among dry-cleaners and laundry workers was slightly higher (e.g., 10%>
higher) than among the general population. Features of the selection of study participants and
study analysis in the available case-control studies reduce the potential for confounding by
smoking, however. Two case-control studies were limited to either nonsmokers or ex-smokers
who had ceased smoking 15 years before diagnosis (Brownson et al.. 1993; Pohlabeln et al..
2000). Both of these studies indicate an approximate twofold increased risk with a history of
work in the dry-cleaning industry (OR: 1.8, 95%> CI: 1.1, 3.0, in Brownson et al. (1993); and OR:
1.83, 95%o CI: 0.98, 3.40, among women in Pohlabeln et al., 2010). The other case-control
studies adjusted for smoking history, and the results for these (somewhat smaller studies) are
similar to the previously cited estimates. The available data pertaining to an exposure-response
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-202 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
gradient are mixed (Blair et al.. 2003; Boice et al.. 1999; Brownson et al.. 1993; Calvert et al.. In
Press; Paulu et al.. 1999; Travier et al.. 2002).
4 6 IMMUNOTOXICITY, HEMATOLOGIC TOXICITY, 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 an altered immune
response that precedes the development of clinically expressed diseases. This section discusses
effects relating to immunotoxicity and hematotoxicity. It also discusses evidence pertaining to
tetrachloroethylene in relation to lymphoid tissue cancers, including childhood leukemia.
4.6.1. Human Studies
4.6.1.1. Noncancer Immune and Hematologic Effects
Adverse effects on the immune system resulting from chemical exposure fall within the
following principal domains: immunosuppression (host resistance), immunostimulation,
autoimmunity, and allergy-hypersensitivity. Various immunologic measurements (e.g., T-cell
counts, immunoglobulin [Ig] E levels, specific autoantibodies) may provide evidence of an
altered immune response that may subsequently be related to risk of clinically expressed diseases
such as infections, asthma, or systemic lupus erythematosus. Tetrachloroethylene exposure via
air or water may result in immune-mediated organ-specific or systemic effects, as described in a
case report of hypersensitivity pneumonitis in a 42-year-old female dry-cleaner worker (Tanios
et al.. 2004). Another case report described severe fatigue, weight loss, myalgia, arthralgia,
cardiac arrhythmia, decreased T-cell count, high-titer (1:160) antinuclear antibodies, and
neurological symptoms that were linked to chemical sensitivity to tetrachloroethylene in a
municipal water supply (Rea et al.. 1991).
4.6.1.1.1. Immunologic and hematologic parameters
Byers et al. (1988) provide data pertaining to immune function from 23 family members
of leukemia patients in Woburn, Massachusetts. In 1979, testing of the wells in this town
revealed that the water in two of the wells was contaminated with a number of solvents,
including tetrachloroethylene (21 ppb) and trichloroethylene (267 ppb) (as cited in Lagakos et
al.. 1986). These wells had been in operation from 1964 to 1979. Byers et al. collected serum
samples in May and June of 1984 and in November of 1985. They determined the total
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-203 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
lymphocyte counts and lymphocyte subpopulations (CD3, CD4, CD8), and the CD4:CD8 ratio 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 autoantibodies (antismooth muscle, antiovarian,
antinuclear, antithyroglobulin, and antimicrosomal antibodies) in the family member samples
and compared the results with laboratory reference values. 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
antinuclear antibodies (compared to <5% expected based on the reference value). Because the
initial blood sample was taken in 1984, and because of the considerable mixture of exposures
that occurred in this setting, it is not possible to determine the patterns at a time nearer to the
time of the exposure, or to infer the exact role of tetrachloroethylene in alterations of the
immunologic parameters.
Other studies have examined immunological parameters in dry-cleaning workers in the
Czech Republic (Andrys et al.. 1997) and in Egypt (Emara et al.. 2010) (see Table 4-25).
Andrys et al. (1997) included 21 dry-cleaning workers (20 women) and 16 office workers in the
dry-cleaning plant (14 women) and compared them to reference values based on samples from
blood donors and —healthy persons in thegane region" (n = 14-311, depending on the test). The
mean ages of the exposed workers and office controls were 45.7 years and 31.9 years,
respectively; no information was provided on the age or sex distribution of the reference
controls. The tests included measures of immunoglobulin (Ig) A, IgG, IgM, and IgE levels,
complement (C3 and C4) levels, phagocyte activity, C-reactive protein, a-macroglobulin,
T-lymphocytes, and a blast transformation test. Several differences were observed between the
exposed workers and the office workers (e.g., higher levels of serum complement C3 and C4,
and of salivary IgA in the exposed), and between the exposed workers and the reference controls
(reduced T-lymphocytes, higher phagocytic activity, higher C3 levels in exposed). However,
there were also many differences noted between the office workers and the reference group
(including reduced T-lymphocytes in office workers). The lack of information about the
reference group adds to the difficulty in interpreting these results.
Table 4-25. Immune and hematological parameters in studies of dry-
cleaning workers or tetrachloroethylene exposure in children
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-204 DRAFT-DO NOT CITE OR QUOTE
-------
Study details
Measure(s)
Results
Authors
Adults
Czech Republic, period not
reported. 21 dry-cleaning workers
(20 women; mean age 45.7 yr); 16
office workers in the dry-cleaning
plant (14 women; mean age
31.9 yr); reference values based on
samples from blood donors and
—healty persons in the same
region" (n = 14-311, depending on
the test)
Ig (IgA, IgG, IgM) levels,
complement (C3 and C4)
levels, phagocyte activity,
C-reactive protein,
a-macroglobulin,
T-lymphocytes
Higher levels serum complement
C3 and C4, salivary IgA in
exposed workers compared with
office workers. Reduced
T-lymphocytes, higher phagocytic
activity, higher C3 levels in
exposed workers compared with
reference controls. Reduced
T-lymphocytes in office workers
compared with reference controls
Andiys et al.
(1997)
Egypt, period not reported. 40 adult
men (ages 20-38 yr), dry-cleaning
workers; 40 healthy male controls
(matched by age and smoking
history): n = 20 in 4 groups
(controls, never smoked; controls,
smoked; PCE-exposed, never
smoked; PCE exposed, smoked).
Amount and duration of smoking
similar among exposed and
nonexposed. Mean years of PCE
exposure 7 yr. Blood PCE levels in
exposed: 1,685 ng/L
RBC counts
RBC counts and hemoglobin levels
decreased with exposure. No
difference in MCV, MCH, or
MCHC
Emara et al.
(2010)
WBC counts
Total white cell and lymphocyte
counts increased with exposure.
No difference in eosinophils,
monocytes, or platelets
lymphocyte subpopulations
(CD3+, CD4+, CD8+,
CD3+CD16CD56+, CD19+
cells)
CD4+ and CD8+ T-lymphocytes
and CD3+CD16CD56+ NK cells
increased with exposure
Ig levels (IgA, IgE, IgG,
IgM)
IgE increased with exposure. No
difference in IgA, IgG, or IgM
levels across groups
serum and lymphocytic
interferon-y and
interleukin-4
Interleukin-4 levels increased with
exposure. No differences with
interferon-y
Germany, 1995-1996. 121
children (ages 36 mo), selected
based on high risk profile for
allergic diseases, blood sample and
indoor air sampling (child's
bedroom) of 26 volatile organic
chemicals (4 wk around age 36 mo)
IgE levels
no association between PCE
measures and total IgE or IgE-
specific allergen antibodies
Lehmann et
al. (2001)
Germany 1997-1999. 85
newborns, cord blood and indoor air
sampling (child's bedroom) of 28
volatile organic chemicals (4 wk
immediately after birth)
CD3 T-cell subpopulations
from cord blood
Decreased interferon-y cells
No association with interleukin-4,
interleukin-2, or tumor necrosis
factor-a
Lehmann et
al. (2002)
Ig = immunoglobulin; MCV = mean corpuscular volume; MCH = mean corpuscular hemoglobin; MCHC = mean
corpuscular hemoglobin concentration; RBC = red blood cells; WBC = white blood cells.
1 Emara et al. (2010) examined immunological and hematologic parameters in 80 men,
2 ages 20-38 years, in Tanta City, Egypt. Forty men were dry-cleaning workers, with a mean
3 duration of work of 7 years. They were matched by age and smoking history to 40 healthy
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-205 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
controls from the same area. The study, thus, included four groups, each with 20 men: controls
who had never smoked; controls who were smokers; tetrachloroethylene-exposed workers who
had never smoked; and tetrachloroethylene-exposed workers who were smokers. The amount
smoked and duration of smoking were similar in the exposed and nonexposed groups (mean:
17.9 and 17.5 cigarettes per day, respectively; mean: 4.5 and 5.0 years smoking, respectively).
Tetrachloroethylene levels were measured at five sites within each worksite, and blood levels of
tetrachloroethylene were also measured in all study participants. The mean air level was <140-
ppm tetrachloroethylene, and mean blood levels were 1,681 and 1,696 |ig/L among nonsmoking
and smoking workers (compared with 0.11 |ig/L in each of the control groups), respectively.
Blood samples were obtained from each study participant for measurement of a differential
blood cell count, serum Ig levels (IgA, IgE, IgG, and IgM), and interferon-y and interleukin-4
levels in serum and lymphocytes.
Red blood cell counts and hemoglobin levels were decreased with exposure (p < 0.05 for
smoking-stratified comparisons), but there was no difference in mean corpuscular volume, mean
corpuscular hemoglobin, or mean corpuscular hemoglobin concentration across groups (Emaraet
al.. 2010). In contrast, total white cell counts and total lymphocytes increased significantly with
exposure (p < 0.05 for smoking-stratified comparisons). There was no difference in eosinophils,
monocytes, or platelets counts across groups. Neutrophil counts were increased in smokers
compared with nonsmokers but did not differ by tetrachloroethylene-exposure group. CD4+ and
CD8+ T-lymphocytes and natural killer (CD3+CD16CD56+) cells were increased in smoking and
nonsmoking exposed workers (p < 0.05), but CD3+ T-lymphocytes were only increased in the
exposed smoking group. This study demonstrated statistically significant effects of
tetrachloroethylene exposure on hematological parameters including decreased red blood cell
counts, increased white blood cells counts, total lymphocytes, and specific T- and NK cell
subpopulations.
Th2 cytokines (e.g., interleukin-4) stimulate production of IgE, and Thl cytokines (e.g.,
interferon-y) act to inhibit IgE production. The results from Emara et al. (2010) indicate that
tetrachloroethylene exposure results in an increase in serum and lymphocytic interleukin-4
levels, as well as increased IgE levels (p < 0.05 for smoking-stratified comparisons). As
determined from Figure 5 of Emara et al. (2010). the mean levels were approximately 90, 160,
170, and 195 IU/mL in nonexposed nonsmokers, nonexposed smokers, exposed nonsmokers, and
exposed smokers, respectively (p < 0.05 for smoking-stratified comparisons). No difference was
seen in IgA, IgG, or IgM levels across groups.
Two studies examined variation in cytokines and in IgE levels in children (Lehmann et
al.. 2001: Lehmann et al.. 2002) (see Table 4-25). Lehmann et al. (2001) examined IgE levels
and cytokine-producing cells (interferon-y, tumor necrosis factor-a, and interleukin-4) in relation
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-206 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
to indoor levels of volatile organic compounds among children (age 36 months) selected from a
birth cohort study in Leipzip, Germany. The hypothesis underlying this work is that a shift in
Thl to Th2 cytokine profile is a risk factor for IgE-mediated allergic disease in children (Tang et
al.. 1994; Warner et al.. 1994). 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 a 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. (2001) 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
highest exposures were seen for limonene (median: 19.1 (j,g/m3), a-pinene (median: 16.3 (j,g/m3),
and toluene (median: 13.3 (J,g/m3). The median exposure of tetrachloroethylene was 2.5 (J,g/m3
(0.87 [ig/m3 and 5.1 (J,g/m3 for the 25th and 75th percentiles, respectively). The only strong
correlation (r > 0.3) between tetrachloroethylene and the other volatile organic compounds
measured was a correlation of 0.72 with trichloroethylene. 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-grass, birch tree). There was no association between tetrachloroethylene
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.
Another study by Lehmann et al. (2002) examined the relationship between indoor
exposures to volatile organic compounds and T-cell subpopulations measured in cord blood of
newborns (see Table 4-25). 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 birth
(a period that 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
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 (j,g/m3), and the median exposure of tetrachloroethylene was 3.4 (J,g/m3
(1.8 |_ig/m3 and 7.3 (J,g/m3 for the 25th and 75th percentiles, respectively). Flow cytometry was
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-207 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
used to measure the presence of CD3 T-cells obtained from the cord blood labeled with
antibodies against interferon-y, tumor necrosis factor-a, interleukin-2, and interleukin-4.
Tetrachloroethylene was the only one of the measured volatile organic compounds that was
associated with a reduced level of interferon-y. In the univariate analysis, the median
percentages of interferon-y cells were 3.6 and 2.6% in the groups that were below the 75th
percentile and above the 75th percentile of tetrachloroethylene exposure, respectively. The odds
ratio between high (above the 75th percentile) tetrachloroethylene exposure and reduced (less
than the 25th percentile) levels of interferon-y cells was 2.9 (95% CI: 1.0-8.6), adjusting for
family history of atopy, gender, and smoking history of the mother during pregnancy. There was
no association between tetrachloroethylene exposure and interleukin-4 cells, but naphthalene and
methylcyclopentane were associated with elevated levels of interleukin-4 cells.
4.6.1.1.2. Immune-related conditions and diseases
Immunosuppression. In 1982, Lagakos et al. (1986) conducted a telephone survey of
residents of Woburn, Massachusetts, collecting information on residential history and history of
14 types of medically diagnosed conditions. 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. 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 history of kidney and urinary tract disorders
(primarily kidney or urinary tract infections) and with lung and respiratory disorders (asthma,
chronic bronchitis, or pneumonia). There are no other human data that characterize the effects of
tetrachloroethylene-only exposure on immunosuppression, as measured by increased
susceptibility to infections.
Allergy and hypersensitivity. Allergy and hypersensitivity, as assessed with measures of
immune system parameters or immune function tests (e.g., asthma, atopy) in humans, have not
been extensively studied with respect to the effects of tetrachloroethylene. Delfino et al. (2003a;
2003b) examined the exacerbation of asthmatic symptoms following exposure to volatile organic
compounds that occurred due to variation in air quality over a 3-month period in 1999-2000 in
Los Angeles. This study included daily repeated exposures to ambient air pollutants and peak
expiratory flow rates over a 3-month period in 21 children (17 males and 4 females) of Hispanic
origin, ages 10-16 years; an additional child participated in the ambient air but not in the exhaled
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-208 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
air portion of the study. Daily diaries were used to record severity of symptoms and asthmatic
episodes. Exposure metrics included exhaled breath measures and ambient levels of eight
volatile organic compounds (benzene, methylene chloride, styrene, toluene, m,p-y,ylene,
o-xylene, /?-dichlorobenzene, and tetrachloroethylene) and eight criteria pollutant gases. An
association between criteria air pollutants and subsequent symptoms of asthma in children in the
Los Angeles area suggests an increased risk of adverse health outcomes with exposure to SO2
and NO2 (Delfino et al.. 2003 a). Although ambient levels of tetrachloroethylene were associated
with bothersome asthma symptoms (OR: 1.37, [95% CI: 1.09, 1.71]) per an interquartile range
change), this association was reduced with the adjustment for SO2 or NO2 (Delfino et al.. 2003a).
In the 21 children who participated in the peak expiratory flow measurements, the mean breath
level of tetrachloroethylene was 4.40 ng/L (SD: 10.77 ng/L), the mean ambient level was 3.52
(SD: 2.17) ng/L, and the correlation between the same-day measures was 0.31 (p < 0.01: Delfino
et al.. 2003b). There was little relation between asthma symptoms and exhaled breath levels of
tetrachloroethylene. The mean exhalation levels of tetrachloroethylene were 2.50 and 2.69 ng/L,
respectively, in the two groups of asthma symptoms (none or not bothersome; bothersome and
more severe). Stronger associations were reported between asthma symptoms and some of the
other volatile organic chemicals, specifically for benzene, toluene, m,p-x.ylene.
Autoimmune disease. In the 1970s, recognition 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) prompted research
pertaining to the role of organic solvents in autoimmune diseases. Exposure to the broad
categories of solvents, organic solvents, or chlorinated solvents has been associated with a two-
to threefold increased risk of systemic sclerosis (scleroderma) in epidemiologic studies
summarized in a recent meta-analysis (Aryal et al.. 2001) and in subsequent studies (Garabrant et
al.. 2003: Maitre et al.. 2004). Similar results were seen in studies of other systemic autoimmune
diseases including undifferentiated connective tissue disease (Lacev et al.. 1999). rheumatoid
arthritis (Lundberg et al.. 1994: Sverdrup et al.. 2005). and antineutrophil-cytoplasmic antibody
(ANCA)-related vasculitis (Beaudreuil et al.. 2005: Lane et al.. 2003). In contrast, 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).
As described in the preceding paragraph, the epidemiologic data in relation to the role of
solvents, as a broad category, in systemic autoimmune diseases, vary among these conditions.
Much more limited data are available pertaining to specific solvents, including
tetrachloroethylene, and risk of autoimmune diseases. One case report describes a condition
similar to vinyl-chloride induced scleroderma in a man who worked as a presser in a dry-
cleaning plant, and who also helped clean the tetrachloroethylene-containing drums on a weekly
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-209 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
basis (Sparrow. 1977). Another case report describes a localized scleroderma in a man who had
worked with tetrachloroethylene as a metal degreaser, with workplace exposures reported to be
between 10-25 ppm (Hinnen et al.. 1995; in German). Among 279 cases with connective tissue
disease, Goldman (1996) observed a higher frequency of individuals who reported employment
as a dry cleaner among systemic sclerosis patients (4 of 33) compared with patients with other
connective tissue diseases (1 of 246; p < 0.01). Similar patterns were seen with self-reported
history of tetrachloroethylene exposure (3 of 33 systemic sclerosis patients compared with 2 of
246 other patients, p < 0.01), but the author noted the difficulty in obtaining this type of
information.
Two registry-linkage studies from Sweden of rheumatoid arthritis (Lundberg et al.. 1994)(Li et
al., 2008) and three case-control studies of undifferentiated connective tissue disease (Lacev et
al.. 1999). scleroderma (Garabrant et al.. 2003). and antineutrophil-cytoplasmic antibody
(ANCA) related diseases (Beaudreuil et al.. 2005) provide data concerning dry-cleaning work or
tetrachloroethylene exposure (see Table 4-26). As expected in population-based studies, the
exposure prevalence is low, with approximately 4% of controls reporting work in dry cleaning
and 1% reporting exposure to tetrachloroethylene. The observed associations are generally weak
for the broad classification of laundry and dry-cleaning work, with odds ratios for dry cleaning of
1.0 in the largest study of rheumatoid arthritis (Li et al., 2008) and 1.4 in two studies of
scleroderma (Garabrant et al.. 2003) and undifferentiated connective tissue disease (Lacev et al..
1999). None of the individual studies are statistically significant. The studies from Sweden
linking occupational census data to risk of rheumatoid arthritis (Lundberg et al.. 1994)(Li et al.,
2008) are also limited by the difficulty in defining time of diagnosis for this disease based on
hospitalization data. The results seen for the exposure to tetrachloroethylene in the three studies
that attempted this kind of assessment were more varied (Beaudreuil et al.. 2005; Garabrant et
al.. 2003; Lacev et al.. 1999). Only the study of ANCA-related diseases resulted in an elevated
odds ratio, but again, this estimate was somewhat imprecise (OR: 2.0. 95% CI: 0.6. 6.9;
Beaudreuil et al.. 2005). These studies are clearly limited by the low prevalence of and difficulty
in accurately characterizing occupational exposure to tetrachloroethylene in population-based or
clinical settings.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-210 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-26. Immune-related conditions in studies of dry cleaning or
tetrachloroethylene exposure in humans3
Condition and study details
Results
Authors
Rheumatoid arthritis
Sweden (13 counties), hospitalized 1981-1983,
896 male cases, 629 female cases; population
comparison (total 370,035 men, 140,139
women), ages 35-74. Registry linkage to 1960
and 1970 Census occupation data
launderers and dry cleaning
men: 1 exposed cases;
OR: 0.8 (95% CI: 0.1-5.0)
women: 7 exposed cases;
OR: 1.5 (95% CI: 0.7-3.2)
Lundberg et al.
(1994)
Sweden, hospitalized 1964-2004 (men) or 1970
to 2004 (women). 13,280 male cases and 14,509
female cases; population comparison (full
population), ages >30 yr, Registry linkage to
1960 or 1970 Census occupation data for men
and women, respectively
launderers and dry cleaning
men: 57 exposed cases;
OR: 0.8 (95% CI: 0.6-1.0)
women: 204 exposed cases;
OR: 1.0 (95% CI: 0.8-1.1)
Li et al., 2008
Other autoimmune diseases
Undifferentiated connective tissue disease,
Michigan and Ohio, diagnosed 1980-1991
(Michigan) 1980-1992 (Ohio). 205 cases, 2,095
population controls. Women, ages 18 and older.
Structured interview (specific jobs and materials;
jobs held 3 or more mo)
dry cleaning
cases: 4.3%, controls 3.8%
OR: 1.4 (95% CI: 0.68,2.8)
PCE
cases: 0%, controls 1% OR: 0.00
Lacey et al.
(1999)
Scleroderma, Michigan and Ohio. Diagnosed
1980-1991 (Michigan), 1980-1992 (Ohio). 660
cases, 2,227 population controls. Women, ages
18 and older. Structured interview (specific jobs
and materials; jobs held 3 or more mo)
dry cleaning
cases: 4.7%, controls 3.7%
OR: 1.4 (95% CI: 0.9,2.2)
PCE
self report cases: 1.1%, controls 1.0%
OR: 1.4 (95% CI: 0.6,3.4)
expert review cases: 0.8%, controls 0.8%
OR: 1.1 (95% CI: 0.4,2.9)
Garabrant et al.
(2003)
ANCA-related diseases,15 France. Diagnosed
1999-2000. 60 patients, 120 hospital controls,
men and women (50% each), mean age 61 yr
PCE
cases: 8.3%, controls 4.1%
OR: 2.0 (0.6-6.9)
Beaudreuil et
al. (2005)
Allergy and hypersensitivity
Exacerbation of asthma symptoms, Los Angeles,
1999-2000. 21 children (ages 10-16 yr), 3 mo
diaries, ambient levels and exhaled breath
measures of 8 volatile organic compounds and 8
criteria pollutants
Little evidence of an association between
ambient PCE exposure or exhaled PCE
measures and asthma symptoms
Delfino et al.
(2003a: 2003b)
" Includes case-control studies and cross-sectional studies but does not include case reports.
b 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).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-211 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
4.6.1.1.3. Summary of human noncancer immune and hematologic effects
The strongest study examining immunologic and hematologic effects of
tetrachloroethylene exposure in terms of sample size and use of an appropriately matched control
group is of 40 male dry-cleaning workers (mean exposure levels <140 ppm; mean duration:
7 years) by Emara et al. (2010). Statistically significant decreases in red blood cell count and
hemoglobin levels and increases in total white cell counts and lymphocyte counts were seen in
the exposed workers compared to age- and smoking-matched controls. In addition, increases in
several other immunological parameters, including T-lymphocyte and natural killer cell
subpopulations, IgE, and interleukin-4 levels were observed. These immunologic effects suggest
an augmentation of Th2 responsiveness. However, the limited available data from studies in
children (Delfino et al.. 2003a; Delfino et al.. 2003b; Lehmann et al.. 2001; Lehmann et al..
2002) do not provide substantial evidence of an effect of tetrachloroethylene exposure during
childhood on allergic sensitization or exacerbation of asthma symptomology. The observation of
the association between increased tetrachloroethylene exposure and reduced interferon-y in cord
blood samples may reflect a sensitive period of development and points to the current lack of
understanding of the potential immunotoxic effects of prenatal exposures. The available data
pertaining to risk of autoimmune disease in relation to tetrachloroethylene exposure are limited
by issues regarding ascertainment of disease incidence and exposure-assessment difficulties in
population-based studies.
4.6.1.2. Cancers of the Immune System, Including Childhood Leukemia
Forty-one epidemiologic studies report on adult lymphopoietic cancer and
tetrachloroethylene exposure. These publications include numerous cohort studies (Andersen et
al.. 1999; Anttila et al.. 1995; Blair et al.. 1998; Blair et al.. 2003; Boice et al.. 1999; Calvert et
al.. In Press; Cano and Pollan. 2001; Chang et al.. 2005; Ji and Hemminki. 2005b. 2006; Lynge
and Thygesen. 1990; Pukkala et al.. 2009; Radican et al.. 2008; Selden and Ahlborg. 2011;
Spirtas et al.. 1991; Sung et al.. 2007; Travier et al.. 2002). and case-control studies Ct Mannetie
et al.. 2008; Aschengrau et al.. 1993; Blair et al.. 1993; Clavel et al.. 1998; Costantini et al..
2008; Costantini et al.. 2001; Fabbro-Perav et al.. 2001; Gold et al.. 2010b; Kato et al.. 2005;
Lynge et al.. 2006; Mai one et al.. 1989; McLean et al.. 2009; Mester et al.. 2006; Miligi et al..
2006; Miligi et al.. 1999; Schenk et al.. 2009; Scherr et al.. 1992; Seidler et al.. 2007;
Siemiatvcki. 1991)Hardell et al., 1989;, and three geographical-based studies (Cohn et al..
1994b; Morton and Marianovic. 1984; Vartiainen et al.. 1993). Some of these papers represent
studies of related populations. For example, three papers examined cancer incidence or mortality
in a cohort of aircraft maintenance workers at an air force base in the United States, with follow-
up through 1982 (Spirtas et al.. 1991). 1990 (Blair etal.. 1998). and 2000 (Radican et al.. 2008).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-212 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Six papers examined cancer risk among occupational groups defined by census or employer-
provided data in Sweden (Cano and Pollan. 2001; Ji and Hemminki. 2005b. 2006; Lynge and
Thygesen. 1990; Selden and Ahlborg. 2011; Travier et al.. 2002). two papers were based on
census data from Sweden, Denmark, Finland, and Norway (Andersen et al.. 1999; Lynge et al..
2006). and a third paper added data from Iceland (Pukkala et al.. 2009). Four papers examined
different subsets of lymphopoietic cancers from a large population-based case-control study in
Italy (Costantini et al.. 2008; Costantini et al.. 2001; Miligi et al.. 2006; Miligi et al.. 1999).
Additionally, five epidemiologic studies—one cohort and four case-control—report on childhood
lymphopoietic cancer and tetrachloroethylene exposure (Costas et al.. 2002; Infante-Rivard et al..
2005; Lowengart et al.. 1987; Shu et al.. 1999; Sung et al.. 2008). Appendix B reviews the
design, exposure-assessment approach, and statistical methodology for each study; the adult
lymphopoietic cancer studies are also summarized in Table 4-27, and the childhood
lymphopoietic cancer studies are summarized in Table 4-28. Most studies were primarily of the
inhalation route, of occupational exposure, and, generally, unable to quantify tetrachloroethylene
exposure. Two studies of contaminated drinking water containing multiple solvents including
tetrachloroethylene were available (Cohn et al.. 1994b; Vartiainen et al.. 1993). Collectively,
these studies have varying sensitivities for identifying cancer hazards.
4.6.1.2.1. Adult lymphopoietic cancer: consideration of exposure assessment
Since the 1960s in Western Europe and the United States, the dry-cleaning industry has
accounted for about 90% of tetrachloroethylene consumption (Gold et al.. 2008; IARC. 1995;
Johansen et al.. 2005). with more infrequent and lower volume use of trichloroethylene and
CFC-113 for specialized cleaning (IARC. 1995). As described previously, eight publications
used occupational data derived from national census data or by the employer for one or more
northern European countries, focusing on dry cleaners and other laundry workers (Andersen et
al.. 1999; Cano and Pollan. 2001; Ji and Hemminki. 2005b. 2006; Lynge et al.. 2006; Lynge and
Thygesen. 1990; Pukkala et al.. 2009; Selden and Ahlborg. 2011; Travier et al.. 2002). Lynge et
al. (2006) used national databases and pension schemes to identify subjects as dry cleaners
versus other job titles held in 1970; however, these databases were not available for subjects
from two of the four countries (i.e., Norway and Finland), nor was information on a subject's
workplace and length of employment available for Swedish subjects. In the absence of national
databases, Lynge et al. (2006) collected this information through interviews, many with a
subject's next of kin. A higher likelihood for recall bias is possible with next of kin or proxy
information, particularly for knowledge of solvent exposures as shown by Boyle et al.(1992).
Additionally, workers who may have switched to jobs as dry cleaners after 1970 would be
misclassified using a classification system based on job held in 1970. Two smaller cohort
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-213 DRAFT-DO NOT CITE OR QUOTE
-------
1 studies examining mortality using cause of death data from death certificates were conducted
2 among laundry and dry-cleaning union members in the United States (Blair et al.. 2003; Calvert
3 et al.. In Press; Ruder et al.. 1994. 2001).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-214 DRAFT-DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
Cohort Studies
Biologically monitored Finnish workers
Anttila et al. (1995)
All subjects
Lymphopoietic
1.38 (0.28, 4.02)
3
849 men and women, blood PCE [0.4 |imol/L in females
and 0.7 |imol/L in males (median)], follow-up
1974-1992, cancer incidence, external referents (SIR).
Non-Hodgkin
3.76 (0.77, 11.0)
3
Multiple myeloma
0
0.38 exp
Leukemia
Not reported
Aerospace workers (Lockheed)
Boice. et al. (1999)
Routine exposure to PCE
Lymphopoietic
1.13 (0.62, 1,89)a
14
77,965 (n = 2,631 with routine PCE exposure and
n = 3.199 with intermittent-routine PCE exposure),
began work during or after 1960, worked at least 1 yr,
follow-up 1960-1996, JEM without quantitative
estimate of PCE intensity, 1987-1988 8-h TWA PCE
concentration (atmospheric monitoring) 3 ppm [mean]
and 9.5 ppm [median], mortality, external referents for
routine exposure (SMR) and internal referents (workers
with no chemical exposures) for routine-intermittent
PCE exposure (RR).
Non-Hodgkin
1.70 (0.73, 3.34)
8
Hodgkin
0
0.98 exp
Multiple myeloma
0.40 (0.01, 2.25)
1
Leukemia
0.55 (0.18, 1.29)
5
Routine-intermittent PCE-exposure duration
0 yr
Non-Hodgkin
1.0b
32
<1 yr
1.25 (0.43, 3.57)
4
1-4 yr
1.11 (0.46, 2.70)
6
>5 yr
1.41 (0.67, 3.00)
10
Test for trend
p > 0.20
a -r
>1
to 0\
2 ^
3
¦t*
to
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
0 yr
Multiple myeloma
1.0b
24
Boice et al. (1999)(continued)
<1 yr
0.46 (0.06, 3.348)
1
1-4 yr
1.13 (0.38, 3.35)
4
>5 yr
0.24 (0.03, 1.84)
1
Test for Trend
p < 0.01
Electronic factory workers (Taiwan)
Chang et al. (2005); Sung et al. (2007)
All subjects
Lympho- and hemato-
poietic
0.67 (0.42, 1.01)
22
86,868 (n = 70,735 female), follow-up 1979-1997,
multiple solvents exposure, does not identify PCE
exposure to individual subjects, lympho- and
hematopoietic cancer incidence, external referents
(SIR)(Chane et al.. 2005); 63.982 females. follow-uD
1979-2001, factory employment proxy for exposure,
PCE not identified to individual subjects, leukemia
cancer incidence, external referents, analyses lagged 5 yr
(SIR1) CSune et al.. 20071
Males
0.73 (0.27, 1.60)
6
Females
0.65 (0.37, 1.05)
16
Females
Leukemia
0.78 (0.49, 1.17)
5
Aircraft maintenance workers from Hill Air Force Base
S Dirt as et al. (1991); Blair et al. (1998); Radican et al.
(2008)
Ever-exposed to PCE
14,066 (10,461 men and 3,605 women) (n = 10,256 ever
exposed to mixed solvents, 851 ever-exposed to PCE),
employed at least 1 yr from 1952 to 1956, follow-up to
2000, PCE used for parachute cleaning, JEM without
quantitative estimate of PCE intensity, mortality, internal
referent (workers with no chemical exposures) (RR).
Males
Non-Hodgkin
2.32 (0.75, 7.15)b
5
Females
2.35 (0.52, 10.71)b
2
Males
Multiple myeloma
1.71 (0.42, 6.91)b
3
Females
7.84 (1.43, 43.06)b
2
a -r
>1
to 0\
2 ^
3
¦t*
to
On
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
Dry-cleaner and laundry worker
Andersen et al. CI999)
All subjects
Lymphopoietic
1.0(0.87, 1.15)°
204
29,333 men and women identified in 1960 Census
Males
1.05 (0.79, 1.38)°
53
(Sweden) or 1970 Census (Denmark, Finland, Norway)
with occupation as launderers or dry cleaners, follow-up
1971-1987 or 1991, PCE not identified to individual
subjects, incidence, country-specific cancer rates referent
(SIR).
Females
0.98 (0.84, 1.16)°
151
All subjects
Non-Hodgkin
1.07 (0.86, 1.34)
82
Males
1.46 (0.96, 2.13)
27
Females
0.95 (0.71, 1.23)
55
All subjects
Hodgkin
1.34(0.81,2.10)
19
Males
0
0.4 exp
Females
1.88 (1.13,2.93)
19
All subjects
Multiple myeloma
1.0 (0.73, 1.34)
45
Males
1.38 (0.75,2.31)
14
Females
0.89 (0.60, 1.26)
31
All subjects
Leukemia
0.85 (0.65, 1.10)
58
Males
0.67 (0.35, 1.17)
12
Females
0.90 (0.66, 1.21)
46
a -r
>1
to 0\
2 ^
3
¦t*
to
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
Blair etal. (^2003^
All subjects
Non-Hodgkin
0.9 (0.5, 1.6)
12
5,369 U.S. men and women laundry and dry-cleaning
union members (1945-1978), follow-up 1979-1993,
PCE exposure potential higher for subcohort entering
union after 1960, semiquantitative cumulative exposure
surrogate to dry clean solvents, cancer mortality, external
referents (SMR).
Hodgkin lymphoma
2.0 (0.6, 4.6)
5
Multiple myeloma
0.8 (0.3, 1.6)
7
Leukemia
0.8 (0.4, 1.4)
12
Semiquantitative exposure score
Any exposure
Lympho- and hemato-
poietic
1.0(0.7, 1.3)
39
Little to no exposure
1.0 (0.6, 1.5)
18
Medium to high exposure
0.9 (0.5, 1.4)
17
Cano and Pollan (2001)
Males
Non-Hodgkin
1.76 (0.97, 3.17)d
11
Swedish men and women aged 25-64 yr reporting
occupation as —dunderers and dry cleaners" in 1970
Census, employed and counted in 1960 Census, follow-
up 1971-1989, NHL incidence from Swedish Cancer
Registry, PCE not identified to individual subjects, all
other occupations referent (RR).
1.85 (0.83, 4.12)e
6
Females
Not reported
Ji and Hemminki (2005b. 2006)
Males
Non-Hodgkinf'8
0.99 (0.75, 1.26)
59
9,255 men and 14,974 women reporting laundry and dry-
cleaning work 1970 Swedish Census, follow-up
1960-2002, cases identified from Swedish Cancer
Registry, PCE not assigned to individual subjects, cancer
incidence from Swedish Cancer Registry, Swedish
cancer rates referent (SIR).
Females
1.05 (0.82, 1.32)
67
Males
Multiple myelomaf
0.99 (0.66, 1.38)
52
Females
1.07 (0.75, 1.45)
36
Males
Leukemia8
0.84 (0.62, 1.90)
47
Females
1.30 (1.03, 1.60)
80
a -r
>1
to 0\
2 ^
3
¦t*
to
00
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
Lvnse and Thvsesen (1990)
All subjects
Non-Hodgkin
1.03 (0.44, 2.02)
8
10,600 men and women reporting work in dry cleaner
and laundries in Swedish 1970 Census, follow-up
1970-1980, job title surrogate for exposure, cancer
incidence from Swedish Cancer Registry, Swedish
cancer rates referents (RR).
Multiple myeloma
1.75 (0.70, 3.61)
7
Leukemia
0.74 (0.30, 1.52)
7
Pukkala et al. (2009)
Launderer and dry cleaner
Lymphopoietic
0.98 (0.83, 1.11)
653
15 million men and women participating in national
census on or before 1990, 5 Nordic countries (Denmark,
Finland, Iceland, Norway, Sweden), 30-64 yr, follow-up
to 2005, occupational title of launderer and dry cleaner in
any census [n - 8,744 men, n - 34,752 women], PCE not
identified to individual subjects, cancer incidence from
national cancer registries, national population cancer
incidence rates referent (SIR).
Male
0.94 (0.79, 1.08)
140
Female
0.99 (0.83, 1.06)
513
Launderer and dry cleaner
Non-Hodgkin
0.98 (0.86, 1.10)
264
Male
0.96 (0.72, 1.25)
54
Female
0.98 (0.86, 1.13)
210
Launderer and dry cleaner
Hodgkin
0.97 (0.67, 1.36)
33
Male
0.77 (0.31, 1.58)
7
Female
1.04 (0.68, 1.53)
26
Launderer and dry cleaner
Multiple myeloma
1.02 (0.86, 1.20)
152
Male
1.31 (0.95, 1.78)
42
Female
0.94 (0.78, 1.33)
110
Launderer and dry cleaner
Leukemia11
0.95 (0.83, 1.09)
204
Male
0.71 (0.50, 0.99)
37
Female
1.03 (0.88, 1.19)
167
a -r
>1
to 0\
2 ^
3
¦t*
to
VO
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
Calvert et. al (In Press)
All subjects
Lympho- and hemato-
0.88 (0.53, 1.38)
19
1,704 U.S. men and women dry-cleaning union member
Exposure duration/time since 1st
employment
poietic
Not reported
in CA, IL, MI, NY follow-up 1940-2004 (618 subjects
worked for one or more years prior to 1960 only at shops
where PCE was the primary cleaning solvent, identified
as PCE-only exposure), cancer mortality (SMR).
PCE-only subjects
Lympho- and hemato-
poietic
1.51 (0.75,2.70)
11
All subjects
Non-Hodgkin
1.57 (0.78, 2.81)
11
Exposure duration/time since 1st
employment
lymphoma
Not reported
PCE-only subjects
Non-Hodgkin
lymphoma
2.46 (0.90, 5.36)
6
Selden and Ahlborg (2011)
Dry-cleaners and laundry workers
Non-Hodgkin
1.38 (1.02, 1.82)
49
9,440 Swedish men (n = 2,810) and women (n = 9,440)
PCE
in 461 washing and dry-cleaning establishments,
identified by employer in mid-1980s, employed
1973-1983, follow-up 1985-2000, exposure assigned
Males
Non-Hodgkin
2.02 (1.13,3.34)
15
Duration of exposure
using company self-reported information on PCE
usage—PCE (dry cleaners and laundries with a
proportion of PCE dry cleaning), laundry (no PCE use),
<1 yr
6.02 (2.21, 13.09)
6
1-4 yr
1.00 (0.12,3.61)
2
and other (mixed exposures to PCE, CFCs, TCE, etc.),
external referents (SIR).
5-11 yr
1.19 (0.64,3.27)
7
Females
1.14(0.68, 1.81)
18
Duration of exposure
<1 yr
1.95 (0.53, 5.00)
4
1-4 yr
1.04 (0.34, 2.44)
5
5-11 yr
1.10(0.46, 1.92)
9
a -r
>1
to 0\
2 ^
3
to
to
o
~n
H
o
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
Laundry
Selden and Ahlbore (2011) (continued)
Males
Non-Hodgkin
2.33 (1.01,4.59)
8
Females
0.99 (0.43, 1.95)
8
Travier et al. (2002)
All subjects
Non-Hodgkin1
0.86 (0.43, 1.72)
8
Men and women with occupation as dry cleaners,
launderers, and pressers in Swedish 1960 or 1970 Census
and employed in laundry, ironing, or dyeing industries,
Males
1.32 (0.75, 2.32)
5
Females
0.52 (0.17, 1.61)
3
followed 1971-1989, cancer incidence from Swedish
Cancer Registry, PCE not identified to individual
subjects, all other occupations/industries referent (RR).
All subjects
Hodgkin1
2.69(1.01,7.19)
4
Males
1.58(0.22, 11.26)
1
Females
3.57(1.15, 11.13)
3
All subjects
Leukemia
1.84 (1.11,3.06)
15
Males
0.93 (0.30, 2.88)
3
Females
2.53 (1.44, 4.46)
12
Case-Control Studies
Upper Cape Cod, MA (United States)
Aschenerau et al. (1993)
Any PCE, no lag
Leukemia
2.13 (0.88, 5.19)
7
34 men and women incident leukemia cases, 737
population controls, stratified by age, vital status, year of
death, sex, telephone or in-person interviews, water
distribution model of Webler and Brown (1993).
adjusted for sex, age, vital status, education, job
exposures (OR).
RDD >90lh percentile, no lag
Leukemia
8.33 (1.53,25.29)
2
Any PCE, >5 yr lag
Leukemia
1.96 (0.71,5.37)
Not
reported
RDD >90lh percentile, >5 yr lag
Leukemia
5.84 (1.37, 24.91)
Not
reported
a -r
>1
to 0\
2 ^
3
to
to
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
Iowa and Minnesota (United States)
Blair etal. ri993,l
Dry-cleaning industry
Non-Hodgkin
2.0 (0.97, 4.3)
16
622 histologically confirmed incident NHL cases in men,
1,245 population controls matched on state, age, and
year deaths [for dead cases], in-person interview, JEM
for solvent group but not PCE individually; adjusted for
age, state, smoking, family history lymphopoietic
disease, agricultural pesticide use, hair dye use, and
proxy respondent (OR).
Solvents other than benzene
Any exposure
Non-Hodgkin
1.1 (0.9, 1.4)
359
Low intensity
1.1 (0.8, 1.4)
334
High intensity
1.4 (0.8, 2.5)
25
France, 18 provinces
Claveletal. (1998)
Launderer and dry cleaner
Hairy cell leukemia
(a type of NHL)
3.0 (0.2, 49.2)
1
226 males histologically confirmed hospital HCL cases,
1980-1990,425 hospital controls from orthopedic and
rheumatological departments matched on sex, birth date,
admission date, residence, self-administered
questionnaire, JEM for solvent exposures, adjusted for
smoking and farming (OR).
Solvents, more confident exposure
assessment
Hairy cell leukemia
(a type of NHL)
0.7 (0.4, 1.2)
32
Italy, 12 regions
Costantini et al. (2001); Miliei et al. (2006); Costantini et
al. (2008)
PCE
2,737 incident lymphomas in men and women (1,450
NHL, 365 HD, 652 leukemia, 270 multiple myeloma)
20-74 yr, 1991-1993, 1,779 population controls
stratified by sex and age, in-person interview, exposure
proxy of job title and JEM for PCE, adjusted for sex,
age, education, and area (OR).
Very low/low intensity
Non-Hodgkin + CLL
0.6 (0.3, 1.2)
18
Medium/high intensity
1.2 (0.6,2.5)
14
Very low/low intensity
Leukemia
0.6 (0.2, 1.6)
6
Medium/high intensity
Leukemia
1.0 (0.4, 2.7)
7
Hodgkin
Not reported
a -r
>1
to 0\
2 ^
3
to
to
to
~n
H
o
O
-------
a
o
<
s?
cs
>!
o
^3
§•
rs
st
3
>!
sT
St
*3
o
>1
>!
Oq
o
>!
Si
Si
s
O
Sfc
I
§•
K) On
2
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
Launderer, dry cleaner, presser
Costantini et al. (2001); Miligi et al. (2006); Costantini et
Males
Non-Hodgkin + CLL
1.6 (0.3,9.1)
3
al. (2008) (continued)
Females
0.7(0.3, 1.5)
10
Males
Hodgkin
2.5 (0.3, 24.6)
1
Females
3.5 (1.5, 8.2)
7
Males
Multiple myeloma
Not reported
Females
1.0(0.3,3.8)
3
Males
Leukemia
3.3 (0.1, 32.4)
2
Females
1.1 (0.4, 3.2)
5
Languedoc-Roussillon region (France)
Fabbro-Perav et al. (2001)
Dry-cleaning solvents
Non-Hodgkin
1.0 (0.6, 1.6)
35
445 histologically confirmed Hodgkin and NHL hospital
cases in men and women recruited, 1992-1996, 1,025
population controls stratified on municipalities size and
population distribution, in-person or telephone interview,
self-reported exposure, exposed defined as duration >1
yr, 5 yr prior to diagnosis, information, adjusted for age,
sex, urban setting, education level (OR).
to
to
LtJ
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
Puget Sound-Seattle (Washington State), Detroit (Michigan) (United States)
Gold et al. (7010^
Ever exposed to PCE
Multiple myeloma
1.5 (0.8, 2.9y
16
180 histologically confirmed multiple myeloma cases in
Cumulative exposure (ppm-wk)
men and women reported to cancer registries,
2000-2002, 481 population controls, RDD or
Medicare/Medicaid services files, in-person interview,
self-reported or proxy-assisted reply to all jobs held >12
mo since 1945, adjusted for age, gender, race, education,
Referent
Multiple myeloma
1.0a
164
1-353
o.3 (0.04,3.oy
1
354-1,430
0.5(0.1,4.4)>
1
study site (OR).
1,431-4,875
1.5 (0.4, 5.4)>
4
4,876-13,500
3.3 (1.2, 9.5)1
10
p-valuc for trend
p = 0.02
Textile, apparel, furnishing machine
operators and tenders (includes dry-
cleaning machine operators)
Multiple myeloma
6.0(1.7,21)
9
Exposure duration
1-5 yr
Multiple myeloma
3.6 (0.7, 1.7)
4
>5 yr
12 (1.3, 110)
5
Trend test
p = 0.001
Dry-cleaning machine operators
Multiple myeloma
Not reported
5 cases, 3
controls
Umea (Sweden)
Hardell et al., 1989
Any styrene, TCE, PCE, benzene
exposure
Non-Hodgkin
4.6(1.9, 11.4)
10
169 men histologically confirmed incident NHL and
Hodgkin cases, 1974-1978, population controls, 25-85
yr, matched for sex, age, and residence, and death [for
dead cases], self-administered questionnaire, OR from
univariate x2 test.
a -r
>1
to 0\
2 ^
3
to
to
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
New York (United States)
Kato et al. ^2005^
Dry-cleaning fluids
Non-Hodgkin
1.59 (0.49, 5.13)
7
376 cases histologically confirmed NHL in women,
20-79 yr, 1995-1998, NY State Cancer registry, 463
population controls stratified on age, telephone
interview, occupation exposure to solvents, dry-cleaning
fluids, adjusted for age, family history hematologic
cancer, education, interview year, proxy respondent,
BMI, prescription/over-counter drugs, pesticide
exposures (OR).
Population of Denmark, Finland, Norway, Sweden
Lvnge et al. (2006)
Dry cleaner
Non-Hodgkin
1.0 (0.7, 1.4)k
42
46,768 subjects with occupation —dundry and dry-
cleaning worker" or industry -laundry and dry cleaning"
in 1970 Censuses in Denmark, Finland, Norway, Sweden
Other job in DC
Non-Hodgkin
0.7 (0.3, 1.6)k
8J
Unclassifiable
Non-Hodgkin
0.9 (0.6, 1.4)k
52J
followed 1970-1971 through 1997-2001; 247 incident
cases NHL, controls randomly selected from cohort,
matched on country, sex, age, and calendar period at
time of diagnosis. Dry cleaner assigned by job title or
employed in shop <10 employees using pension data in
Denmark and Finland or by questionnaire for subjects
from Sweden and Norway; mean PCE during study
period, 24 ppb (165 mg/m3), nested case-control study
(OR).
Dry cleaner, employment duration,
1964-1979
Non-Hodgkin
1.0 (referent)
145
<1 yr
1.35 (0.44, 4.14)
5
2-4 yr
0.61 (0.17, 2.21)
3
5-9 yr
0.92 (0.49, 1.72)
14
>10 yr
0.66 (0.36, 1.22)
15
Unknown
1.47 (0.49, 4.47)
5
a -r
>1
to 0\
2 ^
3
to
to
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
United States (SEER)
Malone et al. (1989)
Dry cleaner occupation
Chronic lymphocytic
leukemia (a type of
NHL)
1.1 (0.6,2.0)
(all respondents)
0.9 (0.4, 1.8)
(self-respondents,
noNOK
information)
14
427 men and women incident CLL cases and 1,683
population controls, <80 yr of age, SEER sites, matched
on sex, race, age, education, study site, questionnaire,
chlorinated HC surrogate exposure metric, adjusted for
race, age, education, sex, study site (OR).
New Zealand
Mannetje et al. (2008); McLean et al. (2009)
Textile bleaching, dyeing and
cleaning machine operators
Non-Hodgkin
0.75 (0.24, 2.32)
5
291 NHL cases ft Mannetie et al.. 2008) and 225
leukemia cases (McLean et al.. 2009). in men and
women, 20 or 25-75 yr, 2003-2004, New Zealand
Cancer Registry, 471 population controls frequency
matched on age, in-person interview, occupational title
as surrogate exposure metric, adjusted for age, sex, and
smoking (OR).
Leukemia
2.07 (0.70, 6.09)
6
Germany, 6 regions
Mester et al. (2006); Seidler et al. (2007)
Launderer, dry cleaner, presser
710 histologically confirmed Hodgkin and NHL in men
and women, 18-80 yr, 1998-2003, 710 population
controls matched on sex, region, and age, in-person
interviews, exposure assessed by job title and JEM for
semiquantitative intensity metric, adjusted for smoking
and alcohol consumption (OR).
Any exposure
Non-Hodgkin and
Hodgkin
1.3 (0.5, 3.2)
11
1-10 yr duration
0.8 (0.3, 2.5)
6
>10 yr duration
3.4 (0.6, 18.5)
5
PCE
0 ppm-yr
Non-Hodgkin and
Hodgkin
1.0 (reference)
667
>0- <9.1 ppm-yr
1.1 (0.5,2.3)
16
>9.1- <78.8 ppm-yr
1.0 (0.5, 2.2)
14
>78.8 ppm-yr
3.4 (0.7, 17.3)
6
Test for trend
p = 0.12
a -r
>1
to 0\
2 ^
3
to
to
On
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
0 ppm-yr
Multiple myeloma
1.0 (reference)
33
Mester et al. (2006); Seidler et al. (2007) (continued)
>0- <9.1 ppm-yr
1.8 (0.5, 6.7)
3
>9.1- <78.8 ppm-yr
0
>78.8 ppm-yr
0
Test for trend
p = 0.34 (negative)
4-SEER reporting sites (CA, 10, MI, WA, United States)
Schenk et al. (2009)
Launderers and ironers
Non-Hodgkin
3.89 (1.06, 14.20)
12
2,046 histologically confirmed NHL in men and women,
20-74 yr, 1998-2000, 1,057 population controls
frequency matched on age, sex, race and study center,
mailed questionnaire, occupational title exposure
surrogate, adjusted for age, group, sex, ethnicity, and
study center (OR).
Montreal, Canada
Siemiatvcki (1991)
Launderer and dry cleaner
215 men and women histologically confirmed incident
NHL cases , 1979-1985, 35-70 yr, 533 population
control group and cancer control group, in-person
interviews, occupational title and JEM for PCE, adjusted
age, family income, and cigarette index, 90% CI (OR).
Any exposure
Non-Hodgkin
0.9 (0.3, 2.4)
3
Substantial exposure
(0.00, 1.7)
0
Geographic-based and Other Studies
Northern New Jersey, 75 Municipalities (United States)
Cohnetal. (1994b)
PCE in town water >5 ppb
1,190 leukemia cases identified from NJ State Cancer
Registry, 1979-1987, residence in 1 of 17 NJ
municipalities, PCE and other chlorinated solvents in
Males
Non-Hodgkin1
1.20 (0.94, 1.52)
78
Females
1.38 (1.08, 1.70)
87
municipal water supplies, log-linear regression adjusted
for age, stratified by sex (RR).
Males
Leukemia
0.84 (0.66, 1.06)
63
Females
1.20 (0.94, 1.52)
56
a -r
>1
to 0\
2 ^
3
to
to
^1
~n
H
O
O
-------
Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No. obs.
Reference(s) and study description
Portland-Vancouver Metropolitan Area,, Oregon (United States)
Morton and Marianovic (1984)
Dry cleaners and launderers
1,622 leukemia cases identified from 24 hospitals and
death certificates, 1963-1977, 16-74 yr, occupational
title as exposure surrogate, 1,611 dry cleaners and
launderers in 1970 population census, age-standardized
rates using 1970 population.
Males
All leukemia
55.7 per 100,000m
2
Females
23.7 per 100,000m
5
Males
Lymphatic leukemia
27.8 per 100,000m
1
Females
20.9 per 100,000m
4
Males
Nonlymphatic leukemia
27.8 per 100,000m
1
Females
9.0 per 100,000m
2
Hausjarvi and Hattula, Finland
Vartiainen et al. (1993)
Hausjarvi
Non-Hodgkin
0.6(0.3, 1.1)
14
Lymphopoeitic cancers, liver cancer and all cancers
among residents with PCE and other solvents in drinking
water, 1953-1991, no subject-level exposure
information, cancer rates of Finnish population referent
(SIR).
Hattula
1.4(1.0,2.0)
31
Hausjarvi
Hodgkin
0.8 (0.3, 1.7)
6
Hattula
1.4 (0.7, 2.5)
11
Hausjarvi
Multiple myeloma
0.7(0.3, 1.3)
7
Hattula
0.7 (0.2, 1.3)
6
Hausjarvi
Leukemia
1.2 (0.8, 1.7)
33
Hattula
0.7(0.4, 1.1)
19
a For Boice et al.(1999), all lymphopoetic cancers is the sum of ICD 9th Edition, 200-208.
b Internal referent population as comparison.
0 For Andersen et al. (19991. all lymphopoeitic cancer is the sum of ICD 7th Edition, 200-204.
d For Cano and Pollan (20011. relative risk for male dry cleaner and launderers in 1970 Census.
e For Cano and Pollan (20011. relative risk for male dry cleaner and launderers in 1960 and 1970 Censuses.
-------
a to ^ Table 4-27. Summary of epidemiologic studies on tetrachloroethylene exposure and hematopoietic cancers,
nil including leukemia (continued)
! §•
s For Ji and Hemminki (20061. female subjects reporting occupation as launderers and dry cleaner in two consecutive censuses, 1960-1970, SIRs for NHL were
1 | | 0.76 (95% CI: 0.39, 1.25) [n = 12] and 0.87 (95% CI: 0.76, 1.10) [n = 64], respectively, and, for multiple myeloma, 1.01 (0.46, 1.78) [n = 9] and 0.88 (0.60,
^ 1.21) [n = 31], respectively.
2 - g For Ji and Hemminki (2005b. 2006). SIR for launderers and dry cleaners in 1960 Census. For lymphopoietic subtypes in launderers and dry cleaners in 1960
^ a Census, for males, SIR: 0.85 (0.51, 1.28) [n = 19] for chronic lymphocytic leukemia, a form of NHL; 0.63 (0.25, 1.18) \n = 7] for acute myelogenous
o_ a, ^ leukemia; 0.91 (0.29, 1.87) [n = 5] for chronic myelogenous leukemia; and, 1.04 (0.41, 1.96) [n = 7] for polycythemia vera; and, for females, SIR: 1.54 (1.05,
^ 2.12) |II = 32] for chronic lymphocytic leukemia; 0.1.36 (0.83, 2.02) [n = 20] for acute myelogenous leukemia; 0.33 (0.03, 0.94) [n = 2] for chronic
^ myelogenous leukemia; and, 1.71 (0.93, 2.73) [n = 14] for polycythemia vera.
o § hFor Pukkala et al. (2009). SIR for chronic lymphatic leukemia, a form of NHL, were 0.90 (95% CI: 0.50-1.49) [males, n = 15 cases] and 1.02 (95% CI: 0.74,
Ull ^ 1.36) [females, n = 46 cases].
:For Travier et al. (2002). RRs for subjects reporting occupation as dry cleaners, launderers, or pressers and employed in dry-cleaning industry in 1960 and 1970
Censuses (Group 2). RRs for these subjects for chronic lymphocytic leukemia, a form of NHL, were 0.67 (0.09, 4.76) [males, n = 1] and 2.89 (1.20, 6.96)
[females, n = 5],
J For Gold et al. (2010b). odds ratio for PCE exposure with jobs assessed with low confidence considered unexposed.
kLynge et al.(2006) is a nested case-control study. RR adjusted for matching criteria (country, sex, 5-yr age group and 5-yr calendar period at the time of
diagnosis of the case).
'For Cohn et al. (1994b). RRs for chronic lymphocytic leukemia, a form of NHL, were 0.98 (0.65, 1.47) [males, n = 28] and 0.93 (0.56, 1.52) [females, n = 19],
m For Morton and Maijanovic (1984). age-standardized incidence rate is statistically significantly different from rate for all men or all women.
CLL = chronic lymphocytic leukemia; Exp = expected number of cancers; JEM = job-exposure matrix; NOK = next of kin; RDD = relative delivered dose.
to
to
vo
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-28. Summary of epidemiologic studies on tetrachloroethylene exposure and childhood hematopoietic
cancers, including leukemia
Exposure group
Cancer site
Relative risk
(95% CI)
No.
obs.
Reference(s) and study description
Cohort Studies
Offspring of Electronic factory workers (Taiwan)
Sung et al. (2009)
Nonexposed
All leukemia (ICD 9, 204-208)
1.0
9
40,647 first singleton births among 47,356 women
employed at factory, 1978-2001, 8,506 births among
women employed 3 mo prepregnancy and 3 mo
postconception, incident childhood cancers from
national cancer registry, 1979-2001, does not
identify PCE exposure to individual mothers,
Poisson regression adjusted for maternal age,
maternal education, sex and birth year, internal
referents [offspring of subjects not employed during
period] (RR).
Exposed pregnancy to organic solvents
3.83 (1.17, 12.55)
6
Case-Control Studies
Residents of ages <19 in Woburn, MA (United States)
Costas et al. (2002)
Maternal exposure 2 yr before conception to diagnosis
19 leukemia, 1969-1989, identified through
physician or hospital records pre-1982 and MA
Cancer Register 1982 onward, 37 local public school
controls matched on race, sex, birth date, residential
status, in-person interview, questionnaire to parents
included information on use of public drinking water
in the home, hydraulic mixing model used to
estimate fraction of month that TCE , PCE and other
solvents in drinking water were delivered to
residence 1964-1979 (Murphy,1991), logistic
regression with composite covariate for
socioeconomic status, maternal smoking during
pregnancy, maternal age at birth of child, and
breastfeeding (OR).
Never
Acute lymphocytic leukemia
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 yr before conception
Never
Acute lymphocytic leukemia
1.00
11
Least
2.48 (0.42, 15.2)
4
Most
2.82 (0.30, 26.4)
5
(p for linear trend)
>0.05
a -r
>1
to 0\
2 ^
3
to
LtJ
o
~n
H
O
O
-------
Table 4-28. Summary of epidemiologic studies on tetrachloroethylene exposure and childhood hematopoietic
cancers, including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No.
obs.
Reference(s) and study description
Birth to pregnancy
Costas et al. (2002) (continued)
Never
Acute lymphocytic leukemia
1.00
7
Least
1.82 (0.31, 10.8)
7
Most
0.90 (0.18,4.56)
5
(p for linear trend)
>0.05
Maternal exposure during pregnancy
Never
Acute lymphocytic leukemia
1.00
9
Least
3.53 (0.22, 58.1)
3
Most
14.3 (0.92, 224)
7
(p for linear trend)
<0.05
Residents of ages <14 yr Quebec (Canada)
Infante-Rivard et al. (2005)
Probable/definite exposure to PCE
Acute lymphocytic leukemia
ICD 9 204.0
0.87 (0.35-2.18)
18
790 acute lymphoblastic leukemia, 1980-2000, 790
population controls from family stipend records,
1980-1993, or health insurance records, 1994-2000,
matched on sex and age, telephone interview with
Maternal exposure 2 yr before
conception to birth
0.96 (0.41-2.25)
11
During pregnancy
0.84 (0.30-2.34)
7
questions on maternal occupation, blinded JEM for
PCE, logistic regression stratified by time period and
adjusted for maternal age and education (OR).
Cumulative exposure score
<4
Acute lymphocytic leukemia
ICD 9 204.0
0.95 (0.35-2.55)
>4
0.55 (0.05-6.34)
-------
Table 4-28. Summary of epidemiologic studies on tetrachloroethylene exposure and childhood hematopoietic
cancers, including leukemia (continued)
Exposure group
Cancer site
Relative risk
(95% CI)
No.
obs.
Reference(s) and study description
Residents of ages <10 yrLos Angeles (CA) Cancer Surveillance Program
Lowengart et al. (1987)
Maternal occupational exposure to PCE
Acute lymphatic and
nonlymphatic leukemia
Not reported
123 case-control pairs—acute lymphocytic and
nonlymphocytic leukemia cases, 1980-1984, and
maternal friend controls or population controls
matched on age, sex, race, nonblinded telephone
Paternal occupational exposure to PCE
1 yr before pregnancy
Acute lymphatic and
nonlymphatic leukemia
co (p = 0.39)
l:0a
interview, self-reported occupational exposure,
logistic regression (OR).
During pregnancy
co (p = 0.39)
l:0a
After pregnancy
co (0.19-co)
2:0a
Children's Cancer Group Study (children <15 yr of age) (Australia, Canada, United States)
Shu et al. (19991
Maternal occupational exposure to PCE
1,842 acute lymphocytic leukemia cases identified in
37 participating institutions, 1989-1933, 1,986
population controls, RDD, matched on age, race and
telephone area code/exchange, telephone interview
with structured questionnaire to assess parental
exposure to PCE using job-industry title and self-
reported exposure history, logistic regression
adjusted for maternal education, race and family
income (maternal exposures) or paternal education,
race, family income, age and sex of case (OR).
Anytime
Acute lymphocytic leukemia
0.4(0.1-1.4)
4
Preconception
1.4 (0.2-8.6)
3
During pregnancy
1.3 (0.2-8.4)
3
Postnatal
0.4(0.1-1.5)
4
Paternal occupational exposure to PCE
Anytime
Acute lymphocytic leukemia
0.9 (0.5-1.6)
25
Preconception
0.8(0.5-1.5)
21
During pregnancy
0.5 (0.2-1.1)
8
Postnatal
0.5 (0.2-1.2)
10
" For Lowengart et al. (19871. the number of case:control pairs.
Exp = expected number of cancers; JEM = job-exposure-matrix; RDD = relative delivered dose; Obs = observed number of cancers.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
The exposure surrogate in studies of dry cleaners and launderers is a broad category and
will have some associated measurement error as this broad category does not account for
individual characteristics that modify one's exposure potential. For example, some variation can
be expected within an occupational group between countries, as Lynge et al. (2006) reported,
average tetrachloroethylene usage in 1960-1970 in Sweden was higher than in Finland or
Norway. The more general the exposure surrogate, such as job title, the greater the likelihood
for misclassification errors, as differences in tasks and exposure conditions within a job title may
be considerable. For some occupations, these differences may be gender related, making it
difficult to interpret differences in relative risk that may be observed between men and women
within a specific occupational group (Messing et al.. 1994). Blair et al. (2003) recruited
members of a laundry and dry-cleaning workers union and attempted to increase the specificity
of the classification of tetrachloroethylene exposure by examining a subgroup who entered the
cohort after 1960, a time of widespread tetrachloroethylene use in dry cleaning. However, this
restriction resulted in a considerable decrease in the number of observed cases of lymphopoietic
cancers, from 39 in the full cohort to 2 in the group that joined after 1960. Blair et al. (2003)
also developed a semiquantitative exposure intensity score using published monitoring data. The
available data indicated a high degree of consistency in exposure levels to tetrachloroethylene
between establishments and provided information that could be used to categorize differences in
potential exposures based on types of jobs. Exposure was characterized with respect to distance
from the washers (cleaners were assigned a high-exposure score, pressers, sewers, and counter
clerks were assigned a medium-exposure score, and those who worked at locations that did not
include washing facilities were assigned a no-exposure score) (Blair et al.. 2003; Blair et al..
1990). Another study by (Calvert et al.. In Press) of unionized dry cleaners in the United States
included an analysis of subjects who worked for one or more years before 1960 in a shop known
to use tetrachloroethylene as the primary solvent. The cohort was stratified into two groups
based on the level of certainty that the worker was employed only in facilities using
tetrachloroethylene as the primary solvent: tetrachloroethylene-only and tetrachloroethylene
plus. Another approach to improving the exposure measure was used by Lynge et al. (2006). In
this study, effect measures were presented for dry cleaners separately from other laundry
workers. Selden and Ahlborg (2011 > obtained information about the dry-cleaning establishment
(e.g., washing techniques, chemicals used, number of employees, and work history of individual
employees) in a questionnaire sent to businesses in Sweden in the 1980s to identify subjects as
either dry cleaners or laundry workers. Travier et al. (2002) presented estimates for launderers,
dry cleaners, and pressers, using job classifications based on the 1960 or 1970 Census data, and
for subjects holding a dry-cleaning job in both census years.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-233 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
A variety of exposure-assessment approaches have been used in studies in other work
settings and in population-based case-control studies. One occupational study assessed
tetrachloroethylene potential for individual subjects using biological monitoring data (Anttila et
al.. 1994). The cohort studies of aerospace workers (Boice et al.. 1999) and aircraft maintenance
workers (Blair et al.. 1998; Radican et al.. 2008; Spirtas et al.. 1991) developed a job exposure
matrix referencing historical industrial monitoring data. In case-control studies, attributes that
strengthen the quality of the exposure assessment include ascertainment of a complete job
history (i.e., all jobs held for >6 or 12 months rather than limiting to most recent job or longest-
held job), inclusion of information on job tasks or duties as well as job title, inclusion of
additional modules for specific jobs that collect more detailed information pertaining to exposure
conditions, and blinded exposure assessment and development of job-exposure matrices focusing
on tetrachloroethylene based on this complete set of information. These attributes were used in
the case-control studies in Italy (Costantini et al.. 2008; Miligi et al.. 2006) and a case-control
study of multiple myeloma in Washington (Gold et al.. 2010a). One case-control study of
potential residential tetrachloroethylene exposure used a statistical model of water distribution
system to estimate delivered dose to a subject's home (Aschengrau et al.. 1993). Because a
nondifferential misclassification of exposure most often leads to an attenuation of the observed
effect estimates(Dosemeci etal.. 1990). the relative specificity of these exposure-assessment
approaches, particularly those that allow assignment of values to individuals within the study,
strengthens their ability to identify cancer hazards compared to studies with broader exposure-
assessment approaches.
4.6.1.2.2. Adult lymphopoietic cancer: consideration of disease subtypes
The broad category of lymphopoietic cancers can be divided into specific types of
cancers, including non-Hodgkin lymphoma, Hodgkin lymphoma, multiple myeloma, and various
types of leukemia (e.g., acute and chronic forms of lymphoblastic and myeloid leukemia). The
classification criteria for these cancers have changed over the past 30 years, reflecting improved
understanding of the underlying stem cell origins of these specific subtypes. For example, hairy
cell leukemia, chronic lymphocytic leukemia, non-Hodgkin lymphoma, and multiple myeloma
may arise from mature B cells. This understanding may help elucidate common etiologic
pathways and exposures. The studies of tetrachloroethylene exposure examine various
outcomes, including the broad category of lymphopoietic cancers, as well as non-Hodgkin
lymphoma, Hodgkin lymphoma, non-Hodgkin lymphoma plus chronic lymphocytic leukemia,
hairy cell leukemia, multiple myeloma, and leukemia.
All of the studies of dry cleaning and other occupations from the Nordic countries
ascertained cancer incidence using national cancer registries. Four other cohort studies from the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-234 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
United States (Blair et al.. 2003; Boice et al.. 1999; Calvert et al.. In Press; Radican et al.. 2008)
relied on cause-of-death data from death certificates or the National Death Index. For diseases
with a relatively high survival rate such as non-Hodgkin lymphoma (5-year survival: 67.4%
based on 1999-2006 data), use of cause-of-death data may underestimate cancer risk. Most of
the case-control studies relied on histologically confirmed cases of incident cancers in a defined
geographic area, as ascertained from cancer registries.
4.6.1.2.3. Adult lymphopoietic cancer: consideration of potential confounding and other
factors
Common behaviors, such as smoking and use of alcohol, have not been strongly
associated with non-Hodgkin lymphoma and multiple myeloma, so there is little reason to be
concerned about potential confounding of observed results pertaining to specific jobs or
tetrachloroethylene measures by these factors. Smoking is a risk factor for some kinds of
leukemia, however, and so its role as a potential confounder for this outcome should be
considered. Tetrachloroethylene was the primary, or in Nordic countries, the exclusive solvent
used in dry cleaning (Johansen et al.. 2005; Lynge et al.. 2006). In studies of some types of
occupations, participants may also have been exposed to other solvents.
4.6.1.2.4. Adult lymphopoietic cancer: summary of results
All of the studies examining the broad category of lymphopoietic cancers were cohort
studies, with the number of exposed cases ranging from 3, in a study of biologically monitored
workers in Finland (Anttila et al.. 1995). to 653, in a study using occupational census codes in
five Nordic countries (Pukkala et al.. 2009) (see Table 4-29). The relative risk estimates among
these seven studies ranged from 0.67 (95% CI: 0.42, 1.01) to 1.51 (95% CI: 0.75, 2.70), with
values from the largest studies around 1.0 (Andersen et al.. 1999; Pukkala et al.. 2009). The
three studies with relative risk estimates greater than 1.0 were studies that used a relatively high
quality exposure-assessment methodology: an standardized incidence ratio (SIR) of 1.39 (95%
CI: 0.28, 4.02) in a small study in Finland examining risk among workers who had been
monitored using blood tetrachloroethylene levels (Anttila et al.. 1995). an SMR of 1.51 (95%
CI: 0.75, 2.70) among laundry and dry-cleaning union workers employed prior to 1960 only in
facilities using tetrachloroethylene as the primary solvent (tetrachloroethylene-only) (Calvert et
al.. In Press), and an SMR of 1.13 (95% CI: 0.62, 1.89) for routine exposure to
tetrachloroethylene, based on a job exposure matrix, in a cohort study of workers in the
aerospace industry (Boice et al.. 1999). In the other study with a relatively detailed exposure-
assessment methodology (a semiquantitative exposure score based on job titles and proximity to
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-23 5 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-29. Results of epidemiologic studies of potential tetrachloroethylene
exposure and adult lymphopoietic cancer and leukemia, by cancer type and
study design
Cancer type,
n exposed
cases
Relative risk
(95% CI)
Design, location, exposure assessment"
Reference
Lymphopoietic (all)
Cohort
3
1.38 (0.28, 4.02)
biological monitored workers (SIR), Finland, blood
PCE3
Antilla et al. (1995)
11
1.51 (0.75,2.70)
laundry and dry-cleaning workers (SMR), United
States, union employment records (PCE-only exposure
based on history of solvent use by shops)
Calvert et al. ; Ruder
et al. (2001)
14
1.13 (0.62, 1.89)
aerospace workers (SMR), United States, job exposure
matrix (PCE routine exposure)3
Boice et al.. (1999)
22
0.67 (0.42, 1.01)
electronic factory workers (SIR), Taiwan
Chang et al.. (2005)
39
1.0 (0.7, 1.3)
laundry and dry-cleaning workers (SMR), United
States, union records (all workers)
Blair et al. (2003)
0.9 (0.5, 1.4)
(medium/high intensity score)3
Blair et al. ^2003^
204
1.0 (0.87, 1.15)
laundry and dry-cleaning workers (SIR), Sweden,
Denmark, Finland, Norway, census occupation codes
Andersen et al. (1999)
653
0.98 (0.30, 1.52)
laundry and dry-cleaning workers (SIR), Sweden,
Denmark, Finland, Norway, Iceland, census occupation
codes
Pukkala et al. (2009)
Leukemia (all)
Cohort
5
0.55 (0.18, 1.29)
aerospace workers (SMR), United States, job exposure
matrix (PCE routine exposure)3
Boice et al.. (1999)
5
0.78 (0.49, 1.17)
electronic factory workers (SIR), Taiwan (females)
Suns et al. (2007)
7
0.74 (0.30, 1.52)
laundry and dry-cleaning workers (SIR), Sweden,
census occupation codes
Lynge and Thygesen
(1990)
12
0.8 (0.4, 1.4)
laundry and R workers (SMR), United States, union
records (all workers)
Blair et al. (2003)
3
0.93 (0.30, 2.88)
laundry and dry-cleaning workers and pressers, Sweden,
census occupation codes, 1960 and 1970 (males)
Travier et al. (2002)
12
2.53 (1.44, 4.46)
laundry and dry-cleaning workers and pressers, Sweden,
census occupation codes, 1960 and 1970 (females)
Travier et al. (2002)
15
1.84 (1.11,2.88)
laundry and dry-cleaning workers and pressers, Sweden,
census occupation codes, 1960 and 1970 (males and
females)
Travier et al. (2002)
58
0.85 (0.65, 1.0)
laundry and dry-cleaning workers (SIR), Sweden,
Denmark, Finland, Norway, census occupation codes
Andersen et al. (1999)
47
0.84 (0.62, 1.90)
laundry and dry-cleaning workers (SIR), Sweden
(males)
Ji and Hemminki
(2005b)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-236 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-29. Results of epidemiologic studies of potential tetrachloroethylene
exposure and adult lymphopoietic cancer and leukemia, by cancer type and
study design (continued)
Cancer type,
n exposed
cases
Relative risk
(95% CI)
Design, location, exposure assessment"
Reference
80
1.30 (1.03, 1.60)
laundry and dry-cleaning workers (SIR), Sweden
(females)
Ji and Hemminki,
(2005b)
204
0.95 (0.83, 1.09)
laundry and dry-cleaning workers (SIR), Sweden,
Denmark, Finland, Norway, Iceland, census occupation
codes
Pukkala et al. (2009)
Leukemia (all)
Case-control
2
3.3 (0.3, 32.4)
Italy, job titles (launderer, dry cleaner, presser) (males)
Costantini et al. (2001)
5
1.1 (0.4,3.2)
Italy, job titles (launderer, dry cleaner, presser)
(females)
Milisi et al. (1999)
6
2.07 (0.70, 6.09)
New Zealand, occupational title (textile bleaching,
dyeing and cleaning machine operators)
McLean etal. (2009)
7
1.0 (0.4, 2.7)
Italy, job exposure matrix (PCE, medium/high
intensity)3
Costantini et al. (2008)
Leukemia (all)
Geographic based
7
2.13 (0.88, 5.19)
Massachusetts, water distribution model (any PCE)a
Aschengrau et al.
(1993)
19
0.7 (0.4, 1.1)
Finland (Hattula), PCE in drinking water
Vartiainen
etal. ri993^
33
1.2 (0.8, 1.7)
Finland (Hausjarvi), PCE in drinking water
Vartiainen et al.
(1993)
56
1.20 (0.94, 1.52)
New Jersey, PCE in town water >5 ppb (females)
Cohnetal. (1994b)
64
0.84 (0.66, 1.06)
New Jersey, PCE in town water >5 ppb (males)
Cohnetal. (1994b)
" Studies with relatively high quality exposure assessment methodologies, based on biological monitoring data,
cohort studies with job exposure matrix based on historical industrial monitoring data, or case-control studies with
job exposure matrix focusing on PCE based on information on job title and tasks or duties, and additional modules
for specific jobs, or studies of residential PCE exposure using a statistical model of water distribution system to
estimate delivered dose to a subject's home.
1
2
3 washers), no increased risk was seen (SMR: 0.9, 95% CI: 0.5, 1.4, for the medium/high intensity
4 score group) (Blair et al.. 2003).
5 Studies of leukemia risk include occupational cohorts and case-control studies and
6 geographic-based studies of residential exposure (see Table 4-30). The cohort studies range
7 from 5 to 204 cases. Two studies using Swedish census data on occupation reported elevated
8 relative risks among women, but not men, who reported jobs as launderers or dry cleaners.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-237 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-30. Results of epidemiologic studies of potential tetrachloroethylene
exposure and adult non-Hodgkin lymphoma, by study design
Cancer type, n
exposed cases
Relative Risk
(95% CI)
Design, location, exposure assessment"
Reference
Adult non-Hodgkin lymphoma
Cohort
3
3.76 (0.77, 11.0)
biological monitored workers (SIR), Finland, blood PCEa
Antillaetal. (1995)
2
2.35 (0.52, 10.7)
Aircraft maintenance workers (RR-internal referent), United
States, job exposure matrix (PCE) (females)3
Radican et al. (2008)
5
2.32 (0.75,7.15)
Aircraft maintenance workers (RR-internal referent), United
States, job exposure matrix (PCE) (males)3
Radican et al. (2008)
6
2.46 (0.90, 5.36)
laundry and dry-cleaning workers (SMR), United States, union
employment records (PCE-only exposure based on history of
solvent use by shops)
Calvert et al.
Ruder et al. (2001)
8
1.70 (0.73, 3.34)
aerospace workers (SMR), United States, job exposure matrix
(routine exposure to PCE)a
Boice et al.. (1999)
8
1.03 (0.44, 2.02)
laundry and dry-cleaning workers (SIR), Sweden, census
occupation codes
Lynge and Thygesen
(1990s)
8
0.86 (0.43, 1.72)
laundry and dry-cleaning workers and pressers, Sweden, census
occupation codes
Travier et al. (2002)
11
1.76 (0.97, 3.17)
laundry and dry-cleaning workers (SIR), Sweden, census
occupation codes
Cano and Pollan,
(2001s)
12
0.9 (0.5, 1.6)
laundry and dry-cleaning workers (SMR), United States, union
records (all workers)
Blair etal. (2003)
15
2.02 (1.13, 3.34)
dry-cleaning workers (SIR), Sweden, census occupation codes
and questionnaire (dry cleaner) (males)3
Selden and Ahlborg,
(2011)
18
1.14 (0.68, 1.81)
dry-cleaning workers (SIR), Sweden, census occupation codes
and questionnaire (dry cleaner) (females)3
Selden and Ahlborg,
(2011)
27
1.46 (0.96,2.13)
laundry and dry-cleaning workers (SIR), Sweden, Denmark,
Finland, Norway, census occupation codes (males)
Andersen et al.
(1999)
55
0.95 (0.71, 1.23)
laundry and dry-cleaning workers (SIR), Sweden, Denmark,
Finland, Norway, census occupation codes (females)
Andersen et al.
(1999)
59
0.99 (0.75, 1.26)
laundry and dry-cleaning workers (SIR), Sweden (males)
Ji and Hemminki
(2006)
67
1.05 (0.82, 1.32)
laundry and dry-cleaning workers (SIR), Sweden (females)
Ji and Hemminki,
(2006)
264
0.98 (0.86, 1.10)
laundry and dry-cleaning workers (SIR), Sweden, Denmark,
Finland, Norway, Iceland, census occupation codes
Pukkala et al. (2009)
42
1.0 (0.7, 1.4)
Nested case-control, Sweden, Denmark, Finland, Norway, census
occupation codes and pension data/questionnaires (dry cleaners)
Lvnse et al. (2006)
Adult non-Hodgkin lymphoma
Case-control
1
3.0 (0.2, 49.2)
France, jobs held 6 or more mo, launderer and dry cleanerb
Clavel et al. (1998)
3
0.9 (0.3,2.4)
Canada, job exposure matrix for PCE (any exposure)
Siemiatvcki (1991)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-23 8 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-30. Results of epidemiologic studies of potential tetrachloroethylene
exposure and adult non-Hodgkin lymphoma, by study design (continued)
Cancer type,
n exposed
cases
Relative Risk
(95% CI)
Design, location, exposure assessment"
Reference
3
1.6(0.3,9.1)
Italy, job titles (launderer, dry cleaner, presser) (males)
Costantini et al,
(2001)
5
0.75 (0.24, 2.32)
New Zealand, occupational title (textile bleaching, dyeing
and cleaning machine operators)
_t Mametje et al.
C2008)
7
1.59 (0.49, 5.13)
United States, self-reported exposure to dry-cleaning fluids
Kato et al. (2005^
9
1.6 (0.6, 4.0)
United States, laundering, dry cleaning, leather products
fabrication0
Scherretal. (1992)
10
0.7(0.3, 1.5)
Italy, job titles (launderer, dry cleaner, presser) (females)
Milisi et al. CI999)
10
4.6(1.9, 11.4)
Sweden, JEM using self-reported information (any styrene,
TCE, PCE, or benzene exposure)
Hardell et al., 1989
12
3.89 (1.06, 14.2)
United States, occupation title (launders and ironers)
Schenk et al. (2009)
14
1.2 (0.6, 2.5)
Italy, job exposure matrix (PCE, medium/high intensity)3'd
Milisi et al. (2006)
14
1.1 (0.6,2.0)
United States, ever employed in dry-cleaning industry"
Malone et al. (1989)
16
2.0 (0.97, 4.3)
United States, all jobs held >1 yr (dry-cleaning industry)
Blair etal. (1993^
35
1.0 (0.6, 1.6)
France, self-reported exposure to dry-cleaning solvents
Fabbro-Peray et al.
(2001)
Adult non-Hodgkin lymphoma
Geographic-based (residential exposure)
Vartiainen et al.
(1993)
14
0.6(0.3, 1.1)
Finland (Hausjarvi), PCE and other solvents in drinking
water
Vartiainen et al.
(1993)
31
1.4(1.0,2.0)
Finland (Hattula), PCE and other solvents in drinking
water
Vartiainen et al.
(1993)
78
1.20 (0.94, 1.52)
New Jersey, PCE in town water >5 ppb (males)
Cohn et al. (1994b)
87
1.38 (1.08, 1.70)
New Jersey, PCE in town water >5 ppb (females)
Cohn et al. (1994b)
a Studies with relatively high quality exposure-assessment methodologies, based on biological monitoring
data, cohort studies with job exposure matrix based on historical industrial monitoring data, or case-control
studies with job exposure matrix focusing on PCE based on information on job title and tasks or duties, and
additional modules for specific jobs, or studies of residential PCE exposure using a statistical model of
water distribution system to estimate delivered dose to a subject's home.
b Includes patients with hairy cell leukemia.
0 Number of exposed cases estimated based on report of a prevalence of 3% in the population (n cases =
303); job history limited to most recent job, job held 15 yr ago, major occupation, and second most major
occupation.
d Includes patients with non-Hodgkin lymphoma and chronic lymphocytic leukemia.
e Includes patients with chronic lymphocytic leukemia.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-239 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Travier et al. (2002) examined cancer incidence from 1971 through 1989. The relative
risk among women who reported work as a launderer, dry cleaner, or presser in the laundry,
ironing, or dyeing industry in 1960 and 1970 was 2.53 (95% CI: 1.44, 4.46), and among men, the
relative risk was 0.93 (95% CI: 0.30, 2.28). Ji and Hemminki (2005b) used a similar approach,
with cancer incidence ascertained through 2002. The start of follow-up began at the time of the
relevant census data (i.e., 1961 for analyses based on jobs held in 1960). The SIR among women
who worked as a launderer or dry cleaner in 1970 was 1.30 (95% CI: 1.03, 1.60), and the SIR
among men who worked as a launderer or dry cleaner in 1960 was 0.84 (95% CI: 0.62, 1.09).
The latter time period was used for women because of the increase of women in the workforce
during the 1960s. A limitation of these studies is the lack of detailed information pertaining to
job tasks for individuals, information that could be particularly useful with respect to the
interpretation of the observed gender-related differences. No increased risk was seen in the
cohort study of aerospace workers using a job exposure matrix to estimate tetrachloroethylene
exposure (SMR: 0.55, 95% CI: 0.18, 1.29 in Boice et al. (1999)). The number of exposed cases
in the case-control studies range from 2 to 7 leukemia cases. The odds ratio in the study with a
relatively strong exposure-assessment methodology was 1.0 (95% CI: 0.4, 2.7) (Costantini et al..
2008). The three geographic-based studies of residential exposure involved 7 to 64 exposed
cases. The case-control study in Cape Cod, MA, that estimated exposure using a statistical
model of the water distribution reported an adjusted odds ratio of 2.13 (95% CI: 0.88, 5.19) for
any tetrachloroethylene exposure and 8.33 (95% CI: 1.53, 25.29) for exposures above the 90th
percentile (Aschengrau et al.. 1993). Relative risk estimates were lower, ranging from 0.7 to 1.2,
in two other residential studies with poorer quality exposure-assessment approaches (Cohn et al..
1994b: Vartiainen et al.. 1993).
The data pertaining to non-Hodgkin lymphoma are more extensive, with 14 cohort
studies ranging in size from 3 (Anttila et al.. 1995) to 264 (Pukkala et al.. 2009) cases,
13 publications based on case-control studies from six countries ranging in size from 3
(Siemiatvcki. 1991) to 35 exposed cases (Fabbro-Perav et al.. 2001). and two geographic-based
studies of residential exposures through drinking water (Cohn et al.. 1994b: Vartiainen et al..
1993) (see Table 4-30). Six of the relative risk estimates from the cohort studies, including the
four with the largest number of non-Hodgkin lymphoma cases, were between 0.95 and 1.05
(Andersen et al.. 1999: Ji and Hemminki. 2005b. 2006: Pukkala et al.. 2009). Among the nine
smaller cohorts (n cases <30) (Andersen et al.. 1999: Anttila et al.. 1995: Blair et al.. 2003: Boice
et al.. 1999: Calvert et al.. In Press: Cano and Pollan. 2001: Lynge and Thygesen. 1990: Radican
et al.. 2008: Travier et al.. 2002). three effect estimates were between 0.86 and 1.03, and six
ranged from 1.46 to 3.76. Five cohort studies using relatively high quality exposure-assessment
methods reported the highest relative risks, but these studies were also based on only 2 to 18
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-240 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
exposed cases, so the estimates are imprecise: RR: 2.35 (95% CI: 0.52, 10.7) for females and
2.32 (95% CI 0.75, 7.15) for males in Radican et al. (2008); RR: 3.76 (95% CI: 0.77, 11.0) in
Antilla et al. (1995): RR: 1.70 (95% CI: 0.73, 3.34) in Boice et al.(1999); SIR: 2.02 (95% CI:
1.13, 3.34) for males and 1.14 (95% CI: 0.68, 1.68) for females in Selden and Ahlborg (2011):
and SMR: 2.46 (95% CI: 0.90, 5.36) in Calvert et al. (Calvert et al.. In Press). Results from the
case-control studies are also quite variable, with ORs ranging from 0.7 to 4.6 (Blair et al.. 1993:
Lynge. 2008: Mai one et al.. 1989: Miligi et al.. 2006: Miligi et al.. 1999)(Hardell et al., 1989;
Siemiatvcki. 1991)(Fabbro-Perav et al.. 2001) (Schenk et al.. 2009). The studies with the higher
quality exposure estimate reported ORs of 1.2 (95% CI: 0.6, 2.5) and 1.0 (95% CI: 0.7, 1.4)
(Lynge et al.. 2006: Miligi et al.. 2006). Both of the geographic studies provide some evidence
of an association between residential exposures via drinking water. Cohn et al. (1994b)reported
RR: 1.38 (95% CI: 1.08, 1.70) in females and RR: 1.20 (95% CI: 0.94. 1.52) for residence in a
town with municipal water supplies containing >5-ppb tetrachloroethylene. In the second, a
study of two towns with tetrachloroethylene and other solvents in the drinking water in Finland,
an association was seen in one town (SIR: 1.4, 95% CI: 1.0, 2.0) but not in the other (SIR: 0.6,
95% CI: 0.3, 1.1) (Vartiainen et al.. 1993). The ability of these studies to provide clear and
specific evidence pertaining to cancer hazard and tetrachloroethylene is limited by their
ecological designs and examination of several solvents in addition to tetrachloroethylene.
Six studies provide data pertaining to tetrachloroethylene and Hodgkin lymphoma (see
Table 4-31). Four cohort studies (Andersen et al.. 1999: Blair et al.. 2003: Pukkala et al.. 2009:
Travier et al.. 2002)) and one case-control study reported in two published papers (Costantini et
al.. 2001: Miligi et al.. 1999) examine risk among laundry and dry-cleaning workers, and one is a
geographic-based study of drinking water exposure in two towns in Finland (Vartiainen et al..
1993). No association is seen in the largest cohort, with 33 cases in the cohort of laundry and
dry-cleaning workers from 5 Nordic countries (SIR: 0.97, 95% CI: 0.67, 1.36) (Pukkala et al..
2009). A two- to threefold increased risk is seen in each of the smaller occupational studies,
with number of cases ranging from 4 to 19 (Andersen et al.. 1999: Blair et al.. 2003: Travier et
al.. 2002). The exposure-assessment methodology in these studies is relatively limited, and none
were considered to be of high quality.
The studies of multiple myeloma are summarized in Table 4-31. As was seen in the
compilation of studies of other types of lymphopoietic cancers, the larger cohort studies that use
a relatively nonspecific exposure measure (broad occupational title of launderers and dry
cleaners, based on census data) do not report an increased risk, with effect estimates ranging
from 0.99 to 1.07 (Ji and Hemminki. 2006: Pukkala et al.. 2009)((Andersen et al.. 1999)).
Results from the cohort and case-control studies with a higher quality exposure-assessment
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-241 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-31. Results of epidemiologic studies of potential tetrachloroethylene
exposure and adult Hodgkin lymphoma and multiple myeloma, by study
design
Cancer type,
n exposed
cases
Relative Risk
(95% CI)
Design, location, exposure assessment"
Reference
Hodgkin
Cohort
4
2.69(1.01,7.19)
laundry and dry-cleaning workers and pressers, Sweden,
census occupation codes
Travier et al. (2002)
5
2.0 (0.6,4.6)
laundry and dry-cleaning workers (SMR), United States,
union employment records
Blair et al. (2003)
19
1.88 (1.13,2.93)
laundry and dry-cleaning workers (SIR), Sweden,
Denmark, Finland, Norway, census occupation codes
(females)
Andersen et al.
(1999)
33
0.97 (0.67, 1.36)
laundry and dry-cleaning workers (SIR), Sweden,
Denmark, Finland, Norway, Iceland, census occupation
codes
Pukkala et al. (2009)
Hodgkin
Case-control
1
2.5 (0.3, 24.6)
Italy, job titles (launderer, dry cleaner, presser) (males)
Costantini et al.
(2001)
7
3.5 (1.5, 8.2)
Italy, job titles (launderer, dry cleaner, presser) (females)
Milieietal. (1999s)
Hodgkin
Geographic-based
6
0.8 (0.3, 1.7)
Finland (Hausjarvi), PCE in drinking water
Vartiainen (1993s)
11
1.4 (0.7, 2.5)
Finland (Hattula), PCE in drinking water
Vartiainen (1993s)
Multiple myeloma
Cohort
1
0.40 (0.01, 2.25)
aerospace workers (SMR), United States, job exposure
matrix (PCE routine exposure)3
Boice et al.. (1999s)
2
7.84 (1.43,43.1)
aircraft maintenance workers (RR-internal referent),
United States job exposure matrix (females)3
Radican et al. (2008s)
3
1.71 (0.42,6.91)
Aircraft maintenance workers (RR-internal referent),
United States, job exposure matrix (males)3
Radican et al. (2008s)
7
0.8 (0.3, 1.6)
laundry and dry-cleaning workers (SMR), United States,
union records (all workers)
Blair et al., (2003)
7
1.75 (0.70, 3.61)
laundry and dry-cleaning workers (SIR), Sweden, census
occupation codes
Lynge and Thygesen
(1990s)
36
1.07 (0.75, 1.45)
laundry and dry-cleaning workers (SIR), Sweden
(females)
Ji and Hemminki
(2006s)
45
1.0 (0.73, 1.34)
laundry and dry-cleaning workers (SIR), Sweden,
Denmark, Finland, Norway, census occupation codes
Andersen et al.
(1999s)
52
0.99 (0.66, 1.38)
laundry and dry-cleaning workers (SIR), Sweden (males)
Ji and Hemminki
(2006s)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-242 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Table 4-31. Results of epidemiologic studies of potential tetrachloroethylene
exposure and adult Hodgkin lymphoma and multiple myeloma, by study
design (continued)
Cancer type,
n exposed
cases
Relative Risk
(95% CI)
Design, location, exposure assessment"
Reference
152
1.02 (0.86, 1.20)
laundry and dry-cleaning workers (SIR), Sweden,
Denmark, Finland, Norway, Iceland, census occupation
codes
Pukkala et al. (2009)
Multiple myeloma
Case-control
3
1.0 (0.3,3.8)
Italy, job titles (launderer, dry cleaner, presser) (females)
Milisi et al. CI999)
9
6.0(1.7,21)
United States, all jobs held >12 mo (textile, apparel,
furnishing machine operators and tenders)
Gold etal., 2010a
16
1.5 (0.8, 2.9)
United States, all jobs held >12 mo, job exposure matrix
(PCE)a,b
Gold etal., 2010b
Multiple myeloma
Geographic-based
6
0.7 (0.2, 1.3)
Finland (Hattula), PCE in drinking water
Vartiainen (1993)
7
0.7 (0.3, 1.3)
Finland (Hausjarvi), PCE in drinking water
Vartiainen (1993)
" Studies with relatively high quality exposure-assessment methodologies, based on biological monitoring
data, cohort studies with job exposure matrix based on historical industrial monitoring data, or case-control
studies with job exposure matrix focusing on PCE based on information on job title and tasks or duties, and
additional modules for specific jobs, or studies of residential PCE exposure using a statistical model of
water distribution system to estimate delivered dose to a subject's home.
b Results for analysis in which low confidence jobs were considered unexposed. Similar results seen in the
primary analysis in which low confidence jobs were included in the exposure group.
methodology, with an exposure measure developed specifically for tetrachloroethylene, do
provide evidence of an association, however, with relative risks of 7.84 (95% CI: 1.43, 43.1) in
women and 1.71 (95% CI: 0.42, 6.91) in men in the cohort of aircraft maintenance workers
(Radican et al.. 2008) and 1.5 (95% CI: 0.8, 2.9) in the case-control study (Gold et al., 2010b).
Boice et al., (1999) also used a relatively high quality exposure measure, but because the results
are based on only one observed case, the imprecision of the estimate (RR: 0.40, 95% CI: 0.01,
2.25) limits this study for insights on multiple myeloma and tetrachloroethylene.
Variation in risk in relation to variation in exposure levels is examined in one study of
lymphopoietic cancer (Blair et al.. 2003). five studies of non-Hodgkin lymphoma (Blair et al..
1993: Boice et al.. 1999: Lynge et al.. 2006)( (Miligi et al.. 2006)) or of non-Hodgkin combined
with Hodgkin lymphoma (Seidler et al., 2006), four studies of multiple myeloma (Boice et al..
1999: Seidler et al.. 2007)(Gold et al., 2010a,b)and two studies of leukemia (Miligi et al.. 2006)
(Aschengrau and Seage. 2003) (see Table 4-32). Gold et al. (2010b) and Seidler et al. (2007)
examined exposure gradients using a cumulative tetrachloroethylene measure. The aerospace
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-243 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
worker cohort study by Boice et al. (1999), the dry cleaners cohort study by Blair et al. (2003),
and the Italian case-control studies (Costantini et al.. 2008; Miligi et al.. 2006)used a
semiquantitative measure of exposure intensity or frequency, and two studies used a less-specific
measure of job duration (Lynge et al.. 2006)(Gold et al., 2010a). Inability to account for
temporal changes in exposure intensity makes duration an inferior exposure surrogate compared
to semiquantitative or quantitative measures. The tetrachloroethylene-based measures in the
non-Hodgkin lymphoma studies (Boice et al.. 1999)((Miligi et al.. 2006) (Seidler et al..
2007)provide evidence of a higher risk at the higher exposure levels, particularly in the highest
category of cumulative exposure (>78.8 ppm-years) in the case-control study by Seidler et al.
(2007). Similar results are seen in one of the multiple myeloma studies (Gold et al., 2010b), but
the smaller study by Seidler et al. (2007) observed no cases among the highest exposure groups
(see Table 4-32).
There is considerable variation in the databases (e.g., number of studies, study design,
and quality of the exposure assessment) for the different types of lymphopoietic cancers. In
general, studies with relatively strong exposure assessments are based on a small number of
observed deaths or incident cases, with a relatively low statistical power resulting from few
observed events, or, for population case-control studies, low exposure prevalence. For
non-Hodgkin lymphoma and multiple myeloma, the presence of higher relative risk estimates in
studies with better exposure-assessment methodologies and evidence of an exposure-response
trend in one or more studies provide the basis for considering the collection of studies as
supportive of a role of tetrachloroethylene as a likely carcinogen. The collection of studies for
leukemia, non-Hodgkin lymphoma, Hodgkin lymphoma, and multiple myeloma is summarized
below.
There is little evidence for an association with leukemia. The two studies with a
relatively high quality exposure-assessment methodology had few exposed cases (<7) and did
not provide evidence of an association (RRs of 0.55 and 1.0 in Boice et al. (1999) and Costantini
et al. (2008). respectively), although a case-control study reported a twofold increased risk of
leukemia with the highest exposure level of tetrachloroethylene-contaminated drinking water
(Aschengrau et al.. 1993). The results from studies using more general (i.e., nonspecific)
exposure methods (e.g., occupational codes for laundry or dry-cleaning workers) generally
showed no association with leukemia (i.e., relative risk estimates <1.0 in 6 of the 9 cohorts)
(Blair et al.. 2003; Boice et al.. 1999; Lynge and Thygesen. 1990; Pukkala et al.. 2009; Sung et
al.. 2007)( (Andersen et al.. 1999)). Two of the increased leukemia relative risks (RR of 2.53
and 1.30) were seen in studies limited to female workers, which may represent a more
homogenous group in terms of potential exposures (Ji and Hemminki. 2005b; Travier et al..
2002).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-244 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-32. Results of epidemiologic studies of potential tetrachloroethylene
exposure and adult lymphopoeitic cancers, with data pertaining to exposure-
response gradients, by cancer type
Results
Design, location, exposure
assessment
Cancer type
Exposure measure
n
RR (95% CI)
Reference
Lymphopoeitic
Exposure score
Little to no
Medium to high
18
17
1.0 (0.6, 1.5)
0.9 (0.5, 1.4)
Cohort, laundry and dry-
cleaning workers, union
records (exposure score based
on proximity to washers)
Blair et al.
(2003)
Non-Hodgkin
Job duration (yr)
0
>0-10
145
5
3
14
15
1.0 (referent)
1.35 (0.44,4.14)
0.61 (0.17,2.21)
0.92 (0.49, 1.72)
0.66 (0.36, 1.22)
Nested case-control within
cohort of laundry and dry-
cleaning workers, Sweden,
Denmark, Finland, Norway,
census occupation codes3
Lynge et al.
(2006)
PCE (duration, yr)
0
<1
1-4
>5
32
4
6
10
1.0 (referent)
1.25 (0.43, 3.57)
1.11 (0.46,2.70)
1.41 (0.67, 3.00)
(trend p > 0.20)
Cohort, aerospace workers,
job exposure matrix (routine
or intermittent exposure to
PCE)
Boice et al.
(1999)
PCE (intensity)
Very low/low
Medium/high
18
14
0.6 (0.3, 1.2)
1.2 (0.6, 2.5)
(trend p = 0.72)
Case-control, Italy, job
exposure matrix
Miligi et al.
(2006)
PCE (duration, yr)
<15
>15
10
3
1.3 (0.5, 3.3)
not reported3
Case-control, Italy, job
exposure matrix
Miligi et al.
(2006)
PCE (cumulative, ppm-yr)
0
>0- <9.1
>9.1- <78.8
>78.8
67
16
14
6
1.0 (referent)
1.1 (0.5,2.3)
1.0 (0.5,2.2)
3.4 (0.7, 17.3)
(trend p = 0.12)
Case-control, Germany (PCE,
job exposure matrix)b
(Includes non-Hodgkin and
Hodgkin lymphoma; similar
results seen with
B-non-Hodgkin)
Seidler et al.
(2007)
Multiple
myeloma
Job duration (yr)
1-5
>5
4
5
3.6 (0.7, 1.7)
12 (1.3, 110)
(trendp < 0.01)
Case-control, United States,
all jobs held >12 mo (textile,
apparel, furnishing machine
operators and tenders)
Gold et al.,
2010a
PCE duration (yr)
0
<1
1-4
^5
24
1
4
1
1.0 (referent)
0.46 (0.06, 3.48)
1.13 (0.38, 3.35)
0.24 (0.03, 1.84)
(trend p < 0.01)
Cohort, aerospace workers,
job exposure matrix (routine
or intermittent exposure to
PCE)
Boice et al.
(1999)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-245 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-32. Results of epidemiologic studies of potential tetrachloroethylene
exposure and adult lymphopoeitic cancers, with data pertaining to exposure-
response gradients, by cancer type (continued)
Results
Design, location, exposure
assessment
Cancer type
Exposure measure
n
RR (95% CI)
Reference
Multiple
myeloma
(continued)
PCE duration (yr)
1-4
5-11
12-29
3-51
3
3
4
6
0.9 (0.2, 3.5)
2.0 (0.4, 9.2)
1.3 (0.3,4.6)
2.1 (0.7,6.8)
(trend p = 0.18)
Case-control, United States,
all jobs held >12 mo (PCE,
job exposure matrix)0
Gold et al.,
2010b
PCE (cumulative)
1-318
319-2,218
2,219-7,713
7,794-57,000
1
1
4
10
0.3 (0.04,3.0)
0.3 (0.1,4.4)
1.5 (0.4,5.4)
3.3 (1.2,9.5)
(trend p = 0.02)
Case-control, United States,
all jobs held >12 mo (PCE,
job exposure matrix)0
Gold et al.,
2010b
PCE (cumulative)
0
>0- <9.1 ppm-yr
>9.1- <78.8 ppm-yr
>78.8 ppm-yr
5
6
0
0
1.0 (referent)
1.8 (0.5,6.7)
(inverse trend
p = 0.34)
Case-control, Germany (PCE,
job exposure matrix)b
Seidler et al.
(2007)
Leukemia
PCE intensity
Very low/low
Medium/high
6
7
0.6 (0.2, 1.6)
1.0 (0.4,2.7)
Case-control, Italy, job
exposure matrix (PCE)
Costantini et
al. (2008)
Any PCE
RDD >90lh percentile
7
2
2.13 (0.88, 5.19)
8.33 (1.53,25.3)
Geographic based, United
States, water distribution
model (any PCE)
Aschengrau
et al. (1993)
" Relative risk estimates only reported for strata with at least five exposed cases.
b Cumulative score based on summation of the product of intensity score (low, 5 ppm; medium, 50 ppm; high,
200 ppm), frequency score (low, 3%; medium, 7.5%; high, 65%) of workweek, and duration for each job.
0 Results for analysis in which low confidence jobs were considered unexposed. Similar results seen in the primary
analysis in which low confidence jobs were included in the exposure group. Cumulative measure based on
summation of the product of intensity (ppm), frequency (h/wk), and duration (yr) for each job.
1 The results from the collection of studies pertaining to non-Hodgkin lymphoma indicate
2 an elevated risk associated with tetrachloroethylene exposure. The results from five cohort
3 studies that used a relatively high quality exposure-assessment methodology generally reported
4 relative risks between 1.7 and 3.8 (Anttila et al.. 1995; Boice et al.. 1999; Calvert et al.. In Press;
5 Radican et al.. 2008; Selden and Ahlborg. 2011) and support an association with
6 tetrachloroethylene. The studies with tetrachloroethylene-specific exposure measures and
7 exposure-response analysis (based on intensity, duration, or cumulative exposure) (Boice et al..
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-246 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
1999; Miligi et al.. 2006) (Seidler et al.. 2007) provide further support for an association,
reporting higher non-Hodgkin lymphoma risks in the highest exposure category, with the
strongest evidence from the large case-control study in Germany, in which a relative risk of 3.4
(95% CI: 0.7, 17.3) was seen in the highest cumulative exposure category (trend^-value = 0.12).
Lynge et al. (2006) distinguished dry cleaners from other workers but used an approach with
greater potential for misclassification because exposure was assigned only for jobs held in 1970.
This study did not report an association between dry cleaners and non-Hodgkin lymphoma, nor
did risk estimates increase with exposure duration. Relative risks in studies with broader
exposure assessments showed a more variable pattern (Blair et al.. 2003; Cano and Pollan. 2001;
Ji and Hemminki. 2006; Lynge and Thygesen. 1990; Pukkala et al.. 2009; S el den and Ahlborg.
2011; Travier et al.. 2002). Confounding by lifestyle factors are unlikely explanations for the
observed non-Hodgkin lymphoma results because common behaviors, such as smoking and
alcohol use, are not strong risk factors for non-Hodgkin lymphoma (Besson et al.. 2006; Morton
and Marianovic. 1984).
With respect to Hodgkin lymphoma, the data are more limited, with only four cohort
studies (Andersen et al.. 1999; Blair et al.. 2003; Pukkala et al.. 2009; Travier et al.. 2002). one
case-control study from Italy reported in two publications (Costantini et al.. 2001; Miligi et al..
1999). and one geographic-based study from Finland (Vartiainen et al.. 1993). None of the
exposure-assessment methods used in these studies were considered to be relatively high quality.
A two- to threefold increased risk is seen in all of the occupational studies except Pukkala et
al.(2009) [SIR: 0.97 (95% CI: 0.67, 1.36)].
The larger cohort studies that use a relatively nonspecific exposure measure (broad
occupational title of launderers and dry cleaners, based on census data) do not report an
increased risk of multiple myeloma, with effect estimates ranging from 0.99 to 1.07 (Andersen et
al.. 1999; Ji and Hemminki. 2006; Pukkala et al.. 2009). Some uncertainty in these estimates
arises from these studies' broader exposure-assessment methodology. Results from the cohort
and case-control studies with a higher quality exposure-assessment methodology, with an
exposure measure developed specifically for tetrachloroethylene, do provide evidence of an
association, however, with relative risks of 7.84 (95% CI: 1.43, 43.1) in women and 1.71 (95%
CI: 0.42, 6.91) in men in the cohort of aircraft maintenance workers (Radican et al.. 2008)and 1.5
(95% CI: 0.8, 2.9) in the case-control study in Washington (Gold et al., 2010b;
tetrachloroethylene exposure). Gold et al. (2010a, b) also reported increasing risks with
increasing exposure duration (based on job titles, Gold et al., 2010a) and based on a cumulative
tetrachloroethylene exposure metric (Gold et al., 2010b). Two smaller studies with
tetrachloroethylene-specific exposure measures based on intensity, duration, or cumulative
exposure did not observe an exposure-response trend: a study by Seidler et al. (2007) observed
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-247 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
no cases among the highest exposure groups, and a study by Boice et al. (1999) of aerospace
workers observed one death among routinely exposed subjects and six deaths among subjects
with a broader definition of routine or intermittent exposure.
4.6.1.2.5. Childhood leukemia
One cohort and four case-control studies are available on childhood leukemia (acute
lymphocytic leukemia, ALL) and parental occupational exposure to tetrachloroethylene or to
drinking water contaminated with trichloroethylene, tetrachloroethylene, and other chlorinated
solvents (Costas et al.. 2002; Infante-Rivard et al.. 2005; Shu et al.. 1999; Sung et al.. 2008)
(Lowengart et al.. 1987); Table 4-28; Appendix B. Some studies suggest a vulnerability for ALL
with maternal exposure either preconception or during pregnancy (Costas et al.. 2002; Lowengart
et al.. 1987; Shu et al.. 1999; Sung et al.. 2009). These studies, however, are insensitive for
assessing association, or lack thereof, between ALL and tetrachloroethylene exposure because
observations are based on a few exposed cases (all studies) or a weak exposure assessment (Sung
et al.. 2008). Only Lowengart et al. (1987) and Shu et al. (1999) examined paternal exposure
and tetrachloroethylene exposure with inconsistent observations. Other studies are needed to
clarify the role of tetrachloroethylene in ALL.
4.6.2. Animal Studies
4.6.2.1. Noncancer Effects
4.6.2.1.1. Immunotoxicity
The animal evidence for immunotoxicity following exposure to tetrachloroethylene is
very limited. These studies consist of mixed solvent exposures and some inhalation and oral
studies in which experimental animals were dosed with tetrachloroethylene alone.
Immune system parameters were altered in a mouse study (female B6C3Fi) administered
tetrachloroethylene by inhalation (maximum concentration: 6.8 ppm) along with a mixture of
24 contaminants frequently found in ground water near Superfund sites. Exposure lasted 14 or
90 days, and mice were sacrificed to assess immune system parameters. Evidence of
immunosuppression was observed, with a dose-related decrease in antibody response to sheep
red blood cells and decreased host resistance following challenge to Plasmodium yoelli. There
were no changes in lymphocyte number, T-cell subpopulations, NK cell activity, or in response
to challenge to Listeria monocytgens or PYB6 tumor cells. While these findings may be
attributed to B-cell/humoral immunity, these effects cannot be attributed to tetrachloroethylene
alone (Germolec etal.. 1989).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-248 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Aranyi et al. (1986) studied the effects of acute inhalation exposures to 25- or 50-ppm
tetrachloroethylene on two measures of immune response (susceptibility to respiratory infection
and mortality due to Streptoccocus zooepidemicus exposure and ability of pulmonary
macrophages to clear infection with Klebsiella pneumoniae). Female CD1 mice that were
5-7 weeks of age at the start of the exposure portion of the experiment were used for both
assays. Up to five replicate groups of about 30 mice were challenged with viable
S. zooepidemicus during simultaneous exposure to tetrachloroethylene or to filtered air. Deaths
were recorded over a 14-day observation period. Clearance of 35S-labeled K. pneumoniae by
pulmonary macrophages was determined by measuring the ratio of the viable bacterial counts to
the radioactive counts in each animal's lungs 3 hours after infection; 18 animals were used per
dose group. A single 3-hour exposure to 50-ppm tetrachloroethylene significantly increased the
susceptibility to respiratory infection and greater mortality following exposure to
S. zooepidemicus (p < 0.01). Forty-four deaths occurred in 140 (31.4%) mice challenged during
a 3-hour exposure to 50-ppm tetrachloroethylene; in contrast, 21 deaths occurred in 140 mice
(15.0%) exposed to filtered air. The 3-hour exposure to 50-ppm tetrachloroethylene was
associated with a statistically significant (p < 0.05) 6.6% decrease in pulmonary bactericidal
activity (80.5 and 13.9% of bacteria killed in controls and 50-ppm group, respectively). No
difference was seen in either mortality rate or bactericidal activity in experiments using a single
3-hour exposure to 25 ppm, or 3-hour exposures to 25-ppm tetrachloroethylene repeated daily for
5 days compared with control animals exposed to filtered air.
In a study by Hanioka et al. (1995b). atrophy of the spleen and thymus was observed in
rats receiving 2,000-mg/kg-day tetrachloroethylene via corn oil gavage for 5 days. No effect was
seen in the 1,000-mg/kg-day group. In a 14-day corn oil gavage (1,000 mg/kg-day) study of
tetrachloroethylene, no effects were observed on thymus and spleen weights of adult rats at a
dose that produced liver toxicity (Berman et al.. 1995). Another study employed 3 daily i.p.
doses of tetrachloroethylene to mice (Schlichting et al.. 1992). No effects were observed on ex
vivo natural killer cell activity or humoral responses of T-cells to exogenous mitogens.
Additional data from inhalation, oral, and dermal exposures of different durations are
needed to assess the potential immunotoxicity of tetrachloroethylene along multiple dimensions,
including immunosuppression, autoimmunity, and allergic sensitization. The data from Aranyi
et al. (1986) suggest that short-term exposures may result in decreased immunological
competence (immunosuppression) in CD-I mice. The relative lack of data, taken together with
the concern that other structurally related solvents (Cooper et al.. 2009) have been associated
with immunotoxicity, contributes to uncertainty in the database for tetrachloroethylene.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-249 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
4.6.2.1.2. Hematologic toxicity
Several studies by Marth et al. (1987; 1985; 1989) and a study by Seidel et al. (1992)
have demonstrated hematopoietic toxicity of tetrachloroethylene in female mice. In the Marth et
al. studies, 135 female NMRI mice were exposed in drinking water to tetrachloroethylene at 0.05
or 0.1 mg/kg per day beginning at 2 weeks of age for 7 weeks and examined 8 or 16 weeks after
exposure cessation. The mice exhibited a reversible hemolytic anemia and had microscopic
evidence of splenic involvement (Marth etal.. 1985c). Tetrachloroethylene was found to
accumulate in the spleen to a significantly greater extent than in the liver, brain, or kidney; levels
of tetrachloroethylene were 20-fold higher in spleen than in liver at the end of the exposure
period (Marth. 1987). Tetrachloroethylene was found in the spleen and fatty tissue of test
animals up to 2 months (56 days) after initial exposure (Marth et al.. 1989). Reversible body-
weight decreases and increases in the relative weight of the spleen compared with the kidneys
were reported. Serum triglycerides increased, and cholesterol levels decreased. These effects
persisted as long as 16 weeks after cessation of exposure. Liver function (as assessed by serum
protein levels) and hepatic protein synthesis were within normal limits, and there was no
evidence of hepatic fatty accumulation or necrosis. Compared with brain, kidney, or liver, the
erythropoietic system was found to be most susceptible to tetrachloroethylene in these studies.
Seidel et al. (1992) exposed female hybrid mice (C57/BL/6 x DBA/2) to
tetrachloroethylene at 270 ppm (11.5 weeks) and 135 ppm (7.5 weeks), 6 hours/day,
5 days/week. Reductions in the numbers of lymphocytes/monocytes and neutrophils were
observed, with a return to control values over the next 3 weeks. There were no effects on spleen
colony-forming units (CFU-Ss), but evidence of a reduction in red cells was supported by
decreases in erythroid colony-forming units and erythroid burst-forming units and evidence of
reticulocytosis. A partial regeneration was seen in the exposure-free follow-up period of
3 weeks. It was noted that the slight CFU-C depression, which persisted in the exposure-free
period, could indicate the beginning of a disturbance at all progenitor cell levels. These data
suggest a reversible bone marrow depression.
Hematological parameters were examined following oral administration of
tetrachloroethylene in sesame oil (3,000 mg/kg-day for 15 days) to male albino Swiss mice with
and without concurrent administration of 2-deoxy-D-glucose (2DG; 500 mg/kg-day i.p.), vitamin
E (400 mg/kg-day oral gavage) or taurine (100 mg/kg-day by oral intubation) (Ebrahim et al..
2001). This study was designed to examine the potential protective properties of 2DG and
vitamin E as well as taurine against tetrachloroethylene-induced cytotoxicity in various organ
systems. Animals exposed to tetrachloroethylene alone demonstrated significantly decreased
hemoglobin and RBC counts (p < 0.01), and significantly decreased HCT (packed cell volume)
and platelet counts (p < 0.001). The WBC count was found to be significantly increased
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-250 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
(p < 0.001). These changes were reverted back to near normal in the animals coexposed to 2DG,
vitamin E, or taurine.
In summary, the limited laboratory animal studies of hematological toxicity demonstrated
an effect of tetrachloroethylene exposure on RBC (decreased RBC (Ebrahim et al.. 2001). or
decreased erythrocyte colony-forming units (Seidel et al.. 1992)). with reversible hemolytic
anemia observed in female mice exposed to low drinking water levels (0.05 mg/kg-bw day) of
tetrachloroethylene beginning at 2 weeks of age in one series of studies (Marth. 1987; Marth et
al.. 1985c; Marth et al.. 1989). Ebrahim et al. (2001) also observed decreased hemoglobin,
platelet counts and packed cell volume, and increased WBC counts. Although limited studies are
available in the peer-reviewed published literature, the results of these studies support the results
seen in the study of dry-cleaning workers by Emara et al. (2010) described in Section 4.6.1.1.1.
4.6.2.2. Cancer Effects
4.6.2.2.1. Mononuclear cell leukemia in rats
The incidence of mononuclear cell leukemia in rats chronically exposed to
tetrachloroethylene is summarized in Table 4-33. The NCI oral gavage study in
Osborne-Mendel rats was considered to be inconclusive because of the high incidence of
respiratory disease, and high mortality with tetrachloroethylene exposure. Lesions indicative of
pneumonia were observed in almost all rats at necropsy. A high incidence of toxic nephropathy
was evident in tetrachloroethylene-exposed male and female rats. Early mortality was also seen
in tetrachloroethylene-exposed animals; 50% of the high dose males and females had died by
Weeks 44 and 66, respectively. Therefore, this bioassay is not considered further in the below
evaluation of the mononuclear cell leukemia induction by tetrachloroethylene in rats.
NTP (1986b) reported that the chronic inhalation administration of tetrachloroethylene at
concentration levels of 0, 200, and 400 ppm caused statistically significant positive trends in the
incidence of MCL in male (p = 0.004) and female (p = 0.018) F344/N rats. The incidences of
MCL in male and female rats exposed to tetrachloroethylene at 0, 200, and 400 ppm
(6 hours/day, 5 days/week, for 104 weeks) were 56, 77, and 74% and 36, 60, and 58%,
respectively. Interpretation of these data is somewhat complicated by the fact that overall
incidences of MCL in the concurrent chamber control groups were high relative to historical
chamber control groups at the performing laboratory (males: 28/50 [56%] vs. 117/250 [47%];
females: 18/50 [36%] vs. 73/249 [29%]). The concurrent control group rates were also higher
than the NTP program historical rate for untreated control groups (males: 583/1,977 [29%];
females: 375/2,021 [18%]).
To evaluate whether the increased MCL incidence contributed to the increase in early
deaths seen with increasing tetrachloroethylene exposure, NTP (1986b) conducted supplemental
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-251 DRAFT-DO NOT CITE OR QUOTE
-------
1 analyses according to their standard methods of data evaluation. These analyses considered the
2 progression of the disease, the effect of tetrachloroethylene on the time of onset of advanced
3 MCL, and the contribution of MCL to early deaths in control and dosed animals. The results of
4 these supplemental analyses showed the following:
Table 4-33. Mononuclear cell leukemia incidence in rats exposed to
tetrachloroethylene
Bioassay
Exposure
Sex
Mononuclear cell leukemia
incidence (%)a
NCI (1977)b
Osborne-Mendel rats
Gavage:
5 d/wk,
78 wk
Vehicle control
500 mg/kg-day
1,000 mg/kg-day
Male
None reported
Vehicle control
500 mg/kg-day
1,000 mg/kg-day
Female
None reported
NTP (1986b)
F344/N rats
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
200 ppm
400 ppm
Male
28/50 (56)
37/50 (77)
37/50 (74)
0 ppm
200 ppm
400 ppm
Female
18/50 (36)
30/50 (60)
29/50 (58)
JISA (1993)
F344/DuCij rats
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
50 ppm
200 ppm
600 ppm
Male
11/50 (22)
14/50 (28)
22/50 (44)
27/50 (54)
0 ppm
50 ppm
200 ppm
600 ppm
Female
10/50 (20)
17/50 (34)
16/50 (32)
19/50 (38)
a Reflects the number of animals with MCL reported under multiple organs," spleen, or liver.
b Gavage doses listed were adjusted several times during the course of the study. Male rats
received the listed TWA daily doses through Week 78, and surviving animals were observed up
to study termination in Week 110.
5 • In both males and females, tetrachloroethylene produced a dose-related increase in the
6 severity of MCL.
7 • Tetrachloroethylene exposure significantly shortened the time to onset of MCL in female
8 rats.
9 • Although there was no notable effect of tetrachloroethylene exposure on survival of
10 female rats, there was an increased incidence of advanced MCL in female rats that died
11 before the scheduled termination of the study. Thus, statistical analyses of only the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-252 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
incidences of advanced MCL in rats were considered. Significantly positive trends and
significant increases in the incidences of advanced MCL were observed in both male and
female rats in the high-dose groups.
Thomas et al. (2007) reanalyzed the NTP (1986b) dose-response data comparing results
with four statistical methods. In their analysis of MCL incidence in rats exposed to 500
chemicals, tetrachloroethylene was one of five chemicals shown by the authors to produce
—defutive" leukemia effects in both sexes of rats. MCL effects were more often than not
confined to one sex, while tetrachloroethylene induced statistically significant increases in both
sexes of the F344 rat. The findings in Thomas et al. (2007) described in more detail later, are
addressed in the context of other considerations in Section 4.6.2.2.2.
In the JISA (1993) study, F344/DuCij rats were exposed via inhalation for 104 weeks to
tetrachloroethylene at concentrations of 0, 50, 200, and 600 ppm. As in the NTP study, there
was a higher control incidence of MCL (22% in males and 20% in females) than the reported
historical rate of MCL for the Japanese laboratory of 147/1,149 [13%] in males and 147/1,048
[14.0%>] in females (see Table 5-10, Section 5). The incidence of MCL in male and female rats
exposed to tetrachloroethylene at 50, 200, and 600 ppm was 28, 44, and 54%> and 34, 32, and
38%), respectively. Both male and female rats displayed a significant dose-dependent increase in
MCL, at/? < 0.01 andp = 0.046 (poly-3 test, conducted for this assessment), respectively. There
was decreased latency in MCLs in female rats of the JISA study, with first appearance in
Week 100 in controls and Weeks 66-70 in treated rats.
4.6.2.2.2. Additional considerations regarding rodent leukemia findings
Under the conditions of the NTP and JISA bioassays, a carcinogenic effect of
tetrachloroethylene in male and female rats was evidenced by significant increases of MCL in
both sexes. The pathology of rat MCL is well characterized and has been well described
(Stromberg. 1985; Thomas et al.. 2007)Ward et al., 1990). MCL is among the most common
causes of death in the aging F344 rat and is readily and unequivocally diagnosed by standard
histopathological techniques. However, the utility of observed increases in MCL in the
chemically exposed rat for human carcinogenic risk assessment has been questioned for several
reasons. In particular, the spontaneous background incidence is both high and variable, and,
thus, can obscure chemical-induced increases. As noted in reviews by Caldwell (1999) and
Ishmael and Dugard (2006). the high background rate of MCL in control (untreated) rats can
limit the ability to separate the background response from possible chemically induced
responses, particularly when the chemically induced response above background is low.
Additionally, because high-incidence MCL occurs only in the F344 rat strain and not in mice,
Caldwell (1999) has stated that marginal increases in incidences are of questionable biological
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-253 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
significance. Supplemental analyses, such as have been conducted by NTP for
tetrachloroethylene and summarized in the preceding section, have been endorsed as a means to
aid in data interpretation for these commonly occurring tumors. In the paragraphs that follow,
issues pertinent to the interpretation of evidence that tetrachloroethylene induces MCL in male
and female rats for the purposes of human health risk assessment are addressed. The discussion
summarizes the findings of a recent analysis by Thomas et al. (2007) and considers the available
evidence for tetrachloroethylene in the context of the approach put forth by those authors. Other
considerations identified by NRC (2010) are also addressed, particularly with respect to
uncertainties surrounding the causes of F344 rat MCL, the biology of the disease, including the
cell type of origin, as well as the mechanisms by which tetrachloroethylene may advance
development of this rodent leukemia.
The significance of MCL findings in multiple NTP bioassays that used the F344 rat was
the subject of a recent reanalysis by Thomas et al. (2007). They examined the incidence of
leukemia in 2-yr bioassays that included untreated male and female F344 rats from 1971 to 1998.
They found that background tumor incidence increased substantially, from 7.9 to 52.5% in males
and from 2.1 to 24.2% in females, over that period. The reanalysis also found that MCL
responses are highly variable and subject to substantial modulation by dietary as well as other, as
yet unidentified, factors.
Their review of the disease pathobiology described MCL as a large granular lymphocytic
(LGL) leukemia that is a rapidly progressing and fatal neoplasm, with death typically occurring
within 2 weeks of onset (Thomas et al.. 2007). The disease is characterized by splenomegaly
upon gross pathological examination. Leukemic cell infiltration of the splenic red pulp with
variable lymphoid cell depletion is consistently seen. The tumor is transplantable; its etiological
factor is unknown. The cell of origin appears to reside in and/or require the splenic
microenvironment, and splenectomy dramatically reduces spontaneous MCL incidence
(Maloney and King, 1973).
Thomas et al. (2007) concluded that the exact cell of origin of F344 rat MCL is unknown.
The pathological characteristics of rat MCL are similar in some respects to one of the human
T-cell leukemias (Caldwell. 1999). and some investigators have proposed that MCL can serve as
an experimental model for human T-cell leukemia (Stromberg. 1985). However, MCLs have
been shown to be heterogenous with respect to cell phenotype and function (e.g., surface antigen
expression, esterase activity, and cytotoxic activity). For example, a study of 10 primary and
10 transplanted MCLs of aging rats found that natural killer (NK) cell activity was variable and
lacked correlation with surface antigens, with poorly differentiated MCL cells exhibiting less
cytotoxic (i.e., NK-cell) activity (Ward and Reynolds, 1983). These and other investigations
(e.g., Stromberg et al., 1983) have provided evidence that MCLs represent a heterogenous group
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-254 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
of leukemias. Thomas et al. note that the use of specific monoclonal anti-rat NK-cell antibodies
and other rat leukocyte specific markers would aid in establishing the cell type of origin. The
lack of assessment of the rodent tumors according to current classification criteria (e.g., as
specified by WHO [2008]) hinders ability to identify cell lineage. In particular, the lack of
immunophenotyping data for MCL occurring spontaneously or as the result of chemical
exposure, and the observed heterogeneity in cell phenotype and function of the spontaneously
occurring tumors studied thus far, greatly limit classification of MCL. Based on the reported
heterogeneity in cell phenotype and function, Thomas et al. (2007) stated that MCL may arise
from either mature LGLs or from a variety of individual LGL subpopulations; alternatively, a
pluripotent LGL precursor may be the cell type of origin.
Acknowledging the limitations that arise from the lack of knowledge about the cell type
of origin for MCL, and the observed heterogeneity in phenotype and function among MCL,
Thomas et al. (2007) characterize MCL as having an NK-cell phenotype based on functional
NK-cell activity in most (but not all) MCL cells. They note that human NK-LGL and F344 rat
MCL have -ame characteristics in common" and conclude that F344 rat MCL —is coiparable to
the aggressive human NK-LGL leukemia on a morphological, functional, and clinical basis."
However, current criteria to identify cell phenotype (e.g., by use of specific monoclonal
antibodies and genomic analysis) were not adopted in this study, and many of the comparison
criteria identify by Thomas et al. (2007) are nonspecific and common to other human leukemia
or lymphoma phenotypes. Although contrary to prior reports that the F344 MCL does not have a
human counterpart (e.g., Caldwell (1999)). a comparable conclusion regarding similarity of F344
rat MCL to human NK-LGL was reached by Stromberg (1985) and Ishmael and Dugard (2006).
Human NK-LGL is a rare form of LGL. NK-LGL usually occurs in younger patients (median
age: 39), has an aggressive clinical course, and is usually fatal within months of diagnosis
despite multiagent therapy. Epstein Barr virus has been implicated in many of the reported
NK-LGL cases, although the mechanism is unknown. In contrast, the majority of other human
LGLs (i.e., T-cell LGL leukemias) follow a chronic indolent course. Due to the paucity of
available data, mechanisms or modes of action contributing to the MCLs arising in untreated or
chemically exposed F344 rats have not been identified.
Thomas et al. (2007) also evaluated MCL incidence in male and female rats exposed to
500 chemicals. On the basis of 34 NTP studies that yielded evidence of a chemically related
increase in the incidence of leukemia, which included the 1986 NTP study of
tetrachloroethylene, the authors conducted a reanalysis of dose-response data by comparing
results with four statistical methods: Fisher exact test for pair-wise comparison of leukemia
incidence between a dose group and a control group, the Cochran-Armitage test for incidence
trend, logistic regression for incidence, and life tables for survival-adjusted incidence.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-255 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Tetrachloroethylene was one of five chemicals shown by the authors to produce —dfinitive"
leukemia effects in both sexes of rats. MCL effects were more often than not confined to one
sex, while tetrachloroethylene induced statistically significant increases in both sexes of the F344
rat.
In their analysis, Thomas et al. (2007) employed the rigid statistical criteria suggested in
Food and Drug Administration (FDA) guidance for testing dose-related cancer incidences of
common tumors (p < 0.01 for pairwise comparison; p < 0.005 for trend test). They noted that
leukemia is generally considered a fatal neoplasm, thus supporting the life table test as more
likely reflecting the true statistical significance of the carcinogenic effect. Life-table analysis
(log-rank test) accounts for time-to-event information, is capable of testing nonlinear dose-
response relationships of arbitrary shapes, and is, therefore, more flexible than the
Cochran-Armitage trend test. The 1986 NTP results in male rats exposed to tetrachloroethylene
revealed a significant dose-response trend when analyzed with a life table analysis (p = 0.004)
assuming that MCL is lethal (a nonsignificant trend with logistic regression (p = 0.097) resulted
if MCL was assumed nonlethal). Pairwise comparisons revealed dose-related incidences
(p = 0.046; Fisher exact test) for both dose groups, and the Cochran-Armitage trend test yielded a
p-yalue of 0.034; neither met the FDA criteria for statistical significance. The borderline
significance of the trend test and nonsignificance of logistic regression for the latter two
comparisons could be explained, in part, by the fact that the incidences did not follow an
incrementally increasing relationship with dose. In female rats in the NTP study, use of a life
table (p = 0.053), logistic regression (p = 0.012), a trend test (p = 0.018), and Fisher exact test
(p = 0.014 and 0.022, respectively, for two doses) all revealed dose-related increases in incidence
that were of borderline significance according to the suggested FDA criteria.
Thomas et al. (2007) note that NTP does not use a rigid statistical rule in interpreting
experimental results, instead relying on consideration of other factors in a weight-of-evidence
approach. These factors include historical control incidences, and whether chemically induced
tumors were sex-specific, dose-responsive, of shorter latency, or of more advanced stage. While
encouraging stringent statistical analysis to reduce false positives, Thomas et al. (2007)
characterized the NTP weight-of-evidence approach as —japropriate" and "rigorous." They
proposed a similar evaluation of the pertinent data, to also include consideration of such factors
as reproducibility of effect across bioassays, and other information to inform biological
plausibility (i.e., evidence of toxic or carcinogenic effects on LGLs or their precursors). An
assessment of the considerations identified by Thomas et al. (2007) and NRC (2010) for
tetrachloroethylene is provided below:
Nature of the dose-response curve in terms of incidence and severity. The NTP study
found that tetrachloroethylene increased the incidence and severity of MCL in male and female
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-256 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
rats. The JISA study reported an increasing trend incidence of MCL in both male and female
rats, and overall the number of early deaths attributed to MCL increased with increasing
exposure.
Appropriate historical control data. Historical control data are available from the
laboratory that performed the NTP study, the NTP program, and from the Japanese laboratory.
A comparison with historical data revealed a higher MCL rate in concurrent controls in the NTP
and Japanese tetrachloroethylene bioassays. Concurrent controls in the NTP studies were higher
than historical chamber control groups at the performing laboratory (males: 28/50 [56%] vs.
117/250 [47%]; females: 18/50 [36%] vs. 73/249 [29%]). The concurrent control group rates
were also higher than the NTP program historical rate for untreated control groups (males:
583/1,977 [29%]; females: 375/2,021 [18%]). As in the NTP study, there was a higher control
incidence of MCL (22% in males and 20% in females) than the reported historical rate of MCL
for the Japanese laboratory of 147/1,149 [13%] in males and 147/1,048 [14.0%] in females (see
Table 5-10, Section 5).
Reduction in latency time. The NTP study found that tetrachloroethylene reduced tumor
latency in female rats. In the JISA study, there was also decreased latency in MCLs in female
rats, with the first appearance in Week 100 in controls and Weeks 66-70 in treated rats.
Reproducibility in another species and routes of exposure. Tetrachloroethylene has
reproducibly been found to be carcinogenic in rats and mice. Tetrachloroethylene was
carcinogenic when tested in mice in an oral gavage study (NCI. 1977) and in two inhalation
studies (NTP and JISA), inducing hepatic neoplasms. Tetrachloroethylene also caused other
types of tumors in the F344 rat. However, tetrachloroethylene has only been found to be
leukemogenic in F344 rat studies. In the JISA study, deaths in female mice due to malignant
lymphomas/total dead (or moribund) mice were 6/18, 4/20, 13/27, and 10/33 in the 0-, 10-, 50-,
and 250-ppm groups, respectively. Tetrachloroethylene exposure did not affect the incidence at
study termination of malignant lymphomas in the lymph nodes or spleen. The NTP study also
did not find an effect of tetrachloroethylene on malignant lymphoma incidence in female mice.
A similar lack of site concordance across rodent bioassays was also seen among many of
the NTP chemicals causing MCL in F344 rats reviewed by Thomas et al. (2007).
Tetrachloroethylene was among six chemicals (the others were allyl isovalerate, bisphenol A,
pyridine, 2,4,6-trichlorophenol, and the benzene metabolite hydroquinone) for which leukemia
was the only neoplastic change for either male or female rats, but for which other sex-species
groups showed evidence of carcinogenicity (Thomas et al.. 2007). (Note that, as discussed in
Section 4.10, elevated incidences of other tumors—specifically, brain gliomas and kidney tubule
adenomas and adenocarcinomas—were observed in male F344/N rats in the tetrachloroethylene
NTP study but were not included in the Thomas et al. (2007) analysis.) For eight other
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-257 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
chemicals evaluated by Thomas et al. (2007), F344 rat MCL was the only carcinogenic effect in
rats or mice. For twenty chemicals, MCL was one of multiple neoplastic changes in F344 rats of
one or both sexes.
Involvement of both sexes. Tetrachloroethylene induced MCL in both sexes of F344 rats
in the NTP and JISA inhalation bioassays. In fact, tetrachloroethylene was one of only
5 chemicals identified in a review of 500 chemicals by Thomas et al. (2007) that were shown to
produce —dafiitive" leukemia effects in both sexes of rats. Tetrachloroethylene was also
hepatocarcinogenic in both sexes of mice in the available oral (NCI. 1977) and inhalation
bioassays (NTP and JISA). Hence, the carcinogenic effects of tetrachloroethylene are evident in
both male and female rodents across multiple data sets and with tumor sites.
Comparative species metabolism. Species differences in metabolism of
tetrachloroethylene have been noted, as reviewed in Section 3. Although thought to be
qualitatively similar, there are clear differences among species in the quantitative aspects of
tetrachloroethylene metabolism (Ikeda and Ohtsuii. 1972; Lash and Parker. 2001; Schumann et
al.. 1980; U.S. EPA. 1991b; Volkel et al.. 1998). These differences are in the relative yields and
kinetic behavior of metabolites (Green et al.. 1990; Ohtsuki et al.. 1983; U.S. EPA. 1985a;
Volkel et al.. 1998). Because metabolites are thought to contribute to the carcinogenicity of
tetrachloroethylene, these differences in metabolism are likely to contribute to species
differences in carcinogenic response, including the types of tumors observed across rodent
bioassays.
The metabolite(s) contributing to the development of MCL from tetrachloroethylene have
not been defined. A role for GSH-derived metabolites was posited based on early reports of fatal
hemorrhagic disease in cattle fed trichloroethylene-extracted soybean oil meal, and the
subsequent finding that the trichloroethylene metabolite »Y-(1,2,-dichlorovinyl)-/.-cysteine
(generated through the GSH pathway) induces renal toxicity, aplastic anemia, and marked DNA
alteration in bone marrow, lymph nodes, and thymus in calves (Bhattacharya and Schultze. 1971.
1972). However, similar effects were not found in a study that administered TCVC, a
GSH-derived metabolite of tetrachloroethylene, to two calves as a single dose (Lock et al..
1996). The first calf received 10 mg/kg i.v. (40 (j,mol/kg) and was observed for 25 days and then
given a second dose of 8 mg/kg (36 (j,mol/kg) and observed for a further week. A second calf
was given 18 mg/kg (72 (j.mol/kg) and observed for 20 days. An initial neutropenia was seen in
the first calf during the first few days after dosing. However, no decline in platelet or neutrophil
count, nor elevation in blood urea nitrogen, was observed. Based on clinical and
histopathological evaluation, TCVC was concluded to lack bone marrow or kidney toxicity. The
authors characterized the lack of toxicity in the kidney as -puzzling" given their prior work
demonstrating the nephrotoxicity of comparable TCVC exposures in the rat (Ishmael and Lock
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-258 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
1986), and their concurrent in vitro studies showing that TCVC, like DCVC, was toxic to renal
transport mechanisms in cortical slices (LocketaL 1996). Toxicokinetic differences among
species were postulated as an explanation for the observed species differences in TCVC
sensitivity, and the unique sensitivity of the calf to DCVC compared with TCVC and other
haloalkene conjugates. Aside from the Lock et al. (1996) evaluation of bone marrow toxicity of
TCVC in the juvenile cow, a species of unknown sensitivity to tetrachloroethylene-induced
leukemia, other studies aimed at elucidating the active metabolites contributing to leukemic
effects have not been reported. In particular, no such studies are available in the F344 rat, the
species and strain in which leukemic effects have been consistently observed in both sexes.
Analyses of how differences in metabolism may lead to differences in the
leukemogenicity of tetrachloroethylene across species are limited by this lack of knowledge
regarding the putative leukemogenic metabolites. As reviewed in Section 3, tetrachloroethylene
is metabolized by two main pathways, oxidation and GSH conjugation. Species differences in
the extent of metabolism, and in the profile of resultant metabolites, have been observed in both
pathways. Metabolism is higher in mice than in rats, predominantly owing to more extensive
metabolism via the oxidative pathway thought to contribute to hepatic toxicity and
carcinogenicity. Rats, in turn, have higher metabolic rates than do larger animals, including
humans. The half-life of tetrachloroethylene is much longer in humans (>100 hours) than in
rodents (<10 hours). Interindividual differences in metabolism, for instance arising from
variability in activity of GSTs and other metabolic enzymes, may also contribute to interspecies
differences in metabolism. Overall, the database is insufficient to characterize how these
metabolic differences may impact species sensitivity to the leukemogenic activity of
tetrachl oroethy lene.
Genotoxicity, cytotoxicity, and any other relevant information. Thomas et al. (2007) note
—iitle evidence to support a mode of action" for F344 rat MCL induced either spontaneously or
by the 34 leukemogens they reviewed, including tetrachloroethylene. However, they propose a
review of evidence that may aid in assessing the biological plausibility for tumor induction. The
genotoxicity of tetrachloroethylene is reviewed in Section 4.8. None of the reviewed studies
have specifically investigated the genotoxicity of tetrachloroethylene in the potential target tissue
(bone marrow or spleen) of the F344 rat of either sex. A study in Sprague-Dawley rats found
only marginal effects on chromosomal aberrations and aneuploidy with tetrachloroethylene
exposure by inhalation (100 and 500 ppm) (Beliles et al.. 1980). However, the overall
conclusion for tetrachloroethylene genotoxicity supports the view that the contribution of
mutagenicity to one or more carcinogenic outcomes cannot be ruled out.
No studies are available that evaluate the toxicity of tetrachloroethylene in the putative
target tissues (bone marrow and/or spleen) or target cells of MCL in the F344 rat. However, as
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-259 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
reviewed in Section 4.6.2.1.2, several studies by Marth et al. (Marth. 1987; Marth et al.. 1985c;
Marth et al.. 1989). Seidel et al. (1992), and Ebrahim (2001) have demonstrated hematopoietic
toxicity of tetrachloroethylene in mice. Ebrahim et al. (2001) found that tetrachloroethylene in
sesame oil (3,000 mg/kg-day for 15 days) significantly decreased hemoglobin, RBC counts,
decreased HCT (packed cell volume) and platelet counts, and significantly increased WBC
count. These findings are similar to those observed in studies of tetrachloroethylene-exposed
humans (Emara et al.. 2010). In the Marth et al. studies, female NMRI mice exhibited a
reversible hemolytic anemia and had microscopic evidence of splenic involvement following
exposure to low drinking water levels (0.05 mg/kg-bw day) of tetrachloroethylene beginning at 2
weeks of age. Seidel et al. (1992) also found evidence of a reduction in red cells, supported by
decreases in erythroid colony-forming units and erythroid burst-forming units and evidence of
reticulocytosis in female hybrid mice (C57/BL/6 x DBA/2) to tetrachloroethylene at 270 ppm
(11.5 weeks) and 135 ppm (7.5 weeks), 6 hours/day, 5 days/week. Reversible reductions in the
numbers of lymphocytes/monocytes and neutrophils were also observed. The slight CFU-C
depression, which persisted in the exposure-free period, could indicate the beginning of a
disturbance at all progenitor cell levels. These data suggest a reversible bone marrow
depression.
A number of leukemogens (e.g., benzene) have been reported to inhibit production of
both red cells and various forms of white cells. A decrease in CFU-Ss, an effect not observed
with tetrachloroethylene exposure (Seidel et al.. 1992). has commonly been reported.
Leukemogens also cause a decrease in bone marrow myeloid progenitors CFU-GEMM,
CFU-GM, and CFU-E/BFU-E, the latter of which was also decreased by tetrachloroethylene
(Seidel et al.. 1992). Thus, Seidel et al. (1992) provides indirect evidence that
tetrachloroethylene induces effects associated with leukemogens (NRC. 2010).
Other studies that may be relevant to leukemia induction in the F344 rat include those of
the immunotoxicity of tetrachloroethylene. However, the available database of such studies, as
summarized in Section 4.6.2.1.1, is limited for establishing whether tetrachloroethylene affects
immune parameters in a manner indicative of potential for inducing leukemia development.
Immunosuppression was seen in female B6C3Fi mice administered tetrachloroethylene
(maximum concentration: 6.8 ppm) with a mixture of 24 frequent contaminants of ground water
near Superfund sites (Germolec et al.. 1989). No changes were evident in lymphocyte number,
T-cell subpopulations, NK cell activity, or with challenge by Listeria monocytgens or PYB6
tumor cells. In a separate inhalation study in mice, exposure to 170-mg/m3 (50-ppm)
tetrachloroethylene for 3 hours increased susceptibility to respiratory streptococcus infection and
significantly decreased pulmonary bactericidal activity (Aranyi et al.. 1986).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-260 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
As reviewed by Thomas et al. (2007). corn oil gavage has been shown to significantly
(p < 0.001) decrease the incidence of MCL in F344 rats, particularly males, by an unknown
mechanism. This complicates interpretation of the few short-term studies in rats administering
tetrachloroethylene in corn oil gavage. These include a finding of atrophy of the spleen and
thymus in rats receiving 2,000 (but not 1,000) mg/kg-day tetrachloroethylene via corn oil gavage
for 5 days (Hanioka et al.. 1995b). In a separate 14-day corn oil gavage study,
tetrachloroethylene did not affect thymus and spleen weights of adult rats at a hepatotoxic dose
(1,000 mg/kg-day) (Berman et al.. 1995).
Summary. This assessment of considerations proposed in Thomas et al. (2007)and by
NRC (2010) highlights several findings that add support to the conclusion that
tetrachloroethylene is a leukemogen in the F344 rat. Particularly pertinent are findings of the
evaluation by NTP of the 1986 inhalation bioassay of tetrachloroethylene, demonstrating dose-
related increases in the incidence of MCL in both sexes and in the severity of MCL in both
sexes, as well as a shortened time to onset of MCL in female rats, and an increased incidence of
advanced MCL in female rats that died before the scheduled termination of the study. These
factors are considered the most important in evaluating the significance of the MCL findings for
tetrachl oroethy lene.
Additional factors supporting the carcinogenicity of tetrachloroethylene include the
observation that tetrachloroethylene has also been found to induce other rare tumors besides
MCL in the F344 rat, as well as tumors at other sites in both sexes of the mouse, in both
inhalation and oral gavage bioassays. As noted by Thomas et al. (2007). chemically induced
MCL has typically been found in only one sex of the F344 rat, and tetrachloroethylene was one
of only 5 chemicals identified in their review of 500 chemicals in the NTP database to
definitively cause the tumor in both males and females. These findings add support to the
conclusion that tetrachloroethylene is a rodent carcinogen, and that increased MCL observed in
both sexes of the F344 rat is not a spurious finding. Although limited, studies demonstrating
hemolysis and bone marrow toxicity in mice add some support to the biologic plausibility of the
observed leukemic effects (NRC. 2010). The pharmacokinetics (metabolites) and
pharmacodynamics (biological mechanisms) that contribute to the development of MCL in the
F344 rat, both spontaneously and with chemical exposure, have not yet been elucidated.
Uncertainties remain regarding the causes of F344 rat MCL, the biology of the disease
including the cell type of origin, as well as the mechanisms by which tetrachloroethylene may
advance development of this rodent leukemia. Further research to clarify the factors that affect
inherent and chemically induced susceptibility to F344 rat MCL is warranted. As proposed by
Stromberg (1985). the F344 rat MCL could serve as a rodent model for human T-cell leukemias,
in which research could be conducted to identify causative factors and disease mechanisms, and
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-261 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
to test and develop novel chemotherapies. Thomas et al. (2007) similarly endorsed additional
research and analyses of F344 leukemogens, such as tetrachloroethylene, to advance
understanding of the mechanisms contributing to the rodent—and by inference, the related
human—diseases.
In summary, although uncertainties remain regarding the pathobiology of MCL and the
mechanisms by which tetrachloroethylene may contribute to disease development and/or
progression, this assessment of additional factors bolsters the support for the finding of
tetrachloroethylene-induced MCL in the F344 rat. It is concluded that the use of this tumor to
identify human carcinogenic hazard and to estimate risks from carcinogen exposure is adequately
supported.
4.6.3. Summary and Conclusions
4.6.3.1. Immunotoxicity, Hematologic Toxicity, and Cancers of the Immune System in
Humans
The strongest epidemiological study examining immunologic and hematopoietic effects
of tetrachloroethylene exposure in terms of sample size and use of an appropriately matched
control group is of 40 male dry-cleaning workers (mean exposure levels <140 ppm; mean
duration: 7 years; mean blood tetrachloroethylene levels: 1,685 |ig/L) by Emara et al. (2010).
Statistically significant decreases in red blood cell count and hemoglobin levels and increases in
total white cell counts and lymphocyte counts were seen in the exposed workers compared to
age- and smoking-matched controls. Similar effects were seen in mice (Ebrahim et al.. 2001). In
addition, increases in several other immunological parameters, including T-lymphocyte and
natural killer cell subpopulations, IgE, and interleukin-4 levels were observed in
tetrachloroethylene-exposed dry-cleaning workers (Emara et al.. 2010). These immunologic
effects suggest an augmentation of Th2 responsiveness. However, the limited available data
from studies in children (Delfino et al.. 2003a: Delfino et al.. 2003b: Lehmann et al.. 2001:
Lehmann et al.. 2002) do not provide substantial evidence of an effect of tetrachloroethylene
exposure during childhood on allergic sensitization or exacerbation of asthma symptomology.
The observation of the association between increased tetrachloroethylene exposure and reduced
interferon-y in cord blood samples may reflect a sensitive period of developmen, and points to
the current lack of understanding of the potential immunotoxic effects of prenatal exposures.
The available data pertaining to risk of autoimmune disease in relation to tetrachloroethylene
exposure are limited by issues regarding ascertainment of disease incidence and exposure-
assessment difficulties in population-based studies. In summary, there is considerable variation
in the extent and quality of the epidemiologic literature (e.g., number of studies, study design,
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-262 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
and quality of the exposure assessment) for lymphopoeitic cancers. In general, studies with
relatively strong exposure assessments are based on a small number of observed deaths or
incident cases, with a relatively low statistical power. For non-Hodgkin lymphoma and multiple
myeloma, the available studies are considered supportive of a role of tetrachloroethylene as a
likely carcinogen. This is based on the presence of higher relative risk estimates in studies with
better exposure-assessment methodologies and evidence of an exposure-response trend in one or
more studies.
Among the specific types of lymphopoeitic cancers, there is considerable variation in the
extent and quality of the epidemiologic literature (e.g., number of studies, study design, and
quality of the exposure assessment). In general, studies with relatively strong exposure
assessments are based on a small number of observed deaths or incident cases, with a relatively
low statistical power. For non-Hodgkin lymphoma and multiple myeloma, the presence of
higher relative risk estimates in studies with better exposure-assessment methodologies and
evidence of an exposure-response trend in one or more studies provide the basis for considering
the collection of studies as supportive of a role of tetrachloroethylene as a likely carcinogen.
For non-Hodgkin lymphoma, there is little evidence of an association in the large cohort
studies examining risk in relation to the broad occupational category of work in laundry or dry
cleaning (i.e., relative risk estimates ranging from 0.95 to 1.05 in females in Andersen et al.
(1999). females and males in Ji and Hemminki (2006). and Pukkala et al.(2009). The results
from the four cohort studies that used a relatively higher quality exposure-assessment
methodology, however, reported relative risks between 1.7 and 3.8 (Boice et al.. 1999) (Anttila et
al.. 1995; Radican et al.. 2008). There is also some evidence of exposure-response gradients in
studies with tetrachloroethylene-specific exposure measures based on intensity, duration, or
cumulative exposure (Boice et al.. 1999; Miligi et al.. 2006; Seidler et al.. 2007). Higher
non-Hodgkin lymphoma risks were seen in these studies in the highest exposure categories, with
the strongest evidence from the large case-control study in Germany in which a relative risk of
3.4 (95% CI: 0.7, 17.3) was seen in the highest cumulative exposure category (trend/>value =
0.12) (Seidler et al.. 2007). Confounding by lifestyle factors are unlikely explanations for the
observed results because common behaviors, such as smoking and alcohol use, are not strong
risk factors for non-Hodgkin lymphoma (Besson et al.. 2006; Morton and Marianovic. 1984).
Results from the multiple myeloma studies are based on a smaller set of studies than
those of non-Hodgkin lymphoma, but results are similar. The larger cohort studies that use a
relatively nonspecific exposure measure (broad occupational title of launderers and dry cleaners,
based on census data) do not report an increased risk of multiple myeloma, with effect estimates
ranging from 0.99 to 1.07 (Ji and Hemminki. 2006; Pukkala et al.. 2009)((Andersen et al..
1999)). Results from the cohort and case-control studies with a higher quality exposure-
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-263 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
assessment methodology, with an exposure measure developed specifically for
tetrachloroethylene, do provide evidence of an association, however, with relative risks of 7.84
(95% CI: 1.43, 43.1) in women and 1.71 (95% CI: 0.42, 6.91) in men in the cohort of aircraft
maintenance workers (Radican et al.. 2008)and 1.5 (95% CI: 0.8, 2.9) in the case-control study in
Washington (Gold et al., 2010b; tetrachloroethylene exposure). Gold et al. (2010a, b) also
reported increasing risks with increasing exposure duration based on job titles, Gold et al.,
2010a) and based on a cumulative tetrachloroethylene exposure metric (Gold et al., 2010b).
Two smaller studies did not observe an exposure-response trend: a study by Seidler et al. (2007)
observed no cases among the highest exposure groups, and a study by Boice et al. (1999) of
aerospace workers observed one death among routinely exposed subjects and six deaths among
subjects with a broader definition of routine or intermittent exposure.
4.6.3.2. Immunological and Hematological Toxicity and Mononuclear Cell Leukemias in
Rodents
Additional data from inhalation, oral, and dermal exposures of different durations are
needed to assess the potential immunotoxicity of tetrachloroethylene along multiple dimensions,
including immunosuppression, autoimmunity, and allergic sensitization. The data from Aranyi
et al. (1986) suggest that short-term exposures may result in decreased immunological
competence (immunosuppression) in CD-I mice. The relative lack of data taken together with
the concern that other structurally related solvents (Cooper et al.. 2009) have been associated
with immunotoxicity contributes to uncertainty in the database for tetrachloroethylene.
The limited laboratory animal studies of hematological toxicity demonstrated an effect of
tetrachloroethylene exposure on RBC (decreased RBC (Ebrahim et al.. 2001). or decreased
erythrocyte colony forming units (Seidel et al.. 1992)). with reversible hemolytic anemia
observed in female mice exposed to low drinking water levels (0.05 mg/kg-bw day) of
tetrachloroethylene beginning at 2 weeks of age in one series of studies (Marth. 1987: Marth et
al.. 1985c: Marth et al.. 1989). Ebrahim et al. (2001) also observed decreased hemoglobin,
platelet counts and packed cell volume, and increased WBC counts.
Cancer findings of primary concern are the statistically significant increases in MCL in
both sexes in the NTP (1986b) and JISA (1993) inhalation bioassays. Section 4.6.2.2.2
addresses issues pertinent to the interpretation of evidence that tetrachloroethylene induces MCL
in male and female rats for the purposes of human health risk assessment. That discussion
summarizes the findings of a recent analysis by Thomas et al. (2007) and considers the available
evidence for tetrachloroethylene in the context of the approach put forth by those authors and by
NRC (2010). This included a summary of the available noncancer studies that may inform the
biologic plausibility of the leukemia findings. In the paragraphs that follow, the findings in and
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-264 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
statistical analyses of the rodent bioassays are presented, and the other factors and data
considered in the analysis presented in Section 4.6.2.2.2 are then summarized. Together, these
analyses informed the conclusions provided concerning the application of the F344 rat leukemia
data to human health risk assessment.
Statistical analysis of the NTP bioassay revealed a statistically significant trend for males
(p = 0.004), and a marginally significant trend for females (p = 0.053). Life table analysis
disclosed statistically significant increases in both the low- and high-dose groups in males. A
significant increase in the low-dose group (p = 0.023) and a marginally significant increase in the
high-dose group (p = 0.053) was seen in females. Additional statistical analyses reported by
Thomas et al. (2007) of the female rat data from the NTP study found the results significant by
logistic regression (p = 0.012), the Cochran-Armitage trend test (p = 0.018), and Fisher exact test
(p = 0.014 and 0.022, respectively, for the lower and higher doses). Similarly, additional
analyses reported by Thomas et al. (2007) supported the statistical significance of the male rat
NTP data [logistic regression (p = 0.097), the Cochran-Armitage trend test (p = 0.034), and
Fisher exact test (p = 0.046 for the lower and higher doses)]. Notably, these statistical analyses
supported the authors' classification of tetrachloroethylene as one of only five chemicals of the
500 examined to produce "definitive" leukemia effects in both sexes of rats. While MCL effects
were more often than not confined to one sex, tetrachloroethylene induced statistically
significant increases in both sexes of the F344 rat.
In the JISA (1993) bioassay, no incremental increase in MCL incidence was seen in
female rats with increasing dose, although MCL showed a marginally significant trend with
dose. In contrast, male rats displayed a significant dose-dependent increase in MCL. Because
MCL is a rapidly progressing and fatal neoplasm, Thomas et al. (2007) and NRC (2010)
supported the life table test as more likely reflecting the true statistical significance of the
carcinogenic effect. However, the Japan bioassay report did not include an analysis of the tumor
latency, and, thus, life table statistical analysis was not possible.
Other factors besides statistical analyses can inform interpretation of bioassay data and
the observed effects of chemical exposures. According to NTP practices, as reviewed in Thomas
et al. (2007). bioassay evaluation includes consideration of factors such as historical control
tumor incidences, and whether chemically induced tumors were sex-specific, dose-responsive, of
shorter latency, or of more advanced stage. NTP analyses of the tetrachloroethylene bioassay
results revealed a dose-related increase in the incidence of MCL in both sexes, in the severity of
MCL in both sexes, a shortened time to onset of MCL in female rats, and an increased incidence
of advanced MCL in female rats that died before the scheduled termination of the study. All of
these findings elevate concern that the MCL findings are related to chemical exposure, and
among factors considered, add significant support to the conclusion that tetrachloroethylene is a
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-265 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
leukemogen in F344 rats. An additional consideration in evaluation of the NTP and JISA studies
is that a higher MCL incidence was seen in concurrent controls compared with historical
controls. The reason for the reportedly higher MCL incidence in concurrent controls in these
bioassays is not known. However, the finding of a chemically induced effect in a bioassay with
a high background rate, which is more likely to obscure chemically induced findings, supports
the conclusion that the observed tumors are due to tetrachloroethylene exposure. The
independent findings of MCL induction in two bioassays conducted by separate laboratories also
strengthen the conclusions.
Available pharmacokinetic data are insufficient to identify the active metabolite(s) of
tetrachloroethylene that contribute(s) to MCL development. Such data are also insufficient to
inform analyses of how interspecies differences in metabolism may affect leukemic outcomes in
other species. In addition, available mechanistic data are insufficient to characterize the
mechanisms or modes of action contributing to either spontaneously occurring or chemically
induced MCL in the F344 rat (Thomas et al.. 2007). including such tumors induced in
tetrachloroethylene-exposed animals. However, the albeit limited studies demonstrating that
tetrachloroethylene induces hemolysis and affects bone marrow function in mice provide indirect
evidence that tetrachloroethylene induces effects associated with MCL and with known
leukemogens (NRC. 2010). These studies support the biological plausibility of
tetrachloroethylene as a leukemogen in rodent species, in general, and provide a basis for
generating hypotheses on how these tumors may be induced. Nonetheless, the paucity of data on
contributing metabolites and mechanisms, and the lack of similar findings in other species,
contribute to uncertainty in interpreting the MCL data in the F344 rat (NRC. 2010).
Knowledge gaps persist regarding the causes of F344 rat MCL, the biology of the disease
including the cell type of origin, as well as the mechanisms by which tetrachloroethylene may
advance development of this rodent leukemia. Large granular lymphocyte (LGL) cells exist in
humans that are morphologically, biochemically, and functionally similar to the cells involved in
MCL in the F344 rat (Stromberg. 1985). In humans, clonal disorders of LGLs represent a
biologically heterogeneous spectrum of lymphoid malignancies thought as originating either
from mature T-cell or natural killer (NK) cells (Sokol and Loughran. 2006). LGL disorders can
clinically present as indolent (chronic) or aggressive diseases (Sokol and Loughran. 2006). The
indolent form of LGL leukemia is a disease of older adults, with a median age at diagnosis of 60
years. A number of clinical conditions have been seen in patients with LGL leukemia. These
include the following: red cell aplasia and aplastic anemia; other lymphoproliferative disorders
such as NHL, Hodgkin lymphoma, multiple myeloma, hairy cell leukemia, and B-cell
lymphoproliferative disorders; and autoimmune diseases such as rheumatoid arthritis and
systemic lupus erythematosus (Rose and Berliner. 2004). The etiology of LGL disorders is not
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-266 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
known (Rose and Berliner. 2004; Sokol and Loughran. 2006). Several possible etiologies have
been proposed including chronic activation of T-cell by a viral antigen or autoantigen in which
case LGL leukemia could be considered as an autoimmune disorder (Sokol and Loughran. 2006).
Lymphoid tumor pathobiology in rats and humans, its historical and current
classification, and epidemiology, including observations in tetrachloroethylene-exposed
populations, have bearing on examination of the human relevance of rat mononuclear cell
leukemia. Important to any examination are the changes in diagnostic and classification criteria
of human lymphoid tumors and lack of data on molecular markers in the tetrachloroethylene
epidemiologic studies, as discussed above. Diagnostic and classification criteria may not be
uniform across studies and hinder comparison of consistency within epidemiologic studies of
lymphoid cancers and tetrachloroethylene exposure and, also, between human and rat lymphoid
tumor observations. Furthermore, adoption of consensus nomenclatures of human lymphoid
tumors, i.e., the WHO scheme, for rats will facilitate cross-species comparisons, as was recently
conducted by the hematopathology subcommittee of the Mouse Models for Human Cancers
Consortium (Morse et al., 2002).
Further research to clarify the factors that affect inherent and chemically induced
susceptibility to F344 rat MCL is warranted, particularly given the morphological, functional,
and clinical similarities of this rodent leukemia to human T-cell leukemias. As proposed by
Stromberg (1985). the F344 rat MCL could serve as a rodent model for the human disease, in
which research could be conducted to identify causative factors and disease mechanisms, and to
test and develop novel chemotherapies. Thomas et al. (2007)similarly endorsed additional
research and analyses of F344 leukemogens, such as tetrachloroethylene, to advance
understanding of the mechanisms contributing to the rodent—and by inference, the related
human—diseases.
In summary, the available bioassay evidence and statistical analyses, together with an
albeit limited database of studies that characterize the biologic plausibility of tetrachloroethylene
as a leukemogen, provide sufficient support of the conclusion that tetrachloroethylene causes
MCL in the F344 rat. Supported, in part, by the similar characteristics of MCL in the F344 rat to
human LGL as described by Thomas et al. (2007) and others, and because no mechanistic or
other data are available that would rule out the relevance of the F344 MCL for assessing human
carcinogenic risk, this finding can be considered to provide evidence of a carcinogenic hazard of
tetrachloroethylene in humans. However, tumor site concordance across species is not always
assumed and may not necessarily be expected in the case of tetrachloroethylene (U.S. EPA,
2005). Tetrachloroethylene also induces other types of tumors within the F344 rat, notably brain
and male kidney tumors. In mice, it is a hepatocarcinogen, but has not been demonstrated to be a
leukemogen. Additionally, known human leukemogens (e.g., benzene, antineoplastic agents)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-267 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
induce a number of different tumors in rodents that may or may not include the particular
leukemia tumor type seen in humans. Several other factors also support the utility of the
bioassay data for estimating carcinogenic risk. In particular, this includes the demonstrated
statistically significant leukemic effects in both sexes in two bioassays, even when the statistical
analyses are subject to stringent statistical criteria (Thomas et al.. 2007). Accordingly, in the
absence of information to indicate that the observed positive effects in these studies are not
relevant to humans, the observation of MCLs in these studies are judged informative in the
weight of evidence for assessing carcinogenic hazard to humans and to estimate carcinogenic
risk of tetrachloroethylene.
4 7 DEVELOPMENTAL AND REPRODUCTIVE TOXICITY AND REPRODUCTIVE
CANCERS
4.7.1. Development
4.7.1.1. Human Developmental Toxicity Data
Epidemiology studies of tetrachloroethylene exposure and effects on reproduction and
development include occupational studies of employment at dry-cleaning establishments in the
Netherlands, Scandinavia, Italy, Canada, and the United States (California) and population-based
studies of exposure through drinking water in the United States (North Carolina, Massachusetts,
and New Jersey). Tetrachloroethylene has been the predominant solvent used in the dry-cleaning
industry in the United States and Europe since the 1970s (Gold et al.. 2008) (Raisanen et al..
2001). Other chemical exposures in dry-cleaning establishments are not widespread; individuals
engaged in spot cleaning may use small amounts of trichloroethylene, acetic acid, ketone, and
acetone solvents, petroleum naphthas, or hydrogen fluoride and hydrofluoric acid (Ruder et al..
2001). Short-term exposure to tetrachloroethylene is highest for dry-cleaning machine operators,
particularly for machines requiring manual transfer of solvent-saturated clothing from a washing
machine to a drying machine. The industry in the United States has gradually switched to dry-
to-dry machines, associated with lower emissions, and in 1993, EPA ruled that all new
establishments must use these machines. However, existing facilities were required to switch to
dry-to-dry machines only if the older machines became inoperable. Other workplace
characteristics influence exposure levels including adequacy of exhaust systems, level of
equipment maintenance, occurrence of tetrachloroethylene spills, and presence of open
containers (Gold et al.. 2008).
Studies of occupational exposure primarily evaluated employees in dry-cleaning
establishments, but a few studied reproductive and developmental outcomes by occupational
groupings more broadly (Windham et al.. 1991) (Lindbohm et al.. 1991; Taskinen et al.. 1989).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-268 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Although some studies identified exposed workers based on the industry they worked in, several
developed more precise classifications for tetrachloroethylene exposure levels based on detailed
information on reported job titles, tasks, and work histories obtained through interviews or
questionnaires. Exposure classification using more detailed information is expected to reduce
error in the assessment of exposure and increase confidence in the reported associations with
health outcomes.
Epidemiology studies also have evaluated reproductive and developmental health effects
stemming from incidents of tetrachloroethylene contamination of drinking water in the United
States (ATSDR. 1998b; Bove et al.. 1995; Lagakos et al.. 1986; Sonnenfeld et al.. 2001)
(Aschengrau et al.. 2008; Aschengrau et al.. 2009a; Aschengrau et al.. 2009b; Janulewicz et al..
2008). In general, drinking water exposures were to multiple pollutants, and most studies were
not able to determine the relative contribution to adverse health effects made by individual
substances. In one incident in Massachusetts, however, investigators were able to evaluate a
—natral experiment" that resulted from scattered water pipe replacements to the water
distribution system in communities and tetrachloroethylene-contaminated water delivered to
specific groups of households (Aschengrau et al.. 2008; Aschengrau et al.. 2009b; Janulewicz et
al.. 2008). The studies of exposure through drinking water are complicated by the occurrence of
other water pollutants, but this literature can provide information about the consistency of health
outcomes reported with those found in the occupational studies.
Studies of developmental effects evaluated low birth weight (Bosco et al.. 1987;
McDonald et al.. 1987; 01 sen et al.. 1990). intrauterine growth restriction (IUGR; also known as
small for gestation age [SGA]) (Bove et al.. 1995; Sonnenfeld et al.. 2001). birth defects
(Ahlborg. 1990b; Bosco et al.. 1987; McDonald et al.. 1987; 01 sen et al.. 1990). and stillbirth
(McDonald et al.. 1987; 01 sen et al.. 1990). A brief summary of each study follows, grouped by
health outcome, population (occupational, population-based), and exposure route (inhalation,
drinking water). Table 4-34 summarizes these studies. Two studies evaluated effects on
postnatal development including learning and behavior, and schizophrenia (Janulewicz et al..
2008; Perrin et al.. 2007). These studies are described in the section on neurotoxicological
effects (see Section 4.1). Studies of effects on immunological development and childhood
cancer are found in Section 4.6.
Overall, no associations were noted in several studies that assessed maternal or paternal
occupational exposure to tetrachloroethylene and increased incidence of stillbirths, congenital
anomalies, or decreased birth weight (Bosco et al.. 1987; Kyyronen et al.. 1989; Lindbohm.
1995; Olsen et al.. 1990; Taskinen et al.. 1989; Windham et al.. 1991). However, the number of
exposed cases for specific types of anomalies was not sufficient to evaluate risk with statistical
precision. When data for adverse birth utcomes identified in Sweden, Norway, and Denmark
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-269 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
were analyzed in relation to low or high tetrachloroethylene exposure among dry cleaners during
their pregnancies, odds ratios for congenital malformation, still birth, and low birthweight
(defined as <1,500 g) were 1.72 (95% CI: 0.40-7.12, 9 cases) for low exposure and 0.87 (95%
CI: 0.20-3.69, 3 cases) for high exposure (Olsen et al.. 1990). Kyyronen et al. (1989) reported
an odds ratio for all congenital malformations of 0.8 (95% CI: 0.2-3.5) among 24 cases and 93
controls. The sample size was not large enough to evaluate specific anomalies or conduct
multivariate analyses. A case-control study by Windham et al. (1991) identified one case of
IUGR with prenatal exposure to both tetrachloroethylene and trichloroethylene among their
sample of women with live births. The studies of occupational exposure also evaluated
associations with spontaneous abortion. More detailed descriptions of these studies and analyses
of spontaneous abortions are provided in Section 4.7.2. A study of parental occupational
exposure has also examined schizophrenia in offspring (Perrin et al.. 2007) and observed an
increased incidence in offspring of parents who worked in dry-cleaning establishments (RR: 3.4,
95% CI: 1.3-9.2), as discussed in Section 4.1.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-270 DRAFT-DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
Zielhuis et al., (19891 (letter to editor)
The Netherlands
Cross sectional study of menstrual
disorders among dry-cleaners and
laundry workers (471 of 592, 80%
response). Sampling frame was not
described. After excluding 72 because
of current pregnancy, lactation, chronic
illness, or gynecological surgery, and
125 exposed and 199 unexposed
because they used oral contraceptives,
final data set included 68 exposed and
76 unexposed
Questionnaire responses
Prevalence in referent group
(%)
Amenorrhea 0
Oligomenorrhea 10
Polymenorrhea 17
Irregular cycle 38
Unusual cycle length 30
Intermenstrual blood loss 17
Menorrhagia 22
Dysmenorrhea 29
Premenstrual syndrome 10
Employment in dry cleaning
compared to employment in
laundries
Linear logistic regression
Dry cleaning vs. laundry
OR (95% CI)
Oligomenorrhea 2.1 (0.9-5.3)
Polymenorrhea 0.8 (0.4-1.7)
Irregular cycle 1.2 (0.7-2.2)
Unusual cycle length
2.3 (1.2-4.4)
Intermenstrual blood loss
1.3 (0.6-2.7)
Menorrhagia 3.0 (1.6-5.6)
Dysmenorrhea 1.9 (1.1-3.5)
Premenstrual syndrome
3.6 (1.5-8.6)
Details concerning
study design and
analysis were not
provided.
to
~n
H
O
O
Eskenazi et al.,(1991b)
United States
Men in the dry-cleaning industry
compared to men working in laundries
recruited from membership lists of two
union locals in San Francisco Bay area
and Greater Los Angeles. Included all
dry cleaners (n = 85) and all laundry
workers 20-50 yr in Local 3 (n = 119)
and random selection of Local 52
(n = 206). Laundry workers were
frequency matched by age to dry
cleaners from same union local.
Eligible were 20-50 yr of age, current
workers, spoke English or Spanish, no
vasectomy and located by telephone or
mail. Participation: 20 exposed (38%
of 53 eligible) and 56 unexposed (34%
of 166 eligible
Semen quality
Semen samples obtained from
34 exposed and 48 unexposed
Brief physical exam by
physician blind to occupational
status to identify any medical
conditions that might affect
semen quality
Direct (expired air levels) and
indirect (index) measure of PCE
exposure
Exhaled air collected 16-19 h
after the end of a workwork
(except 11, which were corrected
to 16 hours using an elimination
model)
LOD: 2.67 (ig/m3, assuming 4 L
breath sample
Exposed: Workers at dry cleaners
or laundries where dry cleaning
was conducted on premises.
Unexposed: Workers at laundries
with no dry cleaning
Confirmed by industrial
hygienists
Analyzed associations with 17
measures of semen quality
Difference in means and
number with abnormal sperm
(<20 million sperm, >40%
abnormal forms, and < 60%
motile sperm)
Oligospermia (<20 million/mL)
approx 25% in both groups
Average percentage motile
sperm —torch fell within
normal limits" in both groups
Less than 60% motile
Exposed: 44%
Unexposed: 31%,/? = 0.23
Breath samples
reflect exposure in
the last week
Laundry workers
averaged less years
education and had
higher proportion
Hispanic (90 vs.
41%). Smoking and
alcohol use were
comparable.
Laundry workers
reported a higher #
days >80°F
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
Eskenazi et al.,(1991b) (continued)
In person interviews Work
history, including job tasks and
exposures in preceding week and
past 3 mo
Exposure score (0-11): estimate
of exposure during 3 mo period of
spermatogenesis
Exhaled PCE (mean, (ig/m3)
Exposed (n = 34)
7,892.9 (1.5-54,949.3)
Unexposed (n = 48)
76.9 (0.6-1,562.4)
Multiple linear regression (13
sperm measures) within 34
exposed, and all 82 men,
adjusted for several potential
confounders
No association within all 82
men for the 3 exposure
measures and clinical quality
measures: sperm concentration,
total count, percentage motility,
or percentage abnormal forms
Associations, adjusted for
confounding (p < 0.05) for
ALH, sperm linearity,
percentage round sperm; and #
narrow sperm and at least one
measure of exposure.
ALH and linearity
measure pattern of
sperm motion.
Authors stated
clinical
interpretation is not
yet —Illy
established"
Result do not
represent experience
of nonunion workers
(>85% of dry-
cleaning industry)
to
o
to
~n
H
O
O
Eskenazi et al.,(1991b)
United States
Wives of dry-cleaners and laundry
workers (extension of Eskenazi et al.,
(1991b)
17 of 20 dry cleaners with wives and
32 of 36 laundry workers with wives
participated
# with index pregnancies or trying to
conceive:
14 dry cleaners, 26 laundry workers
Reproductive outcomes:
* Rate of miscarriage: # of
miscarriages during
husband's employment in
industry/total # of
pregnancies during same
period
* Standardized fertility ratio
(SFR): ratio of O/E based on
U.S. national birth
probabilities for race, birth
cohort, parity, and age of
wives for each person-year
Dates of employment in the
industry and exposure to PCE
from interviews (index
pregnancies ended on average 2 yr
before interviews)
Exposure estimates:
* Expired PCE for husband
* Index of exposure
* Occupation: dry-cleaner vs.
laundry worker
SFR: Comparable between dry-
cleaners and laundry workers
Risk ratio: 1.01, 95% CI:
0.71-2.01
Time to conception (Cox
Proportional Hazard adjusted
for ethnicity and smoking):
Dry cleaners vs. Laundry:
Rate ratio = 0.54 (95% CI:
0.23-1.27)
# pregnancies and
live births similar
between dry-cleaners
and laundry workers
Power to detect
doubling of SA rate
from 12 to 24% was
0.28
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
Eskenazi et al., (1991a) (continued)
Calculated SFR for periods
when the men were
employed and not employed
in the industry
* Time to conception—self
report from wife—number
of months to become
pregnant with index
pregnancy
PCE in expired air was higher
among dry cleaners whose wives
were interviewed (10,245.6 vs.
7,892 ng/m3)
Rachootin and Olsen, (1983)
Denmark
Case-control study of couples
examined or treated for infertility at
Odense University Hospital, Denmark,
1977-1980. Controls selected from
couples with healthy child conceived
within 1 yr born at same hospital,
1977-1979. Eligible couples, residents
of the island of Funen, Denmark,
identified through hospital inpatient
register (1,069 infertile, 4,305 fertile).
Response 87% for both cases (n = 927)
and controls (n = 3,728)
Infertility
Data on reproductive
history, SES and behaviors
from questionnaire, medical
records of infertile couples
reviewed by collaborating
physician blind to
questionnaire responses
to
LtJ
Self-report by women through
mailed questionnaire sent Nov
1980-May 1981. Occupation
held in year prior to hospital
admission and longest held job.
Classified based on job title, type
of workplace and description of
duties. Coded using a 5-digit
Danish Occupational Code and a
5-digit industry code
Exposure defined as contact with
one of 15 specific chemical or
physical agents (included dry-
cleaning chemicals) or
performance of one of 3 work
processes a minimum of one time
per week for at least 1 yr
1. Cases infertile for at least 1
yr compared to controls, all
residing within catchment area
Dry-cleaning chemicals
OR (95% CI)
* Sperm abnormalities:
1.0 (0.5-2.0)
* Women with hormonal
disturbances
1.3 (0.5-3.3)
* Women with idiopathic
infertility
3.0 (1.2-7.4)
2.7 (1.0-7.1) adjusted for
women's age, education,
residence and parity.
*Men with idiopathic infertility
0.2 (0.0-1.4)
A higher percentage
of case couples lived
outside the hospital's
catchment area
Analyzed
associations with 15
chemical or physical
agents, 3 work
processes, noise and
heat
Number of controls
aged >20 yr: <20
Numbers of exposed
cases and controls in
dry cleaning was not
reported
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
Rachootin and Olsen, (19831
(continued)
2. Within control group
comparison; couples who gave
birth after 1 yr compared to
other controls
Delayed conception
Dry-cleaning chemicals
OR (95% CI)
Men 1.2 (0.7-1.9)
Women 1.6 (0.9-2.9)
Adjusted for women's age,
women's education, residence,
parity, women's smoking and
drinking, and past use of oral
contraceptives
to
~n
H
O
O
Sallmen et al. (1995)
Retrospective study, an extension to
Lindbohm et al. (1990)
Finland
Case-control study of women recruited
from Institute of Occupational Health
database of workers biologically
monitored for one or more of 6
solvents linked to national registry of
medically recognized pregnancies,
1965-1983 (n = 3,265)
235 of 355 women responded to
questionnaire (66%); after exclusions
final study population was 197 women
(median age 27, range 17-40 yr)
Time-to-pregnancy (TTP)
(number of menstrual cycles
required to become
pregnant) via self report in
questionnaire
Identified pregnancies from
nationwide database on
medically diagnosed
pregnancies, treated in
hospital from 1973-1983,
and from Finnish Register
of Congenital
Malformations. Used same
pregnancies as for
Lindbohm et al. (1990): SA
(n = 80) or live births
(n = 286), plus 30 referents
from malformation study
Same approach as Lindbohm
study (1990)
Exposure classification based on
self-reported work description and
solvent usage, and on biological
exposure measurements during
year before pregnancy, checked
by independent industrial
hygienist. Each work task
classified by likelihood and level
of exposure with no knowledge of
TTP
Not exposed—no handling of
solvents, not reported by worker
and no measurements.
Potentially exposed: work tasks
may have involved use of
solvents, no or undefined solvent
exposure reported and no
measurements
Analysis combined workers in
potential and low categories
Discrete proportional hazards
regression
IDR: ratio of average incidence
densities of clinically
recognized pregnancies for
exposed compared to
unexposed in each menstrual
cycle class
All solvents
Among women employed at
beginning of TTP (n = 152)
IDR (95% CI)
Not exposed 1.0
Low 0.74 (0.49-1.11)
High 0.44 (0.28-0.70)
Models adjusted for
age, alcohol,
smoking, partner's
smoking, coffee,
recent contraceptive
use, regular
menstruation, length
of menstrual cycle,
age at menarche,
previous induced
abortion or
extrauterine
pregnancy, previous
SA, parity, SA case,
unplanned
pregnancy,
frequency of
intercourse
Adjustment did not
change risk estimates
for organic solvents.
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
Sallmen et al. (19951 (continued)
Exposed: Measurements made
when holding same job and work
tasks implied solvent exposure or
solvent exposure was reported.
High: Handled solvents daily, or
1-4 d/wk and measurements
indicate clear exposure (n = 46)
Low: Handled solvents 1-4 d/wk,
no measurements or low levels, or
handled solvents <1 d/wk (n = 59)
None (n = 92)
PCE
Low (n = 13) 0.63 (0.34-1.17)
High (n = 7) 0.69 (0.31-1.52)
Worked in dry-cleaning shop
Low or high (n = 11)
0.44 (0.22-0.86)
High (n = 6)
0.57 (0.24-1.34)
Adjusted for low and high
exposure to solvents in other
industries, recent use of
IUD/spermicides, and age at
menarche
TTP info collected
8-18 yr after
pregnancy
to
~n
H
O
O
Sallmen et al. (1998). extension of
Taskinen et al. (1989)
Finland
Retrospective time-to-pregnancy study
of paternal exposure to organic
solvents. Wives of workers ever
monitored for organic solvents by
Finnish Institute of Occupational
Health, 1965-1983. Linked ids to
identify wives (n = 1,667) through
Finnish Population Register Centre and
pregnancies (n = 2,687) through
national database of medically
diagnosed pregnancies, treated in
hospital, 1973-1983. Included men in
their first marriage during 1985 with
wives aged 18-40 yr at the end of the
1st trimester of pregnancy.
Self reported by mothers:
Time-to-pregnancy (TTP)
Included pregnancies begun
during the marriage or up to 9
mo before
Only included pregnancies
identified in register and
reported by participants
Self-reported paternal exposure to
solvents at time attempt at
pregnancy began
Paternal exposure via mailed
questionnaires (January 1986) to
both spouses re: occupational
exposure related to study
pregnancy—employment,
occupation including work tasks,
and workplace during year of
conception
Use and frequency of any of the
monitored solvents and any other
materials
Biological measurements
available for 60% of men (during
TTP n = 33, same job but not
during TTP n = 161)
141/282 (50%) of men were
highly or frequently exposed to
organic solvents during TTP,
24.4% (n = 80) were low or
intermediate exposed
Discrete proportional hazards
regression
Paternal exposure to organic
solvents; adj FDR OR (95% CI)
Low/intermediate (n = 80)
0.74 (0.51-1.06)
High/frequent (n = 141)
0.80 (0.57-1.11)
Evaluated several
potential
confounders:
maternal age,
maternal and
paternal alcohol,
maternal and
paternal smoking,
maternal coffee,
recent contraceptive
use, irregular
menstruation,
duration of
menstrual cycle, age
at menarche,
previous induced
abortion or
extrauterine
pregnancy, previous
SA, parity, year of
pregnancy, SA case,
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
to
<1
On
~n
H
O
O
Sallmen et al. (19981. extension of
Taskinen et al. (19891 (continued)
Restricted to cases (n = 110) and
controls (n = 332) who participated in
study on pregnancy outcome.
Excluded 1 case and 3 controls.
316 (72%) of wives participated. After
exclusions (n = 21) and inability to
give TTP (n= 13), final population was
282 couples
Exposure assessment for 80
calendar days preceding study
pregnancy (spermatogenesis)
blind to outcome status. Based on
occupation, job description,
reported solvent or other chemical
use, and biological monitoring
data. New assessment for TTP
needed for 9 men whose job tasks
had changed since last study
Not exposed: Work tasks did not
include handling solvents, worker
did not report exposure and no
biological measurement
Potentially exposed: Work tasks
might have involved solvent use,
but not reported by worker, no
biological measurements
Exposed: Biological measurement
taken while at same job, or tasks
implied solvent exposure, or
solvent exposure reported
Level of Exposure
High: handled solvents daily or
level of biological measurements
above reference value for general
population
Intermediate: Solvent use
1-4 d/wk and biological
measurements indicate
intermediate or low exposure
Low: Handled solvents <1 d/wk
None: all other
Adjusted for paternal and
maternal smoking, maternal
age, age at menarche >15,
duration of menstrual cycle,
frequency of intercourse,
maternal exposure to organic
solvents, year of pregnancy and
variable for missing
information
Paternal exposure to PCE; adj
FDR; OR (95% CI)
Low (n = 9)
0.86 (0.40-1.84)
Intermediate/High (n = 8)
0.68 (0.30-1.53)
Adjusted for short menstrual
cycle, long or irregular
menstrual cycle, older age at
menarche, frequency of
intercourse, maternal age,
maternal exposure to organic
solvents, and variable for
missing information
unplanned
pregnancy,
frequency of
intercourse, maternal
exposure to organic
solvents
Recall: Data
collection on TTP
occurred 8-18 yr
after pregnancy
Participation: Lower
among women with
>2 previous births
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
to
<1
^1
McDonald et al. (1986: 19871
Canada
Hospital-based survey of maternity
departments, 1982-1984. 56,012
women interviewed in 11 obstetrical
units in Montreal (90% of births);
51,885 with term delivery (90%
interviewed) and 4,127 SA (75% of
those admitted)
Treatment in hospital of SA
(4,127 women) plus self report
of previous SA (before Week
28 of pregnancy) (10,910
pregnancies)
Stillbirth without defect:
fetal deaths after the 27th wk
of gestation
Congenital defects:
Information extracted from
medical records at time of
discharge, previous births
obtained from mothers
report at interview and later
review of physician or
hospital records
LBW <2,500 g
Self-reported occupation at time
of conception for current and
previous pregnancies
2nd analysis defined employment
for >30 h/wk at beginning of
pregnancy
Expected numbers calculated for
each occupational category from
effect of individual factors on
probability of SA using logistic
regression: maternal age, parity,
history of previous abortion,
smoking habit, and education
Laundry and dry cleaning:
# current pregnancies: 100
# SA: 8
O/E: 1.18
# previous pregnancies: 123
# SA: 31
O/E: 1.02
2nd analysis combined current
and previous pregnancies:
# pregnancies: 202
# SA: 36
O/E: 1.05
2nd analysis used maternal age,
gravidity, previous spontaneous
abortion, smoking, alcohol,
education, and ethnicity
Stillbirth (n = 3) O/E: 1.86
Congenital defects (n = 9) O/E:
1.41
LBW (« = 15) O/E: 1.17
p-y alue >0.05
Potential bias:
* interviewers were
informed of outcome
status
* recall time to first
wk of pregnancy
different for women
with SA vs. term
birth
Dry-cleaning and
laundry workers
likely included many
not exposed to PCE
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
to
<1
00
~n
H
o
o
Taskinen et al. (19891
Finland
Case-referent study
Workers ever monitored for organic
solvents by Finnish Institute of
Occupational Health, 1965-1983.
Linked IDs to identify wives through
Finnish Population Register Centre and
pregnancy outcomes through national
registers. Included men in their first
marriage during 1985 with wives aged
18-40 yr at the end of the 1st trimester
of pregnancy. Included pregnancies
begun during the marriage or up to 9
mo before
Cases defined as wives with SA (if
multiple, one randomly selected) or
congenitally malformed child.
Referents selected from wives with
healthy birth 1973-1983 (1:3 for SA,
1:5 malformations), age matched
within 30 mo
Only included pregnancies identified in
register and reported by participants
Response rate of SA: cases 136 of 172,
79.1%; referents 370 of 505, 73.3%
Final data set including eligible
pregnancies for SA case-referent sets:
120 cases and 251 referents
Medically diagnosed
pregnancies from Hospital
Discharge Register
(National Board of Health)
or data on SA treated in
hospital polyclinics,
1973-1983
Congenital malformations
recorded in Finnish Register
of Congenital
Malformations
SA rate among all
recognized pregnancies in
the cohort (including
induced abortions) 8.8%
Paternal exposure via mailed
questionnaires (January 1986) to
both spouses re: occupational
exposure related to study
pregnancy—employment,
occupation including work tasks,
and workplace during year of
conception
Use and frequency of any of the
monitored solvents and any other
materials
Exposure assessment for 80
calendar days preceding study
pregnancy (spermatogenesis)
blind to outcome status. Based on
occupation, job description,
reported solvent or other chemical
use, and biological monitoring
data
Not exposed: Work tasks did not
included handling solvents,
worker did not report exposure
and no biological measurement
Potentially exposed: Work tasks
might have involved solvent use,
but not reported by worker, no
biological measurements
Exposed: Biological measurement
taken while at same job, or tasks
implied solvent exposure, or
solvent exposure reported
Categorized into none, low, or
high
Conditional logistic regression
OR for likely exposure to PCE
only presented for unadjusted
model (controlling for potential
exposure to PCE)
OR (95% CI)
Likely exposed: 4 cases, 17
referents
0.5 (0.2-1.5)
Trichloroethylene
Likely exposed 17 cases, 35
referents
1.0 (0.6-2.0)
Potential
misclassification of
exposure but
nondifferential:
Among men with no
monitoring data,
21.5% of cases and
24.2% of referents
reported exposure to
solvents and were
categorized as
exposure likely
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
Taskinen et al. (19891 (continued)
High: handled solvents daily or
level of biological measurements
above reference value for general
population
Intermediate: Solvent use 1-4
d/wk and biological
measurements indicate
intermediate or low exposure
Low: Handled solvents <1 d/wk
None: all other
to
vo
Lindbohm et al. (19911
Finland
All pregnancies and outcomes recorded
in nationwide Hospital Discharge
Register and data requested from
outpatient hospital clinics, 1973-1982.
Pregnancies 1973-1978 linked to 1975
Census and 1979-1982 to the 1980
Census. Central Statistical Office of
Finland (1975 and 1980) Census data
used for occupation and industry, SES
For exposure to any
mutagenic agents, evaluated
pregnancies terminated in
1976 for exposure in 1975
(to approximate 80 d prior
to conception) and May 1,
1980-April 20, 1981 for
1980 Census
99,186 pregnancies among
women, 12-49 yr old, with
information on occupation,
industry and woman's SES.
For exposure to specific
agents, included a 2-yr
period close to the census
(Jan 1, 1976-Dec 31, 1977
and May 1, 1980-April 30,
1982)
Paternal exposure classified using
job-exposure matrix developed in
cooperation with 2 industrial
hygienists. Based on occupation
and industry. Assign prevalence
of chemical exposure to job
groups based on monitoring data
from Institute of Occupational
Health
Classified into 3 levels for
exposure to mutagens:
Moderate/high: 139
Potential/low: 820
None: 7,772
Prevalence of SA: 8.8%
(Similar to national rate in
Finland: 8.9%)
PCE: 3 SA and 45 pregnancies
defined as moderate/high
exposure
Linear logistic regression
controlling for age only
OR (95% CI)
0.7 (0.2-2.4)
Focus of exposure
assessment was on
mutagens"
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
to
00
o
Lindbohm et al. (19901
Finland
Case-control study of women recruited
from Institute of Occupational Health
database of women biologically
monitored for one or more of 6
solvents linked to national registry of
medically recognized pregnancies;
80 cases (78.4% of 102 respondents)
and 286 controls (99.3% of 288) (age
matched 1:3) confirmed pregnancy of
interest
73 cases and 167 controls with
complete information for both cases
and controls
Cases were women with a
spontaneous abortion recorded
in the national register of
pregnancies in Finland and the
Finnish Register of Congenital
Malformations that was
confirmed by the women
Self-report of employment,
occupation, workplace and
exposure to solvents during first
trimester by mailed questionnaire
Exposure assigned by 2
investigators blind to outcome
status using responses and
biological measurements when
available
Not exposed: Work tasks did not
included handling solvents,
worker did not report exposure
and no biological measurement
Potentially exposed: Work tasks
might have involved solvent use,
but not reported by worker, no
biological measurements
Exposed: Biological measurement
taken while at same job, or tasks
implied solvent exposure, or
solvent exposure reported
Level of Exposure:
High: handled solvents daily or
1-4 d/wk and level of biological
or available industrial hygiene
measurements were high
Low: Handled solvents 1-4 d/wk,
and level of exposure low, or
handled solvents <1 d/wk
None: all other
Conditional logistic regression
controlling for previous SA,
parity, smoking, use of alcohol,
and exposure to other solvents
OR (95% CI)
All solvents
2.2 (1.2-4.1)
PCE (8/15 exposed
cases/controls)
Overall 1.4 (0.5-4.2)
Low 0.5 (0.1-2.9)
High 2.5 (0.6-10.5)
Use of PCE in dry cleaning
(4 cases/5controls)
2.7 (0.7-11.2)
Other dry-cleaning work
(1/6)
0.6 (0.1-5.5)
Biological
measurements were
available for only
5% of sample
Blood PCE (mean)
at time nearest
pregnancy
Dry cleaners (n = 6)
2.11 (imol/liter
Other workers
(» = 7)
0.43 (imol/liter
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
to
00
Windham et al. (19911
United States
Hospital-based case-control study
697 women ± 18 yr, June 1986-Feb
1987 (81.8% of 852)
1,359 controls (2 per case) randomly
selected from among residents of Santa
Clara County, California with a live
birth, frequency matched by last
menstrual period (± 1 wk) and hospital
(84% of 1,485)
Analysis limited to 1,361 women who
were employed during pregnancy
(70%)
Medically diagnosed SA
defined as 20 wk gestation
with pathology specimen
submitted to one of 11
hospital laboratories in
Santa Clara County,
California; verified by
review of medical charts
Computer-assisted telephone
interview—exposure during
pregnancy (cases) or first 20 wk
(controls)
Asked whether they used or
worked around any of 10 solvents
(including PCE) once per week or
more, plus asked to name any
other solvents or degreasers. For
each product, number hours per
week, weeks of exposure, skin
contact, smelled odors, or
experience symptoms
Unexposed referent did not use
any of the named solvents
(n = 847)
Exposure metric: average hours
used/week of pregnancy
249 of 1,361 working women
were exposed to solvents
5 PCE exposed cases, 2
exposed controls
9 TCE exposed cases, 15
exposed controls
Crude OR (95% CI)
PCE
4.7(1.1-21.1)
TCE 3.1 (0.55-2.9)
Paint Thinners
2.3 (1.0-5.1)
Paint Strippers
2.1 (0.64-6.9)
PCE and/or TCE
3.4 (1.0-12.0)
Adjusted OR
PCE adj for hours worked 4.2
(0.86-20.2)
PCE adj for age
6.0 (1.4-25.8)
Intensity: respondents reported
skin contact, odor, or symptoms
(headaches, dizziness, or
forgetfulness)
Yes: ORc: 6.3, p = 0.04
None: ORc: 2.1,p = 0.54
Adjustment for
confounders:
Mantel-Haenszel
stratification of
dichotomized
covariates one at a
time: maternal age,
race, education, prior
fetal loss, smoking,
and hours worked
Cases and controls
worked similar hours
and schedules
4 of 7 women
reporting use of PCE
also used TCE
Adjustment did not
alter OR for other
solvents (TCE,
thinners and
strippers)
No consistent trend
by # hours used per
week
~n
H
O
O
-------
a
o
<
s?
rs
>!
o
^3
§•
rs
st
3
>!
sT
St
o
>1
>!
Oq
o
>!
Si
Si
s
O
Sfc
I
§•
K) On
2
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
Bosco et al. HQS?)
Italy
67 women working in 53 of 66 dry-
cleaning shops in 2 neighborhoods in
Rome, Italy (40 dry cleaners and
ironing, 13 ironing service only)
Average age 43 yr employed on
average 20 yr
Self report by standardized
Interview
SA not defined
Self-report by standardized
interview
LBW <2,500 g/live birth
Birth defects/live births
Still births/live births
Self report by standardized
Interview—work activity prior to
and during pregnancy (dry
cleaning, housewife, other)
Presence of trichloroacetic acid in
24-h urine (53 of 67) Mean (|ig/L)
Dry Cleaners 5.01* (n = 40)
Ironers only 1.35 (n = 13)
Controls 1.56 (« = 5)
*p = 0.06 compared dry cleaners
with ironers and controls
combined
5 SA of 56 pregnancies
reported while employed as dry
cleaner (8.9%)
1 SA of 46 pregnancies
reported while house-wife
Fourfold greater risk,
standardized for age, not
statistically significant
Dry cleaners 51 live births
Housewives 44 live births
n (%)
LBW Dry CI Hsewvs
2 (3.9) 9 (6.8)
Birth Defects/LB
2 (3.9) 1 (2.3)
Still births/LB
0 (0) 1 (2.3)
Ascertainment of
exposure and
outcome was not
independent
Asked about
pregnancies
occurring 1>20 yr
previous
to
00
to
~n
H
o
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
to
00
LtJ
Olsen et al. (19901
Scandanavia (Sweden, Finland, and
Denmark)
Nested case-control studies combining
country-specific odds ratios,
1973-1983
Sweden and Denmark: All women
selected from company records of all
active dry-cleaning plants and
laundries (dry cleaners only in
Denmark) working for >1 mo during
1973-1983.
Finland: Dry-cleaning and laundry
workers identified from union registers
and payroll data requested from all
facilities in country, 1973-1983 (%
response not provided). Asked for
names of women employed for at least
3 mo during 1973-1983
Sweden: 169 women with a registered
pregnancy who worked at laundry or
dry cleaner during all or part of the
year before delivery or 6 mo before
SA, 2 matched controls per case; 84%
respondents of 201 contacted. 61.7% of
identified plants participated
Finland: 720 pregnancies (1 randomly
selected per woman) reported in
hospital discharge register and reported
by woman, 3 age-matched controls per
case; 77.2% respondents of 932
contacted
Medically recognized SA
recorded in centralized birth
registries and linked to
participants
Sweden: Central medical birth
register (« = 31)
Finland: Nationwide hospital
discharge registry and
polyclinic data on SA
(n= 118)
Denmark: Central birth register
and standard hospital register
(,n = 10)
Low birth wt: <1,500 g
(Sweden (n = 5), Norway
(n = 7) and Denmark (n = 1)
Congenital malformations
(excluding certain minor
malformations)
Sweden: n = 6
Norway: n = 1
Denmark: n = 1
Finland: n = 24
Perinatal death (Sweden and
Norway)
Sweden: n = 5
Exposure during 1st trimester; Self
report from questionnaires or
interview
Sweden and Denmark:
classification by industrial
hygienist blind to pregnancy
outcome
Finland: classification by study
investigators based on work
history and exposure frequency
Classification:
Unexposed—No exposure to PCE
as defined
Low—worked in dry-cleaning
facility but not high exposure.
High—Conducting dry cleaning
or spot removal >1 h/d
Spontaneous abortion
OR (95% CI)
Combined (weighting by
inverse variance of OR):
Low 1.17(0.74-1.85)
High 2.88 (0.98-8.44)
Sweden:
Low 1.15 (0.43-3.09)
High 0.82 (0.07-9.86)
Denmark:
Low 0.00
High 2.52 (0.26-24.1)
Finland:
Low 1.18(0.71-1.97)
High 4.53 (1.11-18.5)
Combined outcomes (LBW,
malformations and perinatal
death), All countries combined,
all trimesters (combined
variance calculated using
inverse variance of the OR) OR
(95% CI)
Low 1.72 (0.4-7.12)
High 0.87 (0.2-3.69)
In Sweden and
Denmark only 1
exposed case in high
exposure group, in
Finland 6 exposed
cases in high
category
Models adjusted for
parity, smoking and
drinking habits
(Danish model only
for parity and
smoking)
Analyses using
exposure
information from
employers (55% of
sample) stated to
have similar results
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
Olsen et al. (19901 (continued)
Denmark: 143 registered pregnancies
of all women employed at least 1 mo at
listed registered dry cleaners,
1979-1984, 77.3% respondents of 185
in cohort. 74.3% of identified plants
participated
to
oo
Kyyronen et al. (19891
(Also reported in Olsen et al. (1990V)
Finland
679 women confirmed the pregnancy
contained in the register and provided
exposure information for the 1st
trimester; 25.9% of SA cases did not
report the pregnancy in the register and
were not included along with matched
controls
130 SA reported
289 controls (women with
healthy pregnancy and no
SA during study period),
matched by age within ±2 yr
24 cases of congenital
malformation
93 controls
Unexposed—No exposure to PCE
as defined
Low—work tasks included
pressing at a dry cleaners' or spot
removing, or reported handling
PCE less than once per week
High—work tasks included dry
cleaning for at least 1 h daily on
average, or reported handling PCE
at least once per week
Spontaneous abortion:
Multivariate logistic regression
model:
High—3.4 (1-11.2)p < 0.05
Low exposure was not included
in multivariate model:
unadjusted OR: 0.7 (95% CI
not reported)
Model adjusted for frequent use
of solvents other then PCE,
frequent heavy lifting at work,
frequent use of alcohol
Congenital malformation:
Univariate, matched logistic
regression
PCE (any level) 1st trimester
OR (95% CI)
0.8 (0.2-3.5) 2 exposed cases
6 cases and 6
controls reported
exposure to other
solvents: petroleum
benzene, toluene,
acetone, thinner, and
spot remover
mixtures
Other covariates
(including smoking,
temperature, parity,
febrile disease) were
not associated in
univariate models so
not included
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
to
00
~n
H
o
o
Ahlborg, (1990b) (Complementary
study to Olsen et al. (19901')
Sweden
Case-referent study: Two cohorts of
women working for >1 mo during
1973-1983 in dry-cleaning or laundry
work
Primary: 2,181 eligible women
selected from company records of 475
active dry-cleaning plants and
laundries, 263 used PCE and had
women as employees
Linked to Medical Birth Registry and
Inpatient Registry for Somatic Care;
identified 2,438 births and 143 SA
955 pregnancies including 66 cases of
SA, perinatal death, congenital
malformation, or low birth weight
involved employment (at least one
week) during year before delivery or 6
mo before SA. Referents matched to
cases (1:2) by mother's age (± 2 yr),
year of pregnancy, and parity (for
deliveries only)
Responses for 158 pregnancies (48
cases (75%, 110 referents (88%)
Complementary: 5,176 female laundry
and dry-cleaning workers registered as
washers/cleaners in the national census
of 1975 and 1980; linked with medical
registers for 2-yr period following each
census—1,136 pregnancies identified
Pregnancies and
hospitalized SA identified
through national registries
occurring 1974-1983
Identified 2,438 births
Cases defined as
spontaneous abortion,
perinatal death, congenital
malformation, or low birth
weight
Exposure during 1st trimester; Self
reports through mailed
questionnaires; questions re: type
of production (laundry only,
laundry and dry cleaning, or dry
cleaning only), use of specific
agents in dry-cleaning process
(including PCE)
Information obtained from
employers on type of production,
amount of dry cleaning, and use
of specific cleaning agents during
1973-1983, and dates use of PCE
started and ended
Use of PCE:
22 of 48 cases said -don't know,"
19 categorized as exposed by
employer
41 of 110 referents said —dn't
know," 30 categorized as exposed
by employer
Exposure classified by 2
investigators blind to case/referent
status
High: Operating dry-cleaning
machines or spot removing with
PCE ± 2 h/wk, or ironing/pressing
dry cleaned cloth >20 h/wk, or
cleaning and filling the machines
>3 times
Low: Other work at workplaces
where dry cleaning with PCE was
performed
Multivariate conditional logistic
regression model
Primary study:
Dry cleaning (Y/N)
Referents did not work in dry
cleaning or were not working
during 1st trimester
All outcomes combined:
OR (95% CI): 1.1 (0.6-2.0)
Self-report
1.02 (0.47-2.2) Alborg, 1990b)
Employer
1.27 (0.60-2.71)
Use of PCE (Y/N)
OR (95% CI):
Self-report
0.92 (0.36-2.33)
Employer
0.82 (0.32-2.07)
Adding response from
employer to data self-reported
as —dn't know":
1.24 (0.59-2.61)
Highly exposed pregnancies
Primary study: 10 of 55 cases,
27 of 106 referents
Complementary: 9 of 67 cases,
17 of 126 referents
For SA only:
Low 1.0 (0.4-2.2)
High 0.9 (0.4-2.1)
Few highly exposed
pregnancies, few
cases
Validity of self-
reports:
Questionnaire data
compared to
employers response:.
Dry cleaning Y/N:
sensitivity among
cases 0.97 and
controls 0.96
Specificity among
cases 0.75 and
controls 0.69
PCE Use Y/N:
Sensitivity among
cases: 1.0 and
controls: 0.93;
Specificity among
cases: 1.0 and
controls: 0.94
Large plants
participated in the
primary study (dry
cleaning accounted
for <10% of total
production)—air
concentrations likely
to be lower than for
smaller plants
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
Ahlborg, (1990b) (Complementary
study to Olsen et al. (19901')
(continued)
755 pregnancies not found in primary
study, including 55 SA and 28 other
adverse outcomes, response for 68 of
77 cases (88%) and 131 of 150
referents (87%)
Unexposed: Dry cleaning with
PCE was not performed at
workplace
Models adjusted for
smoking, alcohol
consumption,
medical
complications, and
history of adverse
pregnancy outcome
to
oo
On
~n
H
O
O
Doyle et al. (1997)
United Kingdom
7,305 women, 16-45 yr, currently or
previously employed in dry cleaning or
laundry units managed by 4 companies
in the UK, 1980-1995
54.5% of 5,712 questionnaires
successfully delivered were returned
completed
Response rate for current dry-cleaning
and laundry workers: 78 and 65%
Previous workers 46 and 40%
Self report via mailed
questionnaire, self reports
verified with general
practitioner (all women
(114) reporting SA who
worked during pregnancy
and random sample of 58
who reported not working,
comparison for 59).
Distribution of reported
exposures during pregnancy
was similar for validated vs.
not validated;
SA defined as any fetal loss
before 28 wk gestation in a
confirmed pregnancy
Self report via mailed
questionnaire;
For each pregnancy: Work in dry
cleaning or laundry during
pregnancy or 3 mo prior to
conception
Exposure defined as machine
operator during pregnancy or 3
mo prior to conception,
unexposed as nonoperator
Unit of analysis: pregnancy
SA rate: # reported SA/#
liveborn, SAs, and stillbirths
408 pregnancies among
operators
# SA:
Operator: 65
Nonoperator: 29
Laundry: 18
Dry cleaning vs. laundry
Pregnancy completed
1980-1995:
Adjusted OR (95% CI):
0.97 (0.55-1.69)
Operator vs. Nonoperator:
1.63 (1.01-2.66)
Compared to unexposed
pregnancies before 1st exposed
pregnancy:
Laundry: 1.49 (0.87-2.58)
Nonoperators: 1.02 (0.65-1.6)
Operators: 1.67(1.17-2.36)
Models adjusted for
maternal age,
pregnancy order, and
year of event
Separate analyses
also restricted to 1st
and last pregnancies
Were dry cleaners
more likely to report
fetal death or ectopic
pregnancy? No.
Current workers: dry
cleaners vs. laundry
11 vs. 12.9%;
Previous workers:
13.9 vs. 14%
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
Perrin et al. (20071
Israel
Jerusalem Perinatal Study, a
longitudinal study
Examined risk for schizophrenia in a
prospective population-based cohort of
88,829 offspring born in Jerusalem,
1964-1976, followed from birth to age
21-33 yr (January 1, 1998). Included
all births to mothers in a defined
geographic area and linked to Israel's
national Psychiatric Registry
88,060 with complete information
The Psychiatric Registry
contains diagnoses from
multiple sources, including
inpatient wards in
psychiatric and general
hospitals and psychiatric
day-care facilities.
Definition of schizophrenia-
related discharge diagnostic
codes F20-F29. Date of
onset—first psychiatric
admission
4 offspring of parent dry
cleaners with schizophrenia
(2 male, 2 female); 3 cases
had exposed fathers
Occupation and demographic
information from birth certificate
Dry cleaning = 1 if mother or
father occupation listed on birth
certificate, otherwise 0
144 offspring with one or both
parents a dry cleaner (63 female,
81 male)
Time to schizophrenia using
proportional hazards methods
Evaluated potential
confounders: parents' age,
father's social class, duration of
marriage, rural residence,
religion, ethnic origin, parental
immigration status, offspring's
birth order sex, birth weight and
month of birth. Variables
included if changed risk
estimate by >10%. Results
presented as crude because
confounding was minimal
637 diagnosed with
schizophrenia-related
diagnosis; cumulative incidence
= 1%
RR: 3.4 (95% CI: 1.3-9.2)
Models did not
adjust for family
history of mental
illness
to
oo
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
Drinking Water
a -r
>1
to 0\
2 ^
3
to
00
00
Lagakos et al. (19861
United States
Retrospective population-based study
of adverse pregnancy outcomes and
childhood disorders in Woburn,
Massachusetts in relation to drinking
water from two municipal wells
contaminated with chlorinated
organics, 1960-1982. 7,134 of 8,109
possible interviews were completed
(80%). 6,219 distinct residences were
reached and 5,010 interviews were
completed (57% of the towns'
residences with listed telephone
numbers)
235 volunteer interviewers
(approx half were Woburn
residents) conducted a
telephone sample survey of
current and former family
members living in Woburn
household between
1960-1982 using telephone
numbers from the 1982
directory. Interviews were
anonymous and residence
address was not identifiable
For any residents prior to
1979, self-reports on all
pregnancies ending between
1960 and 1982 for women
born since 1920
SA: loss in the first 6 mo of
pregnancy
Perinatal death: Stillbirth or
livebirth surviving fewer
than 7 d
Low birth weight (LBW): 6
lbs (2,722 g)
Exposure estimates for water from
Wells G and H using information
on space—time distribution.
Residence history obtained from
1982 telephone directory and self-
reported residence history.
2 of 8 municipal wells (Wells G
and H in eastern Woburn) were
tested in May 1979 and found to
contain volatile organics and the
wells were shut down.
TCE 267 ppb
PCE 21 ppb
Chloroform 12 ppb
Trichlorotrifluoroethane 23 ppb
Dichloroethylene 28 ppb
Groundwater sampling in 1979,
61 test wells identified
48 EP. I priority pollutants and 22
metals
MA Dept Environmental Quality
and Engineering estimated
regional temporal distribution of
water from Wells G and H during
October 1964-May 1979 using a
model of the Woburn water
distribution system creating 5
zones of graduated exposure
before and after 1970.
Maximum likelihood logistic
regression model adjusting for
maternal age, smoking status
during pregnancy, year of
pregnancy, SES, sex, and
mother's pregnancy history
4,396 pregnancies, 1960-1982
16% were exposed during year
the pregnancy ended
SA: 12% (n = 520)
Perinatal death: 1.5% (46
stillbirths and 21 deaths before
7 d)
LBW among live births >7 d:
6.4% (220/3,462)
Congenital anomalies: 4.6% (n
= 177)
Adjusted OR not presented
SA (p = 0.66)
LBW (p = 0.77)
Perinatal deaths before 1970 (p
= 0.55)
After 1970: OR (/?-value)
10 (0.003) (Based on 3 deaths
out of 88 births in highest
exposure quartile, 1970-1982
Rates of adverse
health effects in East
and West Woburn
among unexposed
(during years when
Well G and H were
not operating) were
not statistically
significantly
different
Authors explored
differences between
East and West
Woburn for possible
selection bias, and
completed calls and
refusals. Checked
accuracy of
interviewers
(recontacting) and
respondents (verified
with medical
records)
Did not ask about
perception of
exposure to Wells G
and H in survey
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
Lagakos et al. (19861 (continued)
Medically diagnosed
congenital anomalies
grouped by involved organ
or system (ICD):
musculoskeletal,
cardiovascular, and eye/ear
defects. Grouped other
organs/systems with few
cases into a group with
potential environmental
links (CNS, chromosomal,
and oral cleft anomalies)
and —dter." Grouped prior
to exposure evaluation
Childhood disorders
grouped into 9 categories
with >20 cases
These data used to estimate the
percentage of annual water supply
from Wells G and H at each
household
Calculated an annual exposure
score corresponding to the
mother's residence in the year the
pregnancy ended
For each child: sum of annual
exposure scores during residence
history in Woburn
Anomalies:
Musculoskeletal (p = 0.78)
Cardiovascular (p = 0.91)
Eye/Ear OR (/j-valuc)
14.9 (0.0001)
CNS/chromosomal/oral cleft
OR (/?-value) 4.5 (0.01)
Other (p = 0.62)
Childhood disorders:
Observed vs. expected
cumulative Wells G and H
exposure by disorder
Kidney/urinary tract (p = 0.02)
Lung/respiratory disorders
(p = 0.05)
Study could not
associate effects with
specific
contaminants
to
oo
VO
~n
H
O
O
Bove et al. (1995)
United States
Cross-sectional study of birth outcomes
and fetal deaths in relation to total
trihalomethanes (TTHM) and
chlorinated organics in public water
supplies in a 4-county area in northern
New Jersey, 1985-1988. 80,938
singleton live births and 594 singleton
fetal deaths (after excluding plural
births, therapeutic abortions and
chromosomal anomalies) from 75 out
of 146 towns primarily served by
public water systems
Live births and fetal deaths
(plus birth weights and
gestational age) identified
through birth or death
certificates occurring during
1/1/85-12/31/88
LBW <2,500 g among term
births (>37 wk)
SGA: live births below
race-, sex-, and gestational
week-specific 5th percentile
weight using NJ data for
1985-1988
Estimated monthly levels of
individual contaminants in each of
75 towns using tap water sample
data collected by the New Jersey
Dept. of Environmental Protection
and Energy and the water
companies. At least 2 samples per
year. Monthly estimates were
assigned to each gestational
month for each live birth and fetal
death. Estimated independently
of birth outcome data
Linear regression for birth
weight, Logistic regression for
categorical outcomes
Adjusted for maternal age,
maternal race, maternal
education, primipara, previous
stillbirth or miscarriage, sex of
the birth, adequacy of prenatal
care. PCE model also adjusted
for TTHM
Results reported with nested CI
(50, 90, and 99%)
During study period,
birth and death
certificates did not
record maternal
occupation,
smoking, and
alcohol consumption
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
to
VO
o
Bove et al. (19951 (continued)
Preterm birth (<37 wk)
Very low birth wt <1,500 g
Birth weight among —etrm
births" (>37 wk and <42
wk)
Birth defects ascertained
using NJ Birth Defects
Registry—a population-
based, passive system—plus
fetal death certificates (>20
wk)
Comparison group
(n = 52,334): all live births
from study population that
were not low birth weight,
SGA, or preterm, and with
no birth defects
Birth defects and fetal deaths in
relation to average exposure
during 1st trimester
PCE
Average 1st trimester: 26 ppb
Average entire pregnancy: 14 ppb
55.6% of study population with
surface water as source of
drinking water, 11.6% had a
mixture of surface and ground
water
82% of comparison group had
PCE concentration in public water
supply <1 ppb, 11.5% >1-5 ppb,
5.1% >5-10 ppb and 1.4%
>10 ppb
Adjusted mean decrease in birth
weight among term births: 27.2
g (50% CI: -13.4- -41.0) for
PCE >10 ppb
No association with fetal
deaths, LBW, SGA, or preterm
birth
Very LBW: OR, 50% CI: 1.49,
1.13-1.97
All surveillance birth defects:
OR (50% CI): 1.14, >10 ppb
CNS defects: no association
Neural tube defects: PCE >5
ppb: 1.16 (0.69-1.83),
association disappeared when
TTHM included in model
Oral cleft defects: PCE
#
<1 67
>1-5 11
>5-10 1
>10 4
OR 50% CI
ref
1.17 0.89-1.53
0.24 0.05-0.63
3.54 2.12-5.57
No monotonic trend
Major cardiac defects: PCE
>5 ppm: OR: 1.13
Information on these
risk factors was
obtained for a small
number of mothers
by phone interview.
For these women,
adjustment for these
risk factors did not
change the
contaminant specific
ORsby >15%
Authors noted that
nondifferential
misclassification
could result in
underestimate or
overestimates of the
true effect for middle
exposure categories
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
to
VO
Sonnenfeld et al. (20011
United States
Retrospective study of birth outcome
among singleton liveborn and stillborn
infants of >20 wk gestation, and
exposure to volatile organic
compounds in drinking water at the
U.S. Marine Corps Base at Camp
Lejeune, North Carolina, 1968-1985.
Included births to mothers living in
base family housing at delivery and for
at least 1 wk prior. Excluded 2 groups
of residents exposed to TCE through a
different water system and residents in
trailer parks because housing records
were incomplete
Outcome data obtained from
birth and fetal death
certificates:
* Mean birth weight
* Small for gestational age:
gestational age calculated
from last menstrual period.
Weight less than the 10th
percentile based on sex-
specific growth curves
* Preterm birth: live births
<37-wk gestation (12
weighing >3,600 g were
recorded as full term)
Birth certificate data were
matched to Camp Lejeune
housing records to confirm
address and that pregnancy
occurred during occupancy
Well, dug in 1958, supplying
residents at Tarawa Terrace
Housing Areas I and II was
contaminated with PCE and other
volatile organic compounds from
a dry-cleaning business that
opened in 1954. Business
practices did not change between
1960 and 1985, when 3
contaminated wells were
disconnected from the TT water
distribution system (February 8).
Data on concentrations available
for 1982 and later. One well
(TT26) of 6 had detectable
contamination and proportion of
water from TT26 varied daily.
Water from all wells was mixed
prior to distribution
Concentration (ppb) in finished
water samples, 1982-85
May-June 1982
PCE 76-1,580
TCE ND-57
Exposed: TT residents
Unexposed: Remaining base
family housing units (minus
exclusions)—based on water
samples from supply wells and
finished water in 1984 and 1985
Potential confounders: infant's
sex and year of birth, mother's
race, age, educational level,
parity, adequacy of prenatal
care, marital status, and history
of fetal death, father's age,
educational level, and military
pay grade. Variable selection
by backward elimination
Exposed vs. unexposed
Difference in mean birth wt:
-26 g (90% CI: -43, -9)
SGA
OR (90% CI): 1.2(1.0, 1.3)
Preterm birth
1.0 (0.9, 1.1)
No discernable pattern with
duration of exposure estimated
by length of residence at TT
prior to giving birth
Adjustment for
confounders did not
alter risk estimates
for exposure
Did not control for
maternal smoking,
alcohol and height
No data on
concentration at tap
in individual homes,
water consumption
or showering
Exposure
misclassification:
Unexposed group
was exposed to PCE
prior to 1972
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
to
VO
to
Aschengrau et al. (2008: 2009a: 2009b)
United States
Population-based retrospective cohort
study of exposure to PCE in drinking
water after installation of water
distribution pipes lined with PCE-
impregnated vinyl liners (VL), selected
all births (index birth), 1969-1983,
from birth certificates with addresses in
one of 8 Cape Cod towns with some
VL/asbestos cement (AC) water
distribution pipes at the birth. Selected
1,492 with addresses with exposure to
VL/AC pipe and 1,704 frequency
matched to —eposed" by month and
year of birth. 959 (64.3% of selected,
70.5% of located) of exposed and
1,087 of referents (63.8% of selected,
69.3% of located) were enrolled
Included only pregnancies with
completely geocoded residential
histories (94.2% of reported
pregnancies)
Clinically recognized
pregnancy outcomes:
* miscarriages, stillbirths up
to Dec 1990 by self report,
self-administered
questionnaire
Final analysis included
5,567 pregnancies from
1,891 women
prevalence of loss among
eligible pregnancies: 11.8%
* Birthweight and
gestational age among
single healthy infants from
birth certificates
* Low birth weight
(<2,500 g)
* Premature birth (gestation
<37 wk)
* Intrauterine growth
retardation (IGR) (Birth
weight <10th percentile)
Congenital anomalies from
questionnaires
Residential history (1969-1983)
by questionnaire during
2002-2003
Could not obtain information on
water consumption and bathing
habits by residence
Estimated annual mass of PCE
delivered to each address before
and during the pregnancy using
EPANET water distribution
system modeling software with
algorithm for tetrachloroethylene
leaching and transport, and GIS
maps of residences and a town's
water distribution system
Estimated water concentration:
1-5,197 (ig/L
Exposure:
Cumulative: up to month and year
of last menstrual period (LMP)
Peak: up to LMP year of
pregnancy
Monthly average during the LMP
year
Before the LMP: 283 losses,
2,112 live births with some
exposure; 376 losses, 2,796 live
births with no exposure
Outcomes among exposed and
unexposed pregnancies
compared for each exposure
period of interest:
Cumulative, peak and average
monthly
Generalized estimating
equation models to account for
lack of independence of
outcomes
Considered several risk factors
for pregnancy loss, associated
with PCE exposure or
nonthinking water sources of
solvent exposure
No associations or patterns
observed for the 3 exposure
measures and pregnancy loss,
birth weight or duration of
gestation
All congenital anomalies
61 exposed, 95 unexposed
OR adjusted for maternal and
paternal age: 1.2 (95% CI:
0.8-1.7)
Nonparticipants
were slightly
younger (26 vs.
27.5 yr) and less
educated 11.3%
less than high
school vs. 3.6%)
but no difference
by exposure
Reproductive
history in medical
records for index
pregnancy
compared to self
reports for 60
women: 92% of
clinically
recognized
miscarriages and
100% of live
births in record
were reported by
participants
Compared
reproductive history
in birth certificates
(n = 2,490) to self
reports: good to
excellent agreement
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
a -r
>1
to 0\
2 ^
0
1
Aschengrau et al. (2008: 2009a: 2009b)
(continued)
During the LMP year: 213 losses
and 1,743 live births with some
exposure; 446 losses, 3,165 live
births with no exposure
Increased odds ratios for any
exposure and neural tube
defects (3.5, 95% CI:
0.8-14.0), oral clefts (3.2, 95%
CI: 0.7-15.0), gastrointestinal
(1.8, 95% CI: 0.7-4.4), and
genitourinary malformations
(1.6, 95% CI: 0.6-3.8)
No increased odds ratios for
cardiac and musculoskeletal
malformations
to
vo
LtJ
~n
H
O
O
Janulewicz et al. (2008)
United States
Population-based retrospective cohort
study of exposure to PCE in drinking
water after installation of water
distribution pipes lined with PCE
impregnated vinyl liners, selected all
births (index birth), 1969-1983, from
birth certificates with addresses in one
of 8 Cape Cod towns with some
VL/AC water distribution pipes at the
birth. Selected 1910 with addresses
from birth certificate with exposure to
VL/AC pipe from a database of all
street locations with VL/AC pipes and
1928 frequency matched to -exposed"
by month and year of birth. 1,240
(64.9% of selected, 70.9% of located)
of exposed and 1,250 of referents
(64.8% of selected, 70.2% of located)
were enrolled and returned self-
administered questionnaire
Learning and behavioral
disorders. Data collection
from mother by self-
administered questionnaire,
2002-2003. Diagnosis of
attention deficit disorder
(ADD) or hyperactivity
disorder (HD), tutoring for
math or reading, a special
class placement for
academic or behavioral
problems, an Individual
Education Plan from the
school system, and if the
child ever repeated a grade
Residential history (1969-1983)
by questionnaire
Estimated cumulative mass of
PCE delivered to each address
during prenatal and postnatal
periods using EPANET water
distribution system modeling
software with algorithm for PCE
leaching and transport, and GIS
maps of residences and a town's
water distribution system
PCE exposure calculated for
94.8% of study children with
completely geocoded residential
histories and date of last
menstrual period
Estimated water concentration:
Main streets: ND-80 |ig/L
Dead-end streets:
1,600-7,750 (ig/L
Prenatal and postnatal periods
analyzed separately
Multivariate GEE analyses with
identity or logit link function to
account for siblings
Final model for BW:
gestational age, maternal
education, race, history of LBW
child, occupational exposure to
solvents, use of self-service dry
cleaning, and proximity of any
residences to dry-cleaning
establishments
Nonparticipants
compared to
participants:
Similar for
distribution of births,
child's sex, race, and
prevalence of
children born with
LBW or premature;
Nonparticipants
were younger, less
educated, and had
more prior births.
Differences did not
vary by exposure
status
-------
a
o
<
s?
rs
>!
o
^3
§•
rs
st
3
>!
sT
St
o
>1
>!
Oq
o
>!
Si
Si
s
O
Sfc
I
§•
K) On
2
Table 4-34. Epidemiology studies on reproduction and development (continued)
Reference, population, study design
Outcomes
Exposure assessment
Key results
Notes
Janulewicz et al. (2008) (continued)
Included only pregnancies with
completely geocoded residential
histories (94.2% of reported
pregnancies)
2,125 subjects in final data set
Exposure:
• Cumulative prenatal: from
month and year of last
menstrual period to the month
and year of birth.
• Cumulative postnatal: from
month and year of birth
through month and year of the
child's 5th birthday
Final data set using refined
exposure assessment
Exposed: 1,349
Nonexposed: 737
Exposure variables divided into
quartiles
Final model for gestational age:
maternal education, race, prior
preterm delivery, obstetric
complications in the current
pregnancy, occupational
exposure to solvents, use of
self-service dry cleaning, and
proximity of any residences to
dry-cleaning establishments
No associations with prenatal or
postnatal exposure and
outcomes; some increased OR
in low exposure groups. For
example, ADD (OR [ 95% CI]):
Low: 1.4 [0.9-2.0]
High: 1.0 [0.7-1.6]
Did not use
information on water
consumption and
bathing habits by
residence—estimates
are not a direct
measure of PCE
intake by individuals
to
VO
~n
H
O
O
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Several studies in the United States of tetrachloroethylene in drinking water have
evaluated developmental risks (Aschengrau et al.. 2008; Aschengrau et al.. 2009b; Bove et al..
1995; Janulewicz et al.. 2008; Lagakos et al.. 1986; Sonnenfeld et al.. 2001). Lagakos et al.
(1986) reported the results of a population-based study in Woburn, Massachusetts, among
residents whose drinking water source was two wells contaminated with chlorinated organic
substances from 1960 to 1982 (see previous study description in discussion of spontaneous
abortion). Of the 3,809 infants that survived more than 7 days, 220 had low birth weights
defined as 6 pounds (not the typical definition of 2,500 g). The 177 medically diagnosed
congenital anomalies (4.6%) were grouped by the involved organ or system using ICD codes.
Sufficient cases existed for musculoskeletal (n = 55), cardiovascular (n = 43), and eye/ear defects
(n= 18) for separate analyses. CNS, chromosomal, and oral cleft anomalies were grouped
together because they contained few cases. The authors felt there was evidence from previous
studies to suggest that these anomalies may be associated with exposure to environmental
contaminants. The rest of the anomalies were grouped into a category called —diter." Childhood
disorders were compiled into nine categories. Incidence of childhood leukemia in relation to
exposure also was assessed and is described in the Section 4.6.1.2.5.
Logistic regression analyses, controlling for other risk factors, found no statistically
significant associations between the annual exposure score for the year a pregnancy ended and
musculoskeletal, cardiovascular, or —other birth anomalies. However, an association was
observed for eye/ear anomalies (OR: 14.9,p< 0.0001) and CNS/chromosomal/oral cleft
anomalies (OR: 4.5, p = 0.01). In an effort to evaluate potential recall bias, the authors checked
66 of 96 disorders (perinatal death post-1970, eye/ear, or CNS/chromosomal/oral cleft anomaly,
other childhood disorders) that had been confirmed in a second interview with medical records.
Of the 66 events, the authors were able to verify 62 using medical records. No relation of
reporting accuracy with exposure was found, thus, there was no evidence of recall bias, although
the authors did not attempt to check birth records among controls.
A prevalence study in four counties in New Jersey evaluated organic contaminants
monitored in the public water supply in relation to birth outcomes (Bove et al.. 1995). All live
births and fetal deaths reported on birth or death certificates between January 1, 1985, and
December 31, 1988, among residents of 75 out of 146 towns were ascertained. The final data set
included 80,938 singleton live births and 594 fetal deaths that were not therapeutic abortions or
chromosomal anomalies. Birth weights and gestational age were obtained from birth or death
certificates. Birth defects for live births were obtained from the New Jersey Birth Defects
Registry, a population-based, passive reporting system. Additional birth defects were
ascertained from fetal death certificates (>20 weeks gestation). Categorical outcomes were
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-295 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
compared to all live full-term births in the study population that were normal weight and had no
birth defects (n = 52,334).
Monthly levels of the contaminants of interest in each town were estimated from
sampling data (at least one sample per 6-month period) obtained from the New Jersey
Department of Environmental Protection and Energy and the 49 water companies that served the
communities. The monthly estimates were assigned to each gestational month for each live birth
and fetal death. Fetal death and birth defects were evaluated in relation to levels averaged over
the first trimester. Other birth outcomes were analyzed in relation to levels averaged over the
entire pregnancy. Average tetrachloroethylene concentrations during the first trimester for all
live births and fetal deaths were 26 ppb.
Tetrachloroethylene concentrations during the first trimester were <1 ppb among 82% of
the comparison group. Concentrations were >1-5 ppb, >5-10 ppb, and >10 ppb for 11.5, 5.1,
and 1.4% of the comparison group, respectively. Infants in the >10-ppb group were 27.2 g
lighter (50% CI: -13.4-41.0). The regression models were adjusted for maternal age, race and
education, primipara, previous stillbirth or miscarriage, sex of the birth, and adequacy of prenatal
care, plus total trihalomethane levels. The odds ratio for very low birth weight was 1.49
(50%) CI: 1.13-1.97) among term births in the >10-ppb group. An odds ratio of 1.16
(50%) CI: 0.69-1.83) was observed for neural tube defects among singleton live births and fetal
deaths in the >5-ppb group. The odds ratio for oral clefts in the >10-ppb group was 3.54
(50%o CI: 2.12-5.57). There were 67, 11, 1, and 4 oral cleft cases in the <1 ppb (referent),
>1-5 ppb (OR: 1.17, 50% CI: 0.89-1.53), >5-10 ppb (OR: 0.24, 50% CI: 0.05-0.63), and >10-
ppb tetrachloroethylene exposure groups, respectively. The authors also reported 90 and 99%>
CIs for odds ratios over 1.5. For oral clefts, the 90 and 99%> CIs for the odds ratio in the >10-ppb
group were 1.28-8.78 and 0.82-12.15, respectively. When multipollutant models including all
contaminants with associations were evaluated, the authors stated that tetrachloroethylene was no
longer associated with neural tube defects, and the odds ratio for oral cleft defects was reduced to
2.0 (CIs were not presented). In the multipollutant model, the odds ratios for trichloroethylene
and total trihalomethanes increased to 3.5. Therefore, while tetrachloroethylene appeared to
increase risk for very low birth weight, neural tube defects, and oral clefts, other monitored
drinking water contaminants also were associated with increased risk, and the contribution of
individual substances cannot be determined.
A study of birth outcomes among singleton liveborn and stillborn infants, >20 weeks,
was conducted at the U.S. Marine Corps Base at Camp Lejeune in North Carolina for the period
1968-1985 (ATSDR. 1998b: Sonnenfeld et al.. 2001). Tetrachloroethylene and other volatile
organic compounds used by a nearby dry-cleaning business contaminated drinking water
supplied to two housing areas on the base (Tarawa Terrace I and II) until the contaminated wells
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-296 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
were disconnected in 1985. Water concentrations measured in samples taken between 1982 and
1985 ranged from 76 to 1,580 ppb for tetrachloroethylene and from not detected (<10 ppb) to
57 ppb for trichloroethylene. The study population included births to mothers living in base
family housing at delivery and for at least 1 week prior. Residents of Tarawa Terrace I and II
were defined as exposed (n = 6,117 births). On the basis of water samples collected from wells
and finished water during 1984 and 1985, residents of the remaining base family housing units
were defined as unexposed (n = 5,681 births). Information on birth weight, gestational age, and
preterm birth (live births less than 37 weeks gestation) was obtained from North Carolina birth
records. To define small for gestational age, a gestationalage specific birth weight distribution
for a Caucasian population in California (Williams et al.. 1982) was found to best describe the
distribution of live births among the nonexposed group. Because standard birth weight
distributions for military populations were not available, the California reference was used to
identify a weight that classified 10% of births as small for gestational age in the nonexposed
group. In models including a term for gestational age, mean birth weight among exposed infants
was 26 g lower than the nonexposed infants (95% CI: -43, -9). The odds ratios for small for
gestational age and preterm birth were 1.2 (95% CI: 1.0-1.3) and 1.0 (95% CI: 0.9-1.1),
respectively. Regression models included several covariates to evaluate confounding, which
were retained after backward elimination; however, some known factors associated with birth
were not evaluated (maternal smoking, alcohol consumption, or height). Because exposure
status was associated with mother and father's education, father's military pay grade, and
mother's age, the unexamined risk factors also may have been associated with exposure and may
have acted as confounders. Final models for mean birth weight included mother's age, history of
one previous fetal loss, history of two or more fetal losses, gestational age, mother's race, living
in an officer's or warrant officer's household, year of birth, and sex of the infant. Final models
for small for gestational age included mother's age, mother's history of one previous fetal loss,
history of two or more previous fetal losses, primiparity, living in an officer's or warrant
officer's household, year of birth, and mother's education. The authors also reported the results
of regression models containing cross-product terms for exposure and maternal age (<35 years,
>35 years) or number of previous fetal losses (none, 1, >2). Among mothers 35 years of age or
older, infants of exposed mothers weighed 104 g less than infants of unexposed mothers (90%
CI: -236, -23). Birth weights of infants born to women less than 35 years of age were not
different between exposure groups. In addition, among women with >2 previous fetal losses,
exposed infants were 104 g lighter than unexposed infants (90% CI: -174, -34). Mother's age
and history of previous fetal loss also appeared to modify the tetrachloroethylene risk for small
for gestational age. The odds ratios for small for gestational age were 1.1 (90% CI: 0.9-1.2) and
2.1 (90% CI: 0.9-4.9) among women <35 and >35 years of age, respectively. There were only
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-297 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
11 exposed and 8 unexposed small for gestational age infants among mothers older than
35 resulting in effect estimates with lower precision. Odds ratios were 1.1 (90% CLO.9-1.2), 1.5
(90% CI: 1.1-2.0), and 2.5 (90% CI: 1.5-4.3) among women with none, 1, and >2 previous fetal
losses, respectively. There were 43 exposed and 14 unexposed small for gestational age infants
among mothers with >2 previous fetal losses. The authors did not present tests for interaction.
The study found small differences in birth weight and a small increased risk of small for
gestational age among live births to mothers living in two housing areas at the military base with
exposure to tetrachloroethylene and other volatile organic compounds in their drinking water.
Although the impact of residual confounding by unmeasured covariates is not known, a possibly
larger problem may be exposure misclassification. Samples were collected over the last 3 years
of the 17-year study period, although the dry-cleaning business operated during the entire period,
and no operational changes occurred. Water pumped from the contaminated well was mixed
with water from five other wells, but the proportion of water provided from the individual wells
varied from day to day. Variation in concentrations delivered to the tap, as well as individual
consumption and exposure through bathing, could not be evaluated in this study. Further, any
movement on the base prior to delivery was not accounted for. During the course of an exposure
reconstruction study, ATSDR learned that some of the cohort initially considered to be
unexposed were in fact supplied with contaminated water from the Hadnot distribution system
between 1968 and 1972 and for a 2-week period in the winter of 1985 (NRC. 2009):
www.atsdr.cdc.gov/HS/lei eune/erratum.htm). Exposed pregnancies during 1968-1972 were
erroneously classified as unexposed. This calls into question the findings in Sonnenfeld et al.
(2001): however, it is likely that as a result of the misclassification, any associations with birth
outcome, if they exist, would have been biased toward the null. Aschengrau et al. (2008) did not
observe an association of tetrachloroethylene in drinking water with either birth weight or
gestational duration. This study, described previously in the discussion of spontaneous abortion,
evaluated effects on pregnancy and development from tetrachloroethylene in drinking water
delivered to homes in the Cape Cod region in Massachusetts between 1968 and 1980. A group
of 1,910 children (1,862 singleton, 24 sets of twins) were born between 1969 and 1983 to
mothers living in one of several Cape Cod towns where tetrachloroethylene leached into drinking
water from vinyl-lined pipes in the water distribution system. Children initially designated as
unexposed (1,853 singleton, 37 sets of twins) were randomly chosen from the remaining resident
births and were frequency matched to the exposed group by month and year of birth. Response
among mothers who were successfully located was comparable between the exposed and
unexposed groups (70%); in the end, 56.4% of selected births designated as exposed were
included, and 54.4% of selected births designated as unexposed were included. After exposure
modeling, 1,353 exposed and 772 unexposed healthy, singleton births were identified.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-298 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
The prevalence of prior low birth weight infants in the cohort was low: 5% (n = 68)
among the exposed and 3.4% (n = 26) among the unexposed group. No differences in mean
birth weight or odds ratios for low birth weight (<2,500 g) or intrauterine growth retardation
(<10th percentile based on U.S. age-, sex-, and race-specific cut-offs, 1970-1976) were observed
by exposure status. Generalized estimating equation regression models for birth weight
differences adjusted for gestational age, maternal race, educational level, history of a low-birth-
weight child, occupational exposure to solvents, use of self-service dry cleaning, and proximity
of any residences to dry-cleaning establishments. Mean birth weights were slightly greater
among exposed infants in almost all quartiles for all of the three exposure measures, but the
estimates were statistically imprecise, and no pattern by exposure amount was observed.
Average monthly maternal exposure during the year of the last menstrual period in quartiles was
associated with increases in birth weight of 20.9, 6.2, 30.1, and 15.2 g compared to no exposure.
Models of gestational age were adjusted for maternal race, educational level, prior preterm
delivery, obstetric complications in the current pregnancy, occupational exposure to solvents, use
of self-service dry cleaning, and the proximity of any residences to dry-cleaning establishments.
Estimates of the difference in duration of gestation with increasing quartiles of exposure during
the year of the last menstrual period were -0.2, 0.1,-0.1, and -0.2 weeks. CIs were wide,
included the null, and did not indicate a pattern by exposure amount.
The study of exposure from leaching tetrachloroethylene in water distribution pipes
installed between 1968 and 1980 in the Cape Cod region in Massachusetts also assessed the risk
of congenital anomalies reported by participants (Aschengrau et al.. 2009b). Congenital
anomalies were coded by two study investigators, blind to exposure status, in consultation with a
pediatrician using guidelines from the Metropolitan Atlanta Congenital Defects Program. Of the
total of 4,657 children reported by the mothers, 643 were excluded because they were born after
1990, were missing prenatal information, were from multiple pregnancies, were exposed to
known teratogens, mothers smoked marijuana daily or weekly, or drank 7 or more alcoholic
drinks during pregnancy. There were 61 children with congenital anomalies among the 1,658
with prenatal exposure, and 95 children with congenital anomalies among the 2,999 with no
prenatal exposure. The unadjusted odds ratio (generalized estimating equation regression) for all
congenital anomalies was 1.1 (95% CI: 0.8-1.6) for any prenatal exposure to
tetrachloroethylene. Simultaneous control for maternal and paternal age did not change the odds
ratio. This also was true when other potential confounders were included one at a time (calendar
year of birth, mother's educational level, cigarette smoking, alcoholic beverage consumption,
prior pregnancy loss, and child's gender). Among children with an average monthly prenatal
exposure greater than or equal to the 75th percentile (2.3 g), the odds ratio was 1.5 (95%
CI: 0.9-2.5). Although case numbers were low, increased odds ratios were observed for several
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-299 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
organ systems, diagnostic groups, and any prenatal exposure compared to none. These included
neural tube defects (3.5, 95% CI: 0.8-14.0, n = 6 exposed cases), oral clefts (3.2, 95%
CI: 0.7-15.0, n = 5 exposed cases), gastrointestinal malformations (1.8, 95% CI: 0.7-4.4, n= 11
exposed cases), and genitourinary malformations (1.6, 95% CI: 0.6-3.8, n = 11 exposed cases).
Odds ratios for cardiac (0.9, 95% CI: 0.4-2.0, n = 9 exposed cases) and musculoskeletal
malformations (0.9, 95% CI: 0.5-1.6, n= 19 exposed cases) were not increased, and risk was not
estimated for eye, ear, respiratory, and other malformations because the number of cases was too
low.
As discussed previously, nondifferential exposure misclassification was likely given the
lack of individual level exposure information, which may have resulted in lower observed risk
estimates. In addition, the authors stated that the prevalence of anomalies, particularly minor
ones, may have been underreported by the mothers because it was lower in the study population
than reported by other monitoring programs. This would affect the statistical power of the study.
The authors did not believe that recall was differential with respect to exposure status because
most of the respondents did not know whether or not they were exposed.
Risk of learning and behavioral disorders was evaluated in relation to prenatal and
postnatal exposure to tetrachloroethylene in the Cape Cod towns with a contaminated water
distribution system (Janulewicz et al.. 2008). The authors did not observe an association with
increasing amount of exposure among children born between 1969-1983 whose mothers lived in
one of the towns with vinyl-lined asbestos-cement pipes at the time of birth. The study is
discussed in detail in Section 4.1.
In summary, some studies of tetrachloroethylene in drinking water suggest that exposure
during pregnancy is associated with low birth weight (Bove et al.. 1995: Lagakos et al.. 1986).
eye/ear anomalies (Lagakos et al.. 1986). and oral clefts (Aschengrau et al.. 2009b: Bove et al..
1995: Lagakos et al.. 1986). No associations with tetrachloroethylene exposure were reported
for small for gestational age (Bove et al.. 1995) or other classifications of congenital anomalies
(e.g., musculoskeletal, cardiovascular) (Aschengrau et al.. 2009b: Lagakos et al.. 1986).
Although a small increase in risk of small for gestational age was reported for infants exposed
prenatally to tetrachloroethylene at the Camp Lejeune military base, the finding remains
inconclusive until ATSDR completes its reanalysis (Sonnenfeld et al.. 2001). Aschengrau et al.
(2008) did not observe associations with birth weight or gestational age in a Cape Cod
population exposed to a wide range of tetrachloroethylene concentrations in drinking water.
Occupational studies of dry-cleaning and laundry workers in Scandinavia could not evaluate
specific congenital anomalies because few cases were identified (Ahlborg. 1990b: Kyyronen et
al.. 1989: Lindbohm. 1995: Olsen et al.. 1990: Taskinen et al.. 1989). The number of cases with
birth anomalies in specific diagnostic groups was very small in all of the studies, and CIs often
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-300 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
included one. In addition, imprecise exposure estimates likely resulted in nondifferential
misclassification and a bias of risk estimates toward the null. Participants in the studies were
exposed to multiple contaminants, and it was not possible to analyze substance-specific risks.
Finally, a more than threefold risk of schizophrenia was associated with dry cleaning as a
surrogate for prenatal tetrachloroethylene exposure (Perrin et al., (2007). discussed in
Section 4.1). The longitudinal design and use of a national registry to identify psychiatric
diagnoses were strengths of the study, but tetrachloroethylene exposure was not directly
analyzed.
4.7.1.2. Animal Developmental Toxicity Studies
Evaluation of the developmental effects of tetrachloroethylene exposure in mammalian
animal models is based on several studies of in utero exposures to maternal animals during
specific periods of pregnancy. Additionally, evaluations of the developmental neurotoxic
potential of tetrachloroethylene have been conducted in rats. These studies are described below.
4.7.1.2.1. In vitro developmental toxicity assay
Saillenfait et al. (1995). using a rat whole embryo (Day 10) culture system, found
tetrachloroethylene-induced embryo toxicity, including mortality, malformations, and delayed
growth and differentiation. No adverse effect was produced at the 2.5-mM concentration, but
concentration-related trends of increasing toxicity occurred from 3.5 through 15 mM. Statistical
tests for a concentration-related trend were not reported. The investigators found that
trichloroethylene produced similar effects, with potency somewhat less than that of
tetrachloroethylene. They also found that TCA and DCA caused a variety of abnormalities in
this culture system.
4.7.1.2.2. Nonmammalian developmental toxicity assay
Spencer et al. (2001) evaluated the effects of tetrachloroethylene on the embryonic
development of Japanese medaka. In this study, 1-day-old in ovo embryos were exposed to
concentrations of 0, 20, 40, 60, or 80 mg/L for 96 hours or to concentrations of 0, 1.5, 3, 6, 12, or
25 mg/L for 10 days. Viability, hatchability, and morphological/developmental abnormalities
were evaluated. A 96-hour LC50 of 27.0 mg/L was identified for egg viability. Following
10 days of exposure, hatchability and larval survival were significantly decreased, and
developmental abnormalities were significantly increased in a concentration-dependent manner.
At the lowest concentration tested (1.5 mg/L), developmental findings included abnormalities of
the circulatory system, yolk-sac edema, pericardial edema, scoliosis, hemorrhaging, blood
pooling, and cardiac morphological defects. The study authors concluded that
tetrachloroethylene is teratogenic to the Japanese medaka.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-301 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
4.7.1.2.3. In vivo mammalian screening study
In a developmental toxicity screening study, timed-pregnant F344 rats were treated by
gavage with tetrachloroethylene at doses of 900 or 1,200 mg/kg-day in corn oil vehicle on GDs
6-19 (Narotsky and Kavlock. 1995). There were 17 dams in each of the tetrachloroethylene-
treated groups and 21 in the control groups. The dams were allowed to deliver, and their litters
were examined on PNDs 1, 3, and 6. At 1,200 mg/kg, no live pups were delivered on GD 22. At
900 mg/kg-day, there was maternal ataxia, and weight gain was markedly less than in the
controls. The number of pups per litter was reduced (p < 0.01) as compared with the controls at
GD 22. On PND 6, the number of pups per litter was reduced (p < 0.001) as compared with the
controls. The investigators noted that full-litter resorptions were not observed with other
chemicals they tested in the presence of maternal toxicity. An increase in micro/anophthalmia
was found in the offspring. There was no evaluation for skeletal changes, and not all available
pups were examined for soft tissue changes.
4.7.1.2.4. In vivo prenatal developmental toxicity studies
Schwetz et al. (1975) conducted an inhalation developmental toxicity study, in which
25-30 Sprague-Dawley rats and 30-40 Swiss-Webster mice were exposed to airborne
tetrachloroethylene at 300 ppm, 7 hours/day, on GDs 6-15. Following laparohysterectomy on
GDs 21 or 18 (for rats and mice, respectively), fetuses were weighed and measured, examined
for external abnormalities, and processed for the evaluation of either soft tissue or skeletal
abnormalities. Three other organic solvents were also tested with the same protocol; the
concentration of all agents was chosen to be approximately twice their threshold limit values.
Although the study authors concluded that there was no significant maternal, fetal, or embryo
toxicity for any of the solvents tested, the maternal and fetal data demonstrated a number of
statistically significant differences from control values following gestational exposures to
tetrachloroethylene in rats and mice. In the rats, exposures to tetrachloroethylene produced
slight, but statistically significant, maternal toxicity (4-5% reductions in mean maternal body-
weight gains) and embryotoxicity (increased resorptions; 9% in treated vs. 4% in controls). In
the mice, maternal toxicity consisted of a significant 21% increase in mean relative liver weight
as compared with controls. The mean fetal weight in mice was significantly (9%) less than in the
concurrent control, and the percentage of litters with delayed ossification of the skull bones,
delayed ossification of the sternebra, and subcutaneous edema was significantly increased. Due
to the single exposure level used in this study, a dose response could not be determined.
Szakmary et al. (1997) exposed CFY rats to tetrachloroethylene via inhalation throughout
gestation (i.e., GDs 1-20) for 8 hours/day at concentrations of 1,500, 4,500, or 8,500 mg/m3. In
the same study, the study authors exposed C57B1 mice via inhalation on GDs 7-15 (i.e., during
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-302 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
the period of organogenesis) to a concentration of 1,500 mg/m3 and New Zealand white rabbits
during organogenesis (GDs 7-20) to a concentration of 4,500 mg/m3. Maternal animals were
killed approximately 1 day prior to expected delivery; a gross necropsy was conducted, organ
weights were recorded, blood was taken by aorta puncture for hematology and clinical chemistry
evaluations, ovarian corpora lutea were counted, and uterine contents were examined (number
and position of living, dead, or resorbed fetuses; and fetal and placental observations and
weights). The numbers of litters available for evaluation were as follows: 20 control and 21 or
22 per treated group in the rat, 77 control and 10 treated in the mice, and 10 control and
16 treated in the rabbit. One-half of the fetuses from each litter were evaluated for visceral
abnormalities, and the other half were evaluated for skeletal development. The study authors
reported that the organs of five dams and five embryos from each group were also evaluated by
routine histological methods. To evaluate the concentration of tetrachloroethylene in maternal
and fetal blood and in amniotic fluid, another subset of rats (number not specified) was studied.
(For the 1,500- and 8,500-mg/m3 exposure levels, maternal blood concentrations of
tetrachloroethylene were 17.8 + 8.9 and 86.2 + 13.0 [iL/mL, respectively. Concentrations in the
fetal blood were 66 and 30% of maternal blood concentrations, and amniotic fluid concentrations
were 33 and 20% of maternal blood concentrations.) In the rat, at 4,500 and 8,500 mg/m3,
maternal body-weight gain during gestation was significantly decreased (37 and 40%,
respectively), relative maternal liver mass was significantly increased (10 and 6%, respectively),
and serum aspartate amino transferase activity was increased (data not provided) as compared to
controls. Percentage preimplantation loss was significantly increased from controls by 133 and
117%) at these exposure levels, while percentage postimplantation loss was increased
nonsignificantly from controls by 80% in each group. Also, at 4,500 and 8,500 mg/m3, fetal
weight was significantly decreased in 98.5 and 100% of all fetuses, the number of fetuses with
skeletal retardation was significantly increased in 98.5 and 100% of fetuses, and the percentage
of fetuses with malformations was both significantly increased to 6.4 and 15.7% as compared to
the control incidence of 2.0%. Although the study authors judged the 1,500-mg/m3 exposure
level to be the NOAEL for the rat study, it is noted that there were concentration-dependent
nonsignificant decreases in maternal body-weight gain (13% lower than control), and increases
in pre- and postimplantation loss (49 and 38% greater than control, respectively). The
percentage of weight-retarded fetuses increased to 3.4 times the control incidence, and the
incidences of fetuses with skeletal retardation (48% increased) or total malformations increased
by 2.3 times the control incidence observed at the low-exposure level of 1,500 mg/m3.
Therefore, these findings are judged to be adverse consequences of treatment. The attribution of
these findings to treatment, and the designation of 1,500 mg/m3 as the study LOAEL is
consistent with the adverse developmental findings of Schwetz et al. (1975). In mice
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-303 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
(1,500 mg/m3) and rabbits (4,500 mg/m3), relative liver mass was significantly increased;
decreased maternal body-weight gain was also observed in the rabbits. In the mice, a
significantly increased number of fetuses with visceral malformations (details not specified) was
observed, while in the rabbits, 2/16 does aborted, total resorption of four litters was reported, and
the percentage of postimplantation loss was significantly increased. The percentage of rabbit
fetuses with malformations (details not provided in the report) was also increased, although not
significantly.
Hardin et al. (1981) [see also (Beliles et al.. 1980)1 exposed Sprague-Dawley rats
(30/group) and New Zealand white rabbits (20/group) via inhalation to 500 ppm of
tetrachloroethylene for 7 hours/day, 5 days/week. Tetrachloroethylene was administered with
and without 3-week pregestation exposures and with both full-term and terminal two-thirds-term
exposure. No maternal or developmental toxicity was identified.
In a developmental toxicity study, Carney et al. (2006) investigated the effects of whole-
body inhalation exposures to pregnant Sprague-Dawley rats at nominal concentrations of 0-, 75-,
250-, or 600-ppm (actual chamber concentrations of 0, 65, 249, or 600 ppm) tetrachloroethylene
for 6 hours/day, 7 days/week on GDs 6-19. This study was conducted under Good Laboratory
Practice (GLP) regulations according to current EPA and OECD regulatory testing guidelines.
Maternal toxicity consisted of slight, but statistically significant, decreases in body-weight gain
during the first 3 days of exposure to 600 ppm, establishing a no-adverse-effect concentration of
249 ppm for dams. A slight, statistically significant decrease in gravid uterine weight at 600
ppm correlated with significant reductions in mean fetal body weight (9.4%) and placental
weight (15.8%) at GD 20 cesarean section. At >249 ppm, mean fetal and placental weights were
significantly decreased by 4.3 and 12.3% from control, respectively. A significant increase in
the incidence of incomplete ossification of the thoracic vertebral centra at this exposure level was
consistent with fetal growth retardation. No treatment-related alterations in fetal growth or
development were noted at 65 ppm. Therefore, the LOAEL for this study is 249 ppm.
4.7.1.2.5. Developmental neurotoxicity
Developmental neurotoxicity data are also discussed in Section 4.1.2.
A cohort of rats from the Szakmary et al. (1997) study (15 litters/group at exposure levels
of 1,500- or 4,500-mg/m3 tetrachloroethylene) was allowed to deliver, and the offspring
(standardized to 8 pups/litter) were maintained on study to PND 100. It was not clearly specified
in the report whether the daily inhalation exposures continued throughout the postnatal period.
Preweaning observations included weekly body weights, developmental landmarks (pinna
detachment, incisor eruption, and eye opening), and functional assessments (forward movement,
surface righting reflex, grasping ability, swimming ontogeny, rotating activity, auditory startle
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-304 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
reflex, and examination of stereoscopic vision). After weaning, exploratory activity in an open
field, motor activity in an activity wheel, and development of muscle strength were assessed.
The study authors reported that adverse findings included a decreased survival index (details not
provided), minimally decreased exploratory activity and muscular strength in treated offspring
(presumably at both exposure levels) that normalized by PND 51, and significantly increased
motor activity on PND 100 of females exposed to 4,500 mg/m3 of tetrachloroethylene.
Nelson et al. (1980) investigated developmental neurotoxicity in Sprague-Dawley rats by
exposing pregnant dams (13-21/group) to tetrachloroethylene at concentrations of 100 ppm or
900 ppm during either early pregnancy (GDs 7 to 13) or late pregnancy (GDs 14 to 20).
Morphological examination of the fetuses (gross, visceral, and skeletal) was performed, and
behavioral testing and neurochemical analyses of the offspring were conducted. There were no
alterations in any of the measured parameters in the 100-ppm groups. At 900 ppm, there were no
skeletal abnormalities, but the weight gain of the offspring as compared with controls was
depressed approximately 20% at postnatal Weeks 3-5. Developmental delays were observed in
both the groups exposed during early and late pregnancy. Offspring of the early pregnancy-
exposed group performed poorly on an ascent test and on a rotorod test, whereas those in the late
pregnancy group underperformed on the ascent test at only PND 14. However, later in
development (Days 21 and 25), their performance was higher than that of the controls on the
rotorod test. These pups were markedly more active in the open field test at Days 31 and 32.
Activity wheel testing on Days 32 and 33 did not reveal statistically significant changes.
Avoidance conditioning on Day 34 and operant conditioning on Days 40-46 did not identify
treatment-related effects. Neurochemical analyses of whole brain (minus cerebellum) tissue in
21-day-old offspring revealed significant reductions in acetylcholine levels at both exposure
periods, whereas dopamine levels were reduced among those exposed on GDs 7-13. All of the
described effects in the 900-ppm group were statistically significant as compared with controls.
Unfortunately, none of the statistics for the 100-ppm treatments were presented. The authors
observed that more behavioral changes occurred in offspring exposed during late pregnancy than
in those exposed during early pregnancy.
Additional evidence of potential developmental neurotoxicity was reported by
Fredriksson et al. (1993). In this study (see Section 4.1.2.2), tetrachloroethylene was
administered to male NMRI mice by gavage at dose levels of 0, 5, or 320 mg/kg-day on PNDs
10-16. At PND 17 and 60, spontaneous activity (locomotion, rearing, and total activity) was
measured over three, 20-minute periods. No treatment-related alterations in activity were
observed at 17 days of age; however, at 60 days of age, all three measures of spontaneous
activity were altered.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-305 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
4.7.2. Reproduction
4.7.2.1. Human Reproduction Data
Studies of tetrachloroethylene exposure have evaluated several outcomes including
effects on menstrual disorders (Zielhuis et al.. 19891 semen quality (Eskenazi etal.. 1991bb).
fertility (Eskenazi et al.. 1991aa; Rachootin and 01 sen. 1983). time to pregnancy (Sallmen et al.,
(1998; 1995). and spontaneous abortion (McDonald et al., (1986; 1987); Lindbohm et al.,
(Ahlborg. 1990b; Bosco et al.. 1987; Doyle et al.. 1997; Kyyronen et al.. 1989; 1991; 1990;
Olsen et al.. 1990; Taskinen et al.. 1989; Windham et al.. 1991). Many of the studies evaluated
exposure during a specific critical window for development, usually the first trimester.
In a letter to the editor, Zielhuis et al. (1989) described the results of a cross sectional
study of menstrual disorders among dry-cleaners and laundry workers in the Netherlands. A
total of 471 of 592 women returned a mailed questionnaire (80%). The sampling frame for
recruitment was not described. After excluding 72 respondents because the woman was
currently pregnant or lactating at the time of administering the questionnaire or reported a
chronic illness or gynecological surgery, and excluding another 324 respondents because the
woman reported use of oral contraceptives, the final data set included 68 exposed and 76
unexposed women. Exposure was defined on the basis of occupation (dry cleaners versus
laundry workers). The authors reported that the exposed and unexposed groups were similar
with respect to age, lifestyle, work conditions, and personal characteristics (body mass index,
number of children, and use of contraceptives). Risk of specific menstrual characteristics by
occupation was evaluated using linear logistic regression adjusting for age, body mass index,
substantive weight changes, number of children, history of diseases, sporting activities, life
events, smoking, alcohol consumption, medical drugs, and work conditions other than exposure
to tetrachloroethylene. Prevalence of menstrual conditions in the population varied between
10% (oligomenorrhea, premenstrual syndrome) to 30% (unusual cycle length) and occurred with
greater frequency among dry cleaners compared to laundry workers for all symptoms except for
one (polymenorrhea). There were no reports of amenorrhea. Elevated odds ratios were observed
for several of the symptoms including oligomenorrhea (2.1, 90% CI: 0.9-5.3), unusual cycle
length (2.3, 90% CI: 1.2-4.4), menorrhagia (3.0, 90% CI: 1.6-5.6), dysmenorrhea (1.9,
90% CI: 1.1-3.5), and premenstrual syndrome (3.6, 90% CI: 1.5-8.6). This study indicates that
working in dry cleaning may adversely affect menstruation, but the lack of detail in reporting
precludes a thorough assessment of selection bias or confounding. In addition, the assignment of
exposure status by industry also precludes a definitive conclusion regarding a potential
association with tetrachloroethylene.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-306 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Semen quality was evaluated among men who worked in the dry-cleaning industry
compared to men working in laundries in California (Eskenazi et al.. 1991bb). The population,
recruited from membership lists of the Laundry and Dry Cleaners Union Locals 3 (San Francisco
Bay area) and 52 (Greater Los Angeles), included all dry cleaners (n = 85) and all laundry
workers, 20-50 years of age, in Local 3 (n = 119) and a randomly selected sample of Local 52
members (n = 206). Laundry workers were frequency matched by age to dry cleaners from the
same Local. Dry cleaners also were recruited from nonunion shops in the San Francisco area.
Eligible individuals were 20-50 years of age, current workers in the industry, spoke English or
Spanish, had not had a vasectomy, and were located by telephone or mail. Respondents included
20 union drycleaners (38% of 53 eligible) and 56 union laundry workers (34% of 166 eligible),
plus 13 nonunion dry cleaners. Men were considered exposed if they worked in the dry-cleaning
industry or a laundry where dry cleaning was performed. The unexposed group included laundry
workers at businesses where dry cleaning was not conducted. After exposure was assessed, the
final data set included 34 exposed workers and 48 unexposed workers with adequate semen
samples and confirmed type of establishment. Information on sociodemographic characteristics,
reproductive and medical history, and personal habits was collected by interview. In addition, a
detailed work history including job tasks and exposures during the previous week and the past
3 months was obtained. A physical exam was conducted by a study physician blind to exposure
status, and participants returned a semen sample collected after at least 2 days of abstinence.
The semen was analyzed for sperm concentration, morphology, and motility. Each sperm
measure was evaluated in relation to three measures of exposure: dry cleaning versus laundry,
tetrachloroethylene in exhaled breath (limit of quantitation: 2.67 |ig/m3), and an exposure index
encompassing the entire period of spermatogenesis (approximately 3 months). Exhaled air was
measured 16-19 hours after the end of the workweek or was corrected to 16 hours using an
elimination model (11 samples). An industrial hygienist assigned an exposure score using
responses to the questionnaire concerning job task (e.g., machine operator, presser, etc.), the type
of dry-cleaning machine used (e.g., wet to dry transfer, dry-to-dry) and other tasks and attributes
known to influence the level of exposure to tetrachloroethylene. The exposure score ranged
from 0 among unexposed men to 11 among the exposed group. The association of semen
parameters with tetrachloroethylene exposure was analyzed using multiple linear regression with
adjustment for potential confounding variables that were associated with both the semen
parameter and any of the exposure measures. Models of three clinically relevant measures of
semen quality, oligospermia (<20 million/mL), >40% abnormal forms, and <60% motile sperm,
were not associated with any exposure measure among the entire cohort. Of four measures of
sperm motility, Ln median amplitude of lateral head displacement was associated with Ln
tetrachloroethylene in exhaled air among all 82 participants (t = 2,0, p = 0.05), adjusting for
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-307 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
ethnicity, education, religion, and physical abnormalities found on exam. Exposure scores and
industry group were not statistically significant predictors of this semen parameter. However, Ln
tetrachloroethylene levels (I = 2.14,p = 0.04) and exposure score (I = 3,07, p = 0.005) were
predictors of amplitude of lateral head displacement among the 34 participants in the exposed
group. Sperm linearity was inversely associated with exposure score in both analytic groups
(t = -2,57, p = 0.02). Percentage of round sperm was statistically significantly associated with
all three exposure measures, controlling for history of STD and working in temperatures over
100°F among all participants but not in the dry-cleaning group alone. Percentage of narrow
sperm was inversely associated with all three exposure measures controlling for ethnicity,
number of days working in temperatures greater than 80°F, and use of marijuana among all
participants. Among the dry cleaners, Ln percentage narrow sperm was inversely related to Ln
tetrachloroethylene levels (t = -2.29,p = 0.03) but not by exposure score (t = 0.92, p = 0.36).
Tetrachloroethylene exposure appeared to alter sperm quality in this population of
unionized dry cleaners. However, the effects were subtle, and the clinical significance of the
semen parameters associated with tetrachloroethylene exposure is not clear. The low response
rate in the primarily unionized cohort limits generalizations to the industry as a whole.
Reproductive outcomes also were evaluated among the wives of the men who participated in the
study of semen quality (Eskenazi et al.. 1991b). Telephone interviews were conducted with 17
wives of the 20 married dry cleaners (85%) and 32 wives of the 36 married laundry workers
(89%) in the original cohort. Pregnancies and miscarriages during the years of their husbands'
employment in the industry were identified among 14 wives of dry cleaners and 26 wives of
laundry workers. Standardized fertility ratios were calculated using the U.S. national birth rates
during periods of employment in the industries and periods when the men were not employed in
the industries as a comparison. Investigators also analyzed the number of months to conception
for the last pregnancy during the period of employment in the industries. The wives of laundry
workers were more likely to be Hispanic, Catholic, to have smoked during the year of the index
pregnancy, and to have a history of reproductive disease or surgery. They had fewer years of
education, and a greater proportion weighed more. The wives also were more likely to work in
dry cleaning and laundries, confounding the source of exposure.
Fertility rates among the wives of dry-cleaners and laundry workers were higher than the
national average for women of the same race, parity, birth cohort, and age. The standardized
fertility ratios were comparable in both industry groups. However, it took longer for the wives of
dry cleaners to achieve the index pregnancy compared to the wives of laundry workers
(8.2 ± 10.2 months versus 4.1 ± 5.8 months, respectively, p = 0.08). In Cox Proportional
Hazards Models with adjustments for ethnicity (Hispanic vs. non-Hispanic) and smoking, the
per-cycle pregnancy rate of wives of dry cleaners was approximately one-half that of the wives
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-308 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
of laundry workers (rate ratio = 0.54, 95% CI: 0.23-1.27). A rate ratio of less than 1 also was
indicated in models using husbands' exhaled tetrachloroethylene (rate ratio = 0.94, 95% CI:
0.85-1.04) and husbands' exposure index (rate ratio = 0.90, 95% CI: 0.78-1.03). The latter two
exposure indices may not have estimated exposure during the sensitive window for the index
pregnancy, however. The small sample size resulted in CIs that included the null hypothesis.
The authors noted that to detect a halving of risk for pregnancy with 80% power (a = 0.05), over
50 women per group would have been required.
A Danish case-control study of couples examined or treated for infertility during
1977-1980 reported evidence of idiopathic infertility among women reporting exposure to dry-
cleaning chemicals (Rachootin and 01 sen. 1983). Controls were couples with a healthy child
born at the same hospital during 1977-1979. Information about occupational and reproductive
history was obtained from 87% of both cases and controls who returned a mailed, self-
administered questionnaire during November 1980 to May 1981. Participants were defined as
exposed if they reported contact with any of 15 types of chemicals and physical agents
(including dry cleaning) and three specific work processes a minimum of once per week for at
least 1 year in the period prior to hospital admission. The medical records of infertile couples
were reviewed by a collaborating physician who had no knowledge of exposures. Three analytic
approaches were used to evaluate subgroups of couples with a medical history anticipated to be
related to occupational exposures. Reported exposure to dry-cleaning chemicals was associated
with idiopathic infertility among women compared to fertile couples with a healthy child
conceived within 1 year (OR: 2.7, 95% CI: 1.0-7.1). The statistical method was not described,
but the authors stated that the odds ratio was adjusted for the women's age, education, residence,
and parity. Cases and controls lived within the catchment area of the hospital. Exposure to dry-
cleaning chemicals was not associated with sperm abnormalities or idiopathic infertility among
male partners or with hormonal disturbances among women. The odds ratio for idiopathic
infertility among women with exposure to dry-cleaning chemicals also was increased when
couples who had been infertile for at least 1 year were compared to other infertile couples with
conditions believed to be unrelated to occupational exposures (crude odds ratio [ORc] = 1.8,
95% CI: 0.5-5.8). A third analysis involved comparison within the control group; couples who
experienced a delay in conception of more than 1 year but who gave birth to a healthy child were
compared to couples who conceived a healthy child in less than 1 year. Again, women reporting
exposure to dry-cleaning chemicals had an increased odds ratio for delayed conception
(ORc: 1.6, 95% CI: 0.9-2.9). Although two of the risk estimates did not reach statistical
significance, all three were greater than 1.5. The consistent increased odds ratios observed using
three different comparison groups suggest an effect of exposure to dry-cleaning chemicals on
conception among women. The study evaluated a large number of chemicals and physical
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-309 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
exposures. The authors did not present the number of cases by subgroup, or the number of
controls who reported exposure to dry-cleaning chemicals, so it is difficult to assess the impact
of sample size on the precision of the effect estimates. Other chemical exposures, as well as
noise, also were associated with idiopathic infertility among the women. In addition, the
statistical analyses for dry-cleaning chemicals did not control for exposure to other chemical or
physical agents.
Sallmen et al. (1995) conducted a retrospective time-to-pregnancy study among Finnish
women biologically monitored at the Institute of Occupational Health in 1965-1983 for one or
more of six solvents (styrene, toluene, xylene, tetrachloroethylene, trichloroethylene, and
1,1,1-trichloroethane). This study was an extension of an investigation of the risk of
spontaneous abortion in the same study population. That study is described later in this section
(Lindbohm et al.. 1990). Pregnancies and their outcomes (live birth, spontaneous abortion, or
fetal loss) between 1973 and 1983 among the women had been identified using a national
register of pregnancies in Finland and the Finnish Register of Congenital Malformations. Time-
to-pregnancy information was obtained through questionnaires mailed to 355 women who were
the cases and controls in the previous study. Information about exposure during the preceding
12 months before each woman's pregnancy began was collected. The response rate was 66%,
and the final data set contained 197 women who had been attempting to become pregnant, had
no other risk factors for infertility, for whom complete information was available on exposure
and time-to-pregnancy. Time-to-pregnancy was defined as the number of menstrual cycles
required to become pregnant and is a measure of fertility, the per cycle probability of conceiving
a clinically detectable pregnancy. Increased time-to-pregnancy can indicate a loss during
pregnancy during any stage from gametogenesis to fertilization to the clinical stage of
pregnancy, including early stage spontaneous abortions.
The same exposure-assessment procedure as was used in the previous study was adopted
for this study, and if the subject reported working in the same job, their previous exposure
classification was used. Self-reported work tasks during the 12 months prior to conception were
assigned to an exposure classification by likelihood and level of exposure for 84 women whose
jobs or exposures were different than reported previously for the first trimester. Classifications
were made without knowledge of reproductive history and were checked by an independent,
experienced industrial hygienist. The three categories for likelihood of exposure were not
exposed, potentially exposed, and exposed. Subjects were grouped according to high (n = 46),
low (n = 59), and none (n = 92) for level of exposure (see description of Lindbohm et al.. 1990).
Exposure to organic solvents during their time-to-pregnancy was reported by more than
one-half of the women (105 out of 197). Incidence density ratios, indicating the likelihood that
exposed women will achieve a clinical pregnancy during the fertile period in each menstrual
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-310 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
cycle class (e.g., 1st menstrual cycle, 2nd, 3rd and 4th, 5th, 6th, etc.) compared to an unexposed
woman, were estimated using discrete proportional hazards regression. Incidence density ratios
(IDRs) were reported for women exposed to tetrachloroethylene (n = 20) or working in dry
cleaning (n = 17). Compared to women with no exposure, the IDRs for low and high exposure
were 0.63 (95% CI: 0.34-1.17) and 0.69 (95% CI: 0.31-1.52), respectively. The statistical
models controlled for exposure to other solvents, recent contraceptive use, and age at menarche.
For workers in dry cleaning, the IDR for 11 women with low or high exposure combined was
0.44 (95% CI: 0.22-0.86) and for 6 women with high exposure was 0.57 (95% CI: 0.24-1.34).
These models controlled for low and high exposure to solvents in other industries, recent use of
IUD/spermicides, and age at menarche. The model for high exposure also adjusted for low
exposure to organic solvents. The authors noted that only 1 of the 11 women who worked in dry
cleaning reported exposure to other solvents in addition to tetrachloroethylene. These results
suggest that exposure to tetrachloroethylene may affect fecundability, however, because the
focus was on a broad range of solvent exposures and industries, the sample size for assessing
tetrachloroethylene was small, and statistical precision was low. However, the study had several
strengths, including collection of detailed work histories. Exposure classifications were based on
the frequency of solvent use, not just reported use ever or job title. In addition, several potential
confounders were assessed, and statistical models controlled for exposure to other solvents. It
was not clear if the models for individual solvents were assessed for confounding by case status
(i.e., pregnancy ended in a spontaneous abortion). However, reduced fecundability was
associated with exposure to organic solvents combined in separate analyses of cases and
controls. The low response rate overall, and evidence that response was higher among cases and
exposed controls, particularly those with lower parity, raises the possibility of selection bias.
Time-to-pregnancy also was evaluated among the wives of men exposed to organic
solvents and monitored by the Finnish Institute of Occupational Health during 1965-1983
(Sallmen et al.. 1998). This was an extension of an earlier case-referent study of risk of
spontaneous abortion (see description of Taskinen et al.. 1989 later in this section). The
investigators used a similar approach as that used in Sallmen et al. (1995). described above.
Cases (n = 110) and referents (n = 332) that participated in Taskinen et al. (1989) were recruited.
Time-to-pregnancy information was obtained through questionnaires mailed to 355 women who
were the cases and controls in the previous study. A detailed history of occupation and work
tasks during the year the pregnancy started had been obtained from the husbands in the previous
study. A similar history was now requested of the wives, focusing on the preceding 12 months
before the pregnancy. The response rate was 72%, and the final data set contained 282 women
who had been attempting to become pregnant, had no other risk factors for infertility, for whom
complete information was available on exposure and time-to-pregnancy. The same exposure-
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-311 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
assessment procedure as was used in the previous study was adopted for this study, and if the
subject reported working in the same job at the beginning of the pregnancy, their previous
exposure classification was used. A new exposure classification was required only for nine men
whose jobs or exposures were different than reported previously for the first 3 months before the
pregnancy began. Classifications were made without knowledge of reproductive history and
were checked by an independent, experienced industrial hygienist (Taskinen etal.. 1989). The
three categories for likelihood of exposure were, not exposed, potentially exposed, and exposed.
Subjects were grouped according to high/frequent (n = 141), intermediate/low (// = 80), and
unexposed (n = 61) for level of exposure to organic solvents during the time-to-pregnancy
period.
Incidence density ratios (IDRs) were reported for exposure to all organic solvents
combined and for specific solvents. The IDRs for low (n = 9) and combined intermediate/high
(n = 8) exposure to tetrachloroethylene were 0.86 (95% CI: 0.4-1.84) and 0.68 (95%
CI: 0.30-1.53). The discrete proportional hazards regression models were adjusted for short
menstrual cycle, long or irregular menstrual cycle, older age at menarche, frequency of
intercourse, maternal age, maternal exposure to organic solvents, and a variable for missing
information. Fecundity appeared most reduced among the wives whose husbands had a high
level and/or frequency of tetrachloroethylene exposure compared to low or no exposure.
However, the study was limited by low statistical precision because of small sample size. Time-
to-pregnancy information and exposures were collected 8 to 18 years after the pregnancy of
interest, which likely resulted in some misclassification. It is less likely that recall bias affected
the risk estimates because the exposures were assigned based on information collected for the
earlier study of spontaneous abortion.
Among studies evaluating effects of tetrachloroethylene on reproduction and
development, the majority of studies assessed effects on risk of spontaneous abortion. These
studies defined spontaneous abortion as a fetal loss prior to 20-28 weeks gestation, although one
study included all fetal loss during the first 6 months of pregnancy (Lagakos et al.. 1986).
Several studies included only clinically recognized spontaneous abortions reported in birth
registers (Ahlborg. 1990b: Kyyronen et al.. 1989: Lindbohm et al.. 1991: Lindbohm et al.. 1990:
McDonald et al.. 1986: McDonald et al.. 1987: Olsen et al.. 1990: Taskinen et al.. 1989:
Windham et al.. 1991). while some included spontaneous abortions reported by participants
(Aschengrau et al.. 2008: Aschengrau et al.. 2009a: Bosco et al.. 1987: Doyle et al.. 1997:
Eskenazi et al.. 1991b: Lagakos et al.. 1986). It should be noted that it is not possible to identify
all spontaneous abortions that occur in populations because a woman may not recognize very
early events and/or may not seek treatment.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-312 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
McDonald et al. (1986; 1987) conducted a large survey of occupation and reproductive
outcomes among 56,012 women in 11 large obstetrical units in Montreal, Canada, over a 2-year
period from May 11, 1982 to May 10, 1984. Interviews were conducted with 51,885 women
with a term delivery and 4,127 women treated in the hospital for a spontaneous abortion, defined
in this study as a fetal loss <28 weeks of gestation. The 11 hospitals included in the survey
treated approximately 90% of all births in Montreal. As part of the interview, women were
asked to describe all previous pregnancies that ended in a spontaneous abortion, and 10,910 were
identified. Interviews were completed for 90% of the women with term births, and 75% of
women admitted for a spontaneous abortion. Information also was collected about occupation at
the time of conception for the current and any previous pregnancies. Nine occupational groups
in the Canadian Classification and Dictionary of Occupations were reduced to six major
groupings and included 42 categories that the investigators concluded were homogenous.
Logistic regression was used to evaluate risk of spontaneous abortion for five nonoccupational
factors: maternal age, parity, history of a previous abortion, smoking habit, and highest
educational level reached, and the expected number of spontaneous abortions for each
occupational group was calculated. The ratio of observed to expected numbers was evaluated for
each occupational group. Among women in the laundry and dry-cleaning occupational grouping,
there were 8 spontaneous abortions out of 100 recent pregnancies (O/E: 1.18; p > 0,1 [CI not
reported]) and 31 out of 123 previous pregnancies (O/E: 1.02). Subsequent analysis of the data
included women who worked at their jobs for at least 30 hours weekly at the beginning of
pregnancy (McDonald et al.. 1987). In this analysis, 36 combined current and previous
spontaneous abortions were observed out of 202 pregnancies. An O/E ratio of 1.05 (p > 0.1; CI
not reported) was reported. The expected number was determined from a logistic regression
model of spontaneous abortion risk including maternal age, parity, history of a previous abortion,
smoking habit, and alcohol consumption. This study is not very informative regarding
tetrachloroethylene risk because the group of dry-cleaners and laundry workers likely included
individuals with no exposure to the solvent.
A case-referent study of adverse pregnancy outcome was conducted among the wives of
male workers who had been monitored for organic solvents by the Finnish Institute of
Occupational Health between 1965 and 1983 (Taskinen et al.. 1989). The cohort included men
in their first marriage during 1985 with wives who were 18-40 years old at the end of the 1st
trimester of pregnancy. Pregnancies and outcomes were identified through national registers.
Eligible pregnancies began during the marriage or up to 9 months before. Cases were defined as
wives with a spontaneous abortion (if multiple, one randomly selected) or a congenitally
malformed child. Referents were selected from wives with a healthy birth between 1973 and
1983 (1:3 for spontaneous abortions, 1:5 for malformations), age matched within 30 months. A
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-313 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
total of 136 of 172 selected cases (79.1%) and 370 of 505 selected referents (73.3%) responded
to a questionnaire mailed in January 1986. Only pregnancies that were identified in the register
and reported by participants were included. Because of this, and because a matched response
was required, the final data set included 120 cases and 251 referents. Information on occupation
and exposure to solvents during the year of conception was requested of the men. Information
on occupational and other exposures during the first trimester of pregnancy was solicited from
the wives. Exposure classifications were made blind to pregnancy outcome. Solvent exposure
for the men was assessed for an 80-day period that preceded the pregnancy, the relevant period
of spermatogenesis, using information on occupation, job description, reported solvent or other
chemical use, and biological monitoring data. Workers were classified as not exposed if work
tasks did not include handling solvents and no exposure was reported, and no biological
measurement for a particular solvent was made. Workers were classified as potentially exposed
if work tasks might have involved solvent use, but use was not reported by the worker, and no
biological measurements for a particular solvent were made. Workers were classified as exposed
if biological measurements for a solvent were taken while at the same job for the reported
pregnancy, reported tasks implied solvent exposure, or solvent exposure was reported. Exposure
was categorized into none, low, intermediate, or high. Workers with high exposure handled
solvents daily, or their biological measurements were above the reference value for the general
population. Workers with intermediate exposure used solvents 1-4 days per week, and
biological measurements indicated intermediate or low exposure. Workers with low exposure
handled solvents <1 day per week. All other scenarios were classified as no exposure.
A spontaneous abortion rate of 8.8% was observed among all recognized pregnancies, a
rate within the range reported for Finland between 1973 and 1983 (Lindbohm and Hemminki.
1988). The unadjusted odds ratio for risk of spontaneous abortion in relation to likely paternal
exposure to tetrachloroethylene was 0.5 (95% CI: 0.2-1.5) using conditional logistic regression.
Likely exposure was assigned to 4 cases and 17 referents. Adjusted odds ratios controlling for
potential paternal exposure to the solvent, likely paternal exposure to other organic solvents and
dusts, maternal exposure to solvents, maternal heavy lifting, and history of previous spontaneous
abortion were presented only for likely exposure to all halogenated hydrocarbons. In addition to
exposure to tetrachloroethylene, this group included exposure to trichloroethylene and
1,1,1-trichloroethane. The adjusted odds ratios for low/rare, intermediate, and high/frequent
exposure were 1.1 (95% CI: 0.5-2.6), 1.3 (95% CI: 0.5-3.1), and 0.8 (95% CI: 0.3-2.2),
respectively. The exposure assessment encompassed a broad range of solvents, and only a small
number reported exposure to tetrachloroethylene. In addition, exposure to multiple chemicals
was possible for much of the cohort, and this was not controlled for in the chemical-specific
models.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-314 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
A subsequent study of paternal occupational exposure and spontaneous abortions
attempted to identify all medically recognized pregnancies (spontaneous abortion, induced
abortion, and healthy births) between 1973 and 1982 through the Finnish nationwide Hospital
Discharge Register and from outpatient hospital clinics (Lindbohm et al.. 1991). Information on
occupation was obtained from 1975 and 1980 national census records. Pregnancies during 1973
to 1978 were linked to the 1975 Census, and pregnancies during 1979 to 1982 were linked to the
1980 Census. A job-exposure classification, developed in cooperation with two industrial
hygienists, assigned chemical exposures commonly used by job groups within industries.
Exposures were assigned to job groupings using a list of 78 exposures, including specific
substances, mixtures, and nonspecific exposures, plus industrial hygiene measurements made by
the Institute of Occupational Health and the Finnish register of employees occupationally
exposed to carcinogens. Exposure assessment focused on mutagens, and three levels were
defined: moderate/high, potential/low, and none.
The susceptible exposure period of interest was an 80-day period prior to conception
corresponding to spermatogenesis. Because the investigators did not have temporally resolved
exposure information, pregnancies that were terminated during a 2-year period close to the
census were selected (January 1, 1976-December 31, 1977 for the 1975 Census, and May 1,
1980-April 30, 1982 for the 1980 Census). A total of 99,186 pregnancies to women aged
12-50 years with complete information about occupation, industry, and socioeconomic status
occurred during these time periods. There were three spontaneous abortions among the wives of
men with moderate or high exposure to tetrachloroethylene (out of 45 pregnancies). The odds
ratio was 0.7 (95% CI: 0.2-2.4) in a linear logistic regression model adjusting only for age. This
large occupational survey was meant to evaluate reproductive risks associated with paternal
exposures to a wide array of substances and mixtures believed to be mutagens. While the focus
was on exposure to mutagens as a whole, specific exposures also were analyzed, and a broad
2-year time period was used to identify pregnancies related to occupation listed in the 1975 or
1980 censuses. The nonspecific exposure window and use of a crude exposure assignment
method based on occupational title in a census, along with the small number of cases, limit the
ability to draw conclusions concerning paternal tetrachloroethylene exposure and risk of
spontaneous abortion.
A case-control study in Finland evaluated the association of medically diagnosed
spontaneous abortions and maternal occupational exposure to specific solvents (Lindbohm et al..
1990). The sampling frame was a database of women biologically monitored at the Institute of
Occupational Health in 1965-1983 for one or more of six solvents (styrene, toluene, xylene,
tetrachloroethylene, trichloroethylene, and 1,1,1-trichloroethane). Pregnancies and their
outcomes between 1973 and 1983 among the women were identified using a national register of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-315 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
pregnancies in Finland and the Finnish Register of Congenital Malformations. Cases were
women with a spontaneous abortion recorded in the database. One to three controls per case
were selected from among women with a live birth (congenital malformations were not included)
matched for age (± 2.5 years). Among the 456 women, overall response to a mailed
questionnaire was 85% for both cases and controls. A lower proportion of cases (78%) than
controls (99%) confirmed the pregnancy selected from the register. The final data set contained
73 cases and 167 controls with complete information about their occupational history and solvent
exposures during the first trimester of pregnancy.
Likelihood and level of exposure to specific solvents was determined by two
investigators blind to pregnancy outcome using responses to the questionnaires and biological
measurements when available. Women were defined as not exposed if work tasks did not
include handling solvents, the worker did not report exposure, and no biological measurements
were available. Women were defined as potentially exposed if work tasks might have involved
solvent use, but exposures were not reported by the worker, and no biological measurements
were available. Women were defined as exposed if biological measurements were taken while at
the same job, reported tasks implied solvent exposure, or solvent exposure was reported. The
level of exposure was categorized into none, low, or high. High exposure involved handling
solvents daily or 1-4 days per week and high-recorded concentrations for biological or available
industrial hygiene measurements. Low exposure involved handling solvents 1-4 days per week
with low biological concentrations, or solvents were handled <1 day per week. All other
exposure scenarios were defined as none. Biological measurements during the first trimester
were available for only 5% of the population, and, therefore, exposure assignments were based
primarily on reports of work tasks and reported solvent use. Exposure classifications were
checked by an experienced industrial hygienist.
Among the exposed women, there were 8 cases and 15 controls with exposure to
tetrachloroethylene. An odds ratio of 1.4 (95% CI: 0.5-4.2) was observed using conditional
logistic regression with adjustment for previous spontaneous abortions, parity, smoking, use of
alcohol, and exposure to other solvents. The adjusted odds ratios for low and high exposure
were 0.5 (95% CI: 0.1-2.9) and 2.5 (95% CI: 0.6-10.5), respectively. Among four cases and
five controls who reported tetrachloroethylene exposure and whose work tasks involved dry
cleaning, the odds ratio for spontaneous abortion, controlling for exposure to other solvents, was
2.7 (95%) CI: 0.7-11.2). The odds ratio for women who reported tetrachloroethylene exposure
but who conducted other work in dry cleaners (1 case and 6 controls) was 0.6 (95% CI: 0.1-5.5).
Blood tetrachloroethylene measurements taken closest to the pregnancy were available for six
women who worked in dry cleaning and seven women in other occupations. The mean
concentration was higher among dry cleaners (2.11 [j,mol/L versus 0.43 (j,mol/L). The authors
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-316 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
reported that the proportion of study subjects who did not report exposure to a specific solvent in
contrast to a biological measurement that indicated that they were exposed was 18% among
cases and 20% among controls, suggesting that recall was not different by exposure. The study,
which is limited by small sample size and a low prevalence of exposure to tetrachloroethylene,
suggests that exposure during the first trimester may increase risk of spontaneous abortion.
Moreover, odds ratios increased in size when the analysis was restricted to more homogenous
exposure groups representing high exposures.
A case-control study in Santa Clara County, California, also focused on occupational
exposure to solvents, including tetrachloroethylene (Windham et al.. 1991). Selection of cases
was hospital based; spontaneous abortions, defined in this study as <20 weeks gestation, among
women 18 years of age or older that occurred between June 1986 to February 1987 were
identified through records of pathology specimens submitted to the 11 hospital laboratories
located in the county. Investigators reviewed medical charts to differentiate spontaneous
abortions from induced abortions. Controls, two per case, were randomly selected from women
with live births, frequency matched by last menstrual period and hospital. A total of 697 of
772 eligible cases (90.3%) and 1,359 of 1,485 controls (91.5%) participated. The analysis was
limited to 1,361 women who were employed during their pregnancy. A higher proportion of
cases was over 35 years of age, reported a prior fetal loss, and consumed more alcohol per week.
Information on exposure during the first 20 weeks of pregnancy or for the duration of the
pregnancy for cases was obtained through a computer-assisted telephone interview. The women
provided detailed information about industry and occupation, job tasks and use of 10 solvents,
plus reported exposure to any other solvents or degreasers. Among the women who reported that
they used tetrachloroethylene during the first weeks of pregnancy, 5 were cases, and 2 were
controls (ORc: 4.7, 95% CI: 1.1-21.1, calculated using Haldane's method for small samples).
Unexposed participants reported no use of any named solvents and did not work in the
microelectronics industry (n = 847). Four of the women exposed to tetrachloroethylene also
reported use of trichloroethylene. The unadjusted odds ratio for use of tetrachloroethylene
and/or trichloroethylene was 3.4 (95% CI: 1.0-12.0). Odds ratios also were calculated in
stratified analyses using Mantel-Haenszel estimation for each of six dichotomous variables
individually (age, race, education, prior fetal loss, smoking, and hours worked). This limited
evaluation of potential confounding does not appear to have resulted in a large decrease of the
summary odds ratios compared to the crude odds ratio, although the adjusted odds ratios were
presented only as a range (e.g., 4.2 [95% CI: 0.86-20.2] controlling for hours worked to 6.0
[95% CI: 1.4-25.8] controlling for age). Estimated risk increased with a higher level or intensity
of exposure when the analyses were stratified by whether exposed participants reported
symptoms, skin contact, or odor versus none (6,3, /;-value for Fisher exact test (1 -tail) = 0.04
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-317 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
compared to 2.1, />-value = 0.54). Despite the small numbers with tetrachloroethylene exposure,
the results suggest an elevated risk for spontaneous abortion. However, several of the exposed
women also were exposed to other solvents, including trichloroethylene, and a detailed
evaluation of potential confounding was precluded by small numbers.
One of the first studies to evaluate adverse reproductive outcomes, including spontaneous
abortions, stillbirths, birth defects, and low birth weight, among female dry cleaners evaluated 53
of 66 small establishments (40 dry cleaning and ironing and 13 ironing only) in two
neighborhoods in Rome, Italy (Bosco et al.. 1987). The study population included all of the 67
women who worked in the participating shops. The women averaged 43 years of age and had
been employed an average of 20 years. Information on the work setting and operations and
reproductive histories were collected through a standardized interview. Participants reported if
they had worked in dry cleaning, as a housewife, or other job prior to and during their
pregnancies. In addition, a 24-hour urine sample was collected on a Friday at the end of the
workweek from 53 of the women. Trichloroacetic acid concentrations were higher among
40 dry cleaners (5.01 |ig/L) compared to 13 ironers (1.35 (J,g/L) and 5 controls (1.56 [ig/L). Of
56 pregnancies reported during employment as a dry cleaner, 5 ended in a spontaneous abortion
(8.9%). One spontaneous abortion was reported among the 46 pregnancies that occurred while
the women were working at home. The fourfold higher incidence of spontaneous abortion
suggests a tetrachloroethylene-related risk among the dry cleaners. However, individual
characteristics and behaviors that may pose a risk of spontaneous abortion were not presented by
exposure status during pregnancy, and potential confounding was not assessed in this very small
study.
A study that used a common protocol to evaluate reproductive outcomes among dry
cleaners in Denmark, Finland, Norway, and Sweden employed a more precise definition of
tetrachloroethylene exposure (Olsen et al.. 1990). All women who had worked at identified
laundries and dry-cleaning plants for at least 1 month during 1973-1983 were included, and a
nested case-referent study was conducted in each country. Identification numbers were linked to
national birth registers and hospital discharge registers to obtain information on births and
outcomes, including spontaneous abortions, in the cohort. In Denmark, all women in the cohort
and every pregnancy that occurred during the study period were included. In Sweden and
Finland, two and three controls per case, matched on maternal age (±2 years), year of pregnancy,
and parity (for Denmark and Sweden), were selected from women with a healthy newborn. In
Norway, information on spontaneous abortions was not available. Women were identified
through company records of active dry-cleaning plants (Sweden and Denmark) and laundries
(Sweden). Approximately 62 and 74% of dry-cleaning plants in Sweden and Norway
participated, respectively. The final study sample consisted of 31 spontaneous abortions and 53
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-318 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
referents in Sweden (84% response) and 10 spontaneous abortions and 119 referents in Denmark
(77.3% response). In Finland, laundry and dry-cleaning workers on the rolls of the Union of
Chemical Workers and the Municipal Workers of Finland and or included in payroll data from
employers for 1973-1983 were identified and linked with the nationwide hospital discharge
register and polyclinic data for information on pregnancies. One pregnancy for each woman was
randomly selected for analysis. The final data set included 118 spontaneous abortions and
264 referents (77.2% response). Information on exposure to tetrachloroethylene was obtained
from the interviews and questionnaires and was classified by an industrial hygienist blinded to
pregnancy status (Sweden and Denmark). The Finnish investigators had more detailed
information and used reported work history and exposure frequency to classify exposure status.
Exposure was categorized into three groups: unexposed (no dry cleaning), low (worked in dry
cleaning but not high exposures), and high (workers who conducted dry cleaning or spot removal
for at least 1 hour per day). Risk of spontaneous abortion in relation to exposure during the first
trimester was analyzed using conditional logistic regression for matched Swedish and Finnish
data, and unconditional logistic regression for the Danish data set. Models were adjusted for
parity, smoking, and alcohol consumption (Sweden and Finland only).
Odds ratios greater than 1 were observed for the high exposure group in Denmark (OR:
2.52, 95%) CI: 0.26-24.1) and Finland (OR: 4.53, 95%> CI: 1.11-18.5). The high exposure group
contained small numbers of cases and controls with one case each in Sweden and Denmark, and
six cases in Finland. The odds ratios were combined using the inverse variance of the odds ratio.
The odds ratios for low and high exposure (95% CI) were 1.17 (0.74-1.85) and 2.88
(0.98-8.44), respectively. The authors stated that similar results were obtained when exposure
information provided by the employers (55% of sample) was used instead of responses from the
participants.
A separate report of the Finnish study population was published, evaluating 130 cases of
spontaneous abortions and 289 controls matched for maternal age (Kvvronen et al.. 1989).
Slightly different categorizations were used to define exposure. High exposure included women
whose tasks included dry cleaning at least 1 hour daily, and who handled tetrachloroethylene at
least once a week (n = 15). Low exposure included women whose work tasks involved pressing
at a dry cleaners or spot removing, or who handled tetrachloroethylene less than once a week
(n = 31). Blood tetrachloroethylene measurements, taken within 10 months of the first trimester
of pregnancy, were available for seven of the participants (except for one more distant
measurement). These data corresponded well to their reported exposure. Exposure to other
solvents, including petroleum, benzene, toluene, acetone, thinner, and spot remover mixtures,
was reported by six cases (5.9% of women who worked during their pregnancy) and six controls
(2.9%). The odds ratio for high exposure to tetrachloroethylene was 3.4 (95% CI: 1.0-11.2,
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-319 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
p < 0.05) in a multivariate model adjusted for frequent use of solvents other than
tetrachloroethylene (OR: 1.5, 95% CI: 0.4-5.4), frequent heavy lifting at work (OR: 1.9;
95% CI: 1.0-2.8), and frequent use of alcohol (OR: 2.0, 95% CI: 1.0-4.0). Selection bias did
not appear to be a major factor; when exposure information obtained from employers was used
to classify eight cases and six controls instead of self-reports, the proportion returning the
questionnaire was similar (0.25 and 0.17, respectively).
Ahlborg et al. (1990b) published the Swedish results separately along with a
complementary study designed to be more representative of the entire dry-cleaning and laundry
sector. Laundry and dry-cleaning establishments, identified from the Swedish Post Address
Register in 1984, were mailed a questionnaire to obtain names and contact information for all
women employees who had worked for at least 1 month during 1974 and 1983. Cases of
spontaneous abortion (defined in this study as fetal death at <28 weeks gestation), perinatal
death, congenital malformation, or low birth weight (<1,500 g) were identified among deliveries
during 1974-1983 recorded in the Medical Birth Registry, the Swedish Registry of Congenital
Malformations, and the Inpatient Registry for Somatic Care (spontaneous abortion treated in a
hospital). Dates of delivery or spontaneous abortion were used to identify women who had been
working while they were pregnant (at least 1 week of the year before delivery or 6 months before
a spontaneous abortion). A total of 67 cases were identified among 955 pregnancies, and two
referents per case were selected, matched on mother's age, year of pregnancy, and parity (only
for deliveries). Responses were received from 48 cases {15%) and 110 referents (88%).
Recruitment for the complementary study involved the identification of women registered as
washers/cleaners via an occupational code in the 1975 and 1980 Censuses. A total of
755 additional pregnancies were identified via linkage with the medical registers for the 2-year
period after each census. Responses to the mailed questionnaire were received from 68 cases
(88%>) and 131 referents (87%). Exposure to tetrachloroethylene during the first trimester was
classified independently by two investigators who were unaware of the worker's case/control
status. High exposure included operating a dry-cleaning machine or conducting spot removing
using tetrachloroethylene at least 2 hours per week, or ironing/pressing dry-cleaned cloth for
over 20 hours per week, or cleaning and filling the machines at least three times. Low exposure
included other work in dry-cleaning businesses where tetrachloroethylene was used. Unexposed
workers were employed in companies that did not dry clean using tetrachloroethylene. In the
combined data set, 31 and 19 cases (all outcomes) were classified as having low and high
exposure, respectively. The numbers of spontaneous abortions by exposure category were not
reported. Odds ratios for spontaneous abortion among workers with low and high exposure
using conditional logistic regression were 1.0 (95% CI: 0.4-2.2) and 0.9 (95% CI: 0.4-2.1),
respectively. The models adjusted for smoking, alcohol consumption, medical complications,
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-320 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
and history of adverse pregnancy outcome. This study did not find an increased risk of
spontaneous abortion among workers reporting tetrachloroethylene exposure during the first
trimester.
A relatively large study in the United Kingdom evaluated the risk of spontaneous
abortions among current and former employees of dry-cleaning and laundry establishments
managed by four companies between 1980 and 1995 (Doyle et al.. 1997). Information about
workplace exposure and reproductive history were obtained in 1995-1996 via mailed
questionnaires sent to 7,301 women, aged 16-45 years, who were identified by the employers.
Of the 5,712 questionnaires successfully delivered, 54.5% were completed (n = 3,110). The
responses by current dry-cleaners and laundry workers were 78 and 65%, respectively, but were
lower among former workers (46.1 and 39.7%, respectively). The authors reported that the age
distribution of responders was comparable to that of nonresponders. The final data set included
3,092 respondents with complete information about 3,517 total pregnancies. Pregnancies were
included in the analysis if the women reported that it had been confirmed by a doctor, hospital
treatment was required, or it ended in a live birth. The rate of spontaneous abortions was
evaluated in relation to the woman's employment during her pregnancy or the 3 months prior.
Work at a dry cleaner and as a dry-cleaning machine operator was used as an exposure surrogate
for tetrachloroethylene. This was compared to work at a laundry or no employment at a dry
cleaners or laundry during the pregnancy or 3 months prior.
Spontaneous abortions were compared to total pregnancies (spontaneous abortions,
stillbirths, and live births) excluding ectopic and molar pregnancies and induced abortions. For
the 325 reported spontaneous abortions between 1980-1995, the odds ratio for dry cleaning
compared to laundry work was 0.97 (0.55-1.69). However, among 93 spontaneous abortions to
women employed in dry cleaning, machine operators had a 63% higher risk of spontaneous
abortion compared to nonoperators (OR: 1.63, 95% CI: 1.01-2.66). The unconditional logistic
regression models controlled for maternal age, pregnancy order, and year of birth. A similar
pattern of risk was observed when the analyses were restricted to the women's first or last
pregnancies. These latter analyses were meant as a check to address the lack of independence of
multiple pregnancies reported by the same woman. For example, among dry-cleaning machine
operators, when the last exposed pregnancy was compared to pregnancies that occurred later
during periods with no exposure to tetrachloroethylene, risk of spontaneous abortion was 82%
higher (OR: 1.82, 95% CI: 1.09-3.05). An elevated risk also was observed when pregnancies
during work as a dry-cleaning machine operator were compared to unexposed pregnancies before
the first exposed pregnancy. Laundry workers also experienced more spontaneous abortions
when employed in laundries compared to periods when they had other employment or were not
employed; however, the CIs included one. The investigators were not able to compare risks
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-321 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
between dry cleaning generally and laundry work because the number of spontaneous abortions
reported for pregnancies while working in a laundry was low (n = 19). Doyle et al. (1997) found
an elevated risk of spontaneous abortion for work as a dry-cleaning machine operator during or 3
months before a pregnancy compared to work in other dry-cleaning jobs or work in other
industries or in the home during this sensitive period.
The rate of self-reported spontaneous abortions was comparable among the wives of dry
cleaners (n= 14) and laundry workers (n = 26) in a cohort of primarily unionized men in
northern and southern California who participated in a study of semen quality (Eskenazi et al..
1991b). Rates of spontaneous abortion during the time periods when their husbands worked in
the industry were 11.1 and 15.2% among the wives of dry-cleaners and laundry workers,
respectively (X2 = 0.32, p = 0.57). Although the authors presented the rates as spontaneous
abortion rates, it does not appear that the fetal deaths reported were limited to <28 weeks of
gestation. The rate was calculated as the total number of miscarriages during the husband's
employment in the industry divided by the total number of pregnancies during the same time
period, multiplied by 100. It was not stated how many years the women, whose average age was
midthirties, had to recall previous miscarriages.
A population-based study in Woburn, Massachusetts, evaluated outcomes during
pregnancy and effects in children among residents whose drinking water source was two wells
contaminated with chlorinated organic substances from 1960 to 1982 (Lagakos et al.. 1986). The
two wells were operated as a single water source. The contamination of the two wells, located in
eastern Woburn, was discovered in May, 1979. Levels of trichloroethylene (267 ppb),
tetrachloroethylene (21 ppb), and chloroform (12 ppb) were detected, and the wells were shut
down. The other six wells that supplied Woburn were located in the southwest part of town, and
testing did not find levels above state and federal standards. Information was collected through a
telephone survey of former and current family members residing in Woburn from 1960-1982
and listed in the 1982 town directory. The survey was conducted by 235 volunteers trained in
interview techniques who successfully contacted 6,219 residences. In the end, 5,010 completed
interviews were obtained, approximately 57% of the town's residences with listed telephone
numbers. All pregnancies ending between 1960 and 1982 to women born since 1920 were
ascertained, and information was collected on pregnancy outcomes and the health of offspring,
maternal characteristics, and residence history. Regional and temporal distribution of the water
from the two contaminated wells was determined by the Massachusetts Department of
Environmental Quality and Engineering during October 1964 to May 1979. The town was
partitioned into five zones of graduated exposure to water from the wells. The study
investigators estimated the proportion of each household's annual water supply that came from
the two wells. Each pregnancy was assigned an annual exposure score using the mother's
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-322 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
residence during the year the pregnancy ended. An exposure history was constructed for each
child consisting of the sum of annual scores accumulated during their residence in Woburn.
Of the 4,396 pregnancies that occurred during 1960 to 1982, 16% were exposed during
the year the pregnancy ended. There were 520 spontaneous abortions (12%), defined in this
study as a fetal loss in the first 6 months, and 67 perinatal deaths (1.5%), defined as a stillbirth or
a live birth that survived fewer than 7 days. Logistic regression analyses, controlling for other
risk factors, found no statistically significant associations between the annual exposure score for
the year a pregnancy ended and spontaneous abortion, or perinatal deaths before 1970. An odds
ratio of 10 (p = 0.003) was observed for perinatal deaths after 1970, when changes in industrial
water demand occurred, and a different set of five zones representing exposure to water from the
contaminated wells was constructed. This was due to 3 perinatal deaths that occurred in
households with the highest exposure score category of 0.51-1.0.
A population-based retrospective study of tetrachloroethylene in drinking water evaluated
effects on pregnancy and development from exposure resulting from leaching of
tetrachloroethylene from vinyl linings in water distribution pipes installed between 1968 and
1980 in the Cape Cod region in Massachusetts (Aschengrau et al.. 2008; Aschengrau et al..
2009a; Aschengrau et al.. 2009b). Because the pipes were used to replace existing pipes or to
extend the distribution system to serve a growing population, population exposure was
irregularly distributed, and a wide range of tetrachloroethylene concentrations were detected in
samples collected in 1980. In addition, only one town used a chlorinated surface water supply,
resulting in a low probability that drinking water was contaminated with chlorinated byproducts.
Water concentrations ranged from 1.5 to 80 |ig/L along main streets, and from 1,600 to 7,750
[j,g/L along dead end streets where water flow was low. All births between 1969 and 1983 were
identified from birth certificates, and women residing in one of eight Cape Cod towns with vinyl-
lined water distribution pipes at the time of the index birth were eligible for the study. A total of
1,492 women with addresses along streets where the pipes had been installed or with connections
to such pipes were initially defined as exposed. A comparison group of 1,704 births, frequency
matched to the exposed group by month and year of birth, was selected. Follow-up of the
selected individuals occurred during 2002-2003. The final data set contained 959 women with
potential exposure and 1,087 potentially unexposed women who returned a self-administered
questionnaire, comprising 64% of the selected sample and 69% of those who were located.
Response did not vary by potential exposure status. The study population was primarily
Caucasian, with an average age of 27 years, and most had adequate prenatal care (72-73%). The
annual mass of tetrachloroethylene delivered to each address before and during pregnancy was
estimated using self-reported residential histories mapped using GIS (94% of reported
pregnancies), a leaching and transport model developed for the study, and EPA's EPANET
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-323 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
modeling software estimating water flow and direction. Estimated water concentrations of
tetrachloroethylene ranged between 1 and 5,197 (J,g/L.
Self-reported clinically recognized pregnancy loss (659 spontaneous abortions and
stillbirths) and 4,908 live births up to December 1990 were eligible for analysis. Pregnancy
outcomes were analyzed in relation to three measures of exposure: cumulative exposure up to the
month and year of the last menstrual period (prepregnancy window), peak exposure up to the last
menstrual period year of the pregnancy (prepregnancy window), and average monthly exposure
during the year containing the last menstrual period (time of conception). Risk of pregnancy loss
associated with exposure measures, divided into quartiles, was evaluated using generalized
estimating equations to account for lack of independence of multiple pregnancies by the same
woman. Risk estimates for pregnancy loss by increasing quartiles of exposure were similar
across the three exposure measures. For example, the multivariate GEE odds ratios for average
monthly exposure in increasing quartiles during the year of the last menstrual period were 1.1
(95% CI: 0.8-1.6), 0.7 (95% CI: 0.5-1.1), 0.8 (95% CI: 0.6-1.2), and 0.7 (95% CI: 0.5-1.0),
respectively. Several covariates were evaluated for potential confounding, including risk factors
for pregnancy loss, those associated with tetrachloroethylene exposure, and nondrinking water
sources of solvent exposure. Maternal age, year of pregnancy, paternal age, maternal history of
gynecologic infections, and the number of prior live births were included in the final models.
The authors checked the validity of self-reported birth outcomes by comparing the reproductive
histories reported by the women for all of the index pregnancies with information from birth
certificates. Further, information from medical records about pregnancies reported by 60 women
also was compared to self-reported histories. The authors reported good-to-excellent agreement
including for gestational duration and birth weight, prenatal cigarette smoking, number of prior
live births, and spontaneous and induced abortions. The study evaluated a large number of
pregnancy losses using a detailed exposure model and carefully assessed potential confounding.
It is important to note, however, that exposure estimates were not based on household
measurements, and individual consumption was not known. Therefore, exposure
misclassification may not have allowed detection of a small increase in risk. Finally, use of
exposure prior to the last menstrual period or during that year may not have had the required
precision to identify a risk associated with a particular susceptible window for pregnancy loss
(e.g., the first trimester).
In summary, the literature contains few studies of effects on spermatogenesis or
menstruation among subjects with exposure to tetrachloroethylene. One study of primarily
unionized workers in the dry-cleaning and laundry industries in California observed subtle
deficits in sperm quality in relation to tetrachloroethylene in exhaled breath, an exposure index,
and occupational group (dry-cleaning or laundry worker) (Eskenazi et al.. 1991b). However,
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-324 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
three clinically recognized measures of sperm quality were not associated with exposure in the
study population. The results of Eskenazi et al. (1991b) are compelling, but more studies are
needed to understand the spectrum of effects on sperm and their impact on fecundity. Two other
studies that evaluated effects on sperm, hormonal disturbances, or menstruation among men and
women with occupational exposure were not adequate to draw conclusions concerning the
association (Rachootin and Olsen. 1983; Zielhuis et al.. 1989). Some studies that relied on
detailed work histories and monitoring data to classify exposure suggested that maternal or
paternal exposure to tetrachloroethylene or work in dry cleaning reduces fertility or delays
conception (Eskenazi etal.. 1991a; Sallmen et al.. 1998; Sallmen et al.. 1995). However, the risk
estimates were imprecise because the number of participants reporting exposure to
tetrachloroethylene was small. As a consequence, the existing literature is inconclusive
concerning effects of tetrachloroethylene on reproduction and fertility.
A number of studies have evaluated the risk of spontaneous abortions in relation to
maternal and paternal occupational exposure to tetrachloroethylene. Results of several studies of
maternal occupational exposure to tetrachloroethylene suggest an increased risk of spontaneous
abortion, particularly at higher levels (Doyle et al.. 1997; Kyyronen et al.. 1989; Lindbohm et al..
1990; Olsen et al.. 1990; Windham et al.. 1991). Most of the studies evaluated exposure during
the first trimester of pregnancy. Some of the studies observed an increased odds ratio ranging
between 1.4 to 4.7, but had low statistical power because the cohort contained small numbers of
exposed cases and controls, and were limited in their ability to evaluate potential confounding
(Bosco et al.. 1987; Lindbohm et al.. 1990; Olsen et al.. 1990; Windham et al.. 1991). In general,
the studies that used a more precise definition of exposure, or categorized exposure into levels of
increasing dose or intensity, observed higher risk estimates (Doyle et al.. 1997; Kyyronen et al..
1989; Lindbohm et al.. 1990; Olsen et al.. 1990; Windham et al.. 1991). Increased risks were not
found among dry cleaners in Sweden (Ahlborg. 1990b; Olsen et al.. 1990). Three studies of
paternal occupational exposure prior to the beginning of the pregnancy did not observe an
association (Eskenazi et al.. 1991b; Lindbohm et al.. 1991; Taskinen et al.. 1989). Two of these
surveyed occupational exposure to a broad array of substances and, consequently, had low
statistical power for chemical-specific analyses (Lindbohm et al.. 1991; Taskinen et al.. 1989).
Although there is no evidence of an increased risk associated with paternal exposure, the studies
were not of sufficient size or detail in exposure estimates to draw conclusions. No associations
with incidence of spontaneous abortion were observed among two populations exposed to
tetrachl or ethylene in drinking water (Aschengrau et al.. 2008; Aschengrau et al.. 2009a; Lagakos
et al.. 1986). The populations were likely exposed to lower levels compared to the occupational
populations. In addition, the window of exposure used to assess risk in both studies may not
have had been precise enough to detect a small elevation in risk for spontaneous abortion.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-325 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
4.7.2.2. Animal Reproductive Toxicity Studies
Evaluation of the reproductive effects of tetrachloroethylene exposure in mammalian
animal models is based on a two-generation reproduction studies in rats, an in vivo sperm assay,
and an in vitro oocyte fertilization assay following in vivo exposure of adult female rats. These
studies are described below.
4.7.2.2.1. In vitro fertilization assay
In a study designed to examine the fertilizability of rat oocytes, female rats were exposed
to inhaled tetrachloroethylene at 12,000 mg/m3 (2 hours/day, 5 days/week) for 2 weeks (Berger
and Horner. 2003). The percentage of extracted oocytes that were fertilized in vitro was reduced
for tetrachloroethylene-treated females as compared with controls.
4.7.2.2.2. In vivo reproductive toxicity studies
Beliles et al. (1980) described an experiment in which male rats and mice (12/group)
were exposed via inhalation to tetrachloroethylene concentrations of 100 and 500 ppm, for
7 hours/day, for 5 days. Sperm head abnormalities and abnormal sperm were evaluated at 1, 4,
and 10 weeks after the last dose. Rats were unaffected. In mice, at 4 weeks, but not at 1 or 10
weeks after exposure, there was a significant increase (p < 0.05) in the percentage of males with
abnormal sperm heads (19.7%) in the 500-ppm exposure group. For the 100-ppm and control
groups, the percentages were 10.3 and 6% (not statistically significant at thep < 0.05 level),
respectively. A positive control group administered triethylene melanime was adversely affected
(11.1%). The authors suggested that the temporal appearance of the abnormal sperm heads
indicated that the spermatocyte and/or spermatogonia were the stages most sensitive to the
effects of inhaled tetrachloroethylene. In this study, the NOAEL was 100 ppm, and the LOAEL
was 500 ppm.
A multigeneration study of the effects on rats of exposure to airborne concentrations of
tetrachloroethylene was performed by Tinston (1994). Although this study has not been
published, it was submitted to EPA (Office of Prevention, Pesticides, and Toxic Substances and
to the IRIS Office as a result of the data call-in for the IRIS update). It was conducted under
GLP standards and received frequent quality assurance audits. In this study, weanling male and
female (Alpk:APfSD) rats (F0) (24/sex/group) were exposed to airborne tetrachloroethylene
concentrations of 0, 100, 300, or 1,000 ppm, 6 hours/day, 5 days/week, for 11 weeks prior to
mating and then for 6 hours/day during mating and through GD 20. There were no exposures
from GD 21 through Day 5 postpartum. One litter was produced in the first generation (F1 A).
The first-generation dams and their litters were exposed to tetrachloroethylene from PND 6
through 29, at which time, parental animals for the second generation were selected. The
second-generation parents (Fl) were then exposed 5 days/week during the 11-week premating
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-326 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
period. In the second generation, three litters were produced: F2A, F2B, and F2C. The F2A
dams and litters were exposed from Days 6 to 29 (control and 100 ppm) or Days 7 to 29
(300 ppm). The 1,000-ppm exposure for the F1 dams stopped after the F2A littering.
F2B litters were generated by mating the F1 parental males and females in the control,
300-, and 1,000-ppm groups; the dams and F2B litters were not exposed to tetrachloroethylene
during lactation. An F2C litter was produced by mating F1 males exposed to 1,000 ppm with
unexposed females. These females and the F2C litters were killed on PND 5 and discarded
without further examination. Overall, the F0 males were exposed for 19 weeks, and the F1
males were exposed up to 35 weeks. Postmortem evaluation in adults and selected weanlings
included organ weight and histopathology examination of liver, kidney, and reproductive organs;
sperm measures were not assessed.
Table 4-35 summarizes the results of the Tinston study. Signs of CNS depression
(decreased activity and reduced response to sound) were observed at 1,000 ppm for the first 2
weeks in both adult generations and again when the exposure was resumed on Day 6 postpartum
in the F1 generation (adults and pups). Other signs of overt tetrachloroethylene toxicity in the
adults included irregular breathing and piloerection at both 1,000 and 300 ppm and salivation
and tip-toe gait (in one F1 female) at 1,000 ppm. These changes stopped with the cessation of
exposure or within approximately 30 minutes thereafter.
There were a number of changes relative to controls that were of minor biological
significance. One change, transient statistically significant reductions of mean body weights
(originating from treated males and nontreated females), suggests the absence of male-mediated
effects on reproductive outcome. Nevertheless, the alterations in testes weight cannot be
discounted as a possible effect of treatment.
In females, dystocia was noted in one F0 dam at 100 ppm, two F1 dams at 300 ppm, and
a total of four dams (two each F0 and Fl) at 1,000 ppm; these dams were terminated without
completion of delivery. From the data for surviving dams and litters, it can be assumed that the
difficulties in parturition were not associated with or attributable to alterations in mean gestation
length or increased mean pup or litter weights. In fact, mean pup body weights showed a
statistically significant decrease throughout the lactation period at 300 and 1,000 ppm for Fl A
litters and in early lactation for F2A and F2B litters. Additionally, mean Fl A male pup body
weight was significantly decreased (5% less than controls; p < 0.05) at 100 ppm on PND 29.
These PND 29 mean body-weight deficits in all treated groups were observed in the animals
Table 4-35. Exposure concentrations (ppm) at which effects occurred in a
two-generation study
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-327 DRAFT-DO NOT CITE OR QUOTE
-------
Parameter
Generation
F0
F1A
F1
F2A
F2B
F2Ca
Clinical signs
(piloerection, irregular
breathing)
1,000, 300
1,000, 300
Behavioral effects
(decreased activity;
reduced response to sound)
1,000
1,000
1,000
Transient decreased body-
weight gains
1,000, 300
1,000, 300
Decreased mean testes
weight
1,000
1,000
Increased liver and kidney
weights
1,000
1,000
Renal histopathology
1,000
1,000
Decreased pups born alive
(percentage)
l,000b
l,000c
l,000c
Decreased mean
percentage pup survival
Days 1-5
1,000
l,000c
l,000c
Decreased mean
percentage pup survival
Days 5-22
l,000b
l,000b
NA
Decreased mean male pup
weight Day 1
l,000c
l,000c
l,000c
Decreased mean female
pup weight Day 1
l,000c
l,000b
l,000c
Decreased mean male pup
weight Day 29
l,000b,
300b, 100
b,d
NA
Decreased mean female
pup weight Day 29
l,000b,
300b, 100d
NA
a Not exposed after delivery.
hp < 0.05.
><0.01.
d trend/? <0.05.
NA = Not applicable (pups terminated on Day 5 postnatal).
Source: Adapted from Tinston (19941.
1 selected as parents of the second generation, but by the second week of the F1 premating period,
2 mean body weights were similar to those of controls for both 100- and 300 ppm-animals.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-328 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Mean litter size was decreased at 1,000 ppm for F2A and F2B litters. Statistically
significant decreases in the number of live pups on PND 1 (25 and 37% lower than controls for
F2A and F2B, respectively) are suggestive of either an adverse effect on fertilization or on in
utero survival. Early postnatal survival (i.e., on PND 1 and between PNDs 1 and 5) was also
compromised in F2A and F2B pups at 1,000 ppm, with mean litter sizes decreasing to 48% and
53%) of those of controls, respectively. The number of dead pups and litters with dead pups was
also increased, although not significantly, at 300 ppm for F2A litters. Clinical observations data
for 1,000-ppm litters reported an increased incidence of F2A and F2B pups that were found
dead, were killed in extremis, or were missing and presumed dead. The apparent increase in
adverse survival findings at 300 and 1,000 ppm in the second generation as compared with the
first generation could not be definitively attributed to any particular aspect of study design or
conduct (e.g., differences in the duration of treatment), although it is noted that, unlike the
second generation (Fl) parental animals, the first generation (F0) rats were not exposed to
tetrachloroethylene during preconception and in utero development.
A deficiency of the Tinston study is that the pregnant rats were not exposed from
gestation Day 21 through lactation Day 6 or 7, and the exposure at the 1,000-ppm treatment level
stopped for the Fl dams at the littering of the F2B pups. The F2B pups were not exposed
postnatally. It is additionally noted that this study was conducted according to the pre-1998 EPA
harmonized two-generation reproduction study guideline and, thus, did not assess a number of
sensitive endpoints such as estrous cyclicity, sperm measures, age of sexual maturation, and
enhanced reproductive organ pathology.
A summary of the doses at which treatment-related effects were observed in the Tinston
(1994) study is presented in Table 4-35. Overall, the parental systemic toxicity was observed at
300 and 1,000 ppm, with a NOAEL of 100 ppm. For offspring, the LOAEL of 100 ppm was
based upon decreased body weight in F1A pups at PND 21; no NOAEL was established. There
was no evidence of treatment-related effects on reproductive function at any exposure level
tested.
4.7.2.3. Reproductive Cancers in Humans
Thirteen epidemiologic studies reporting data on breast cancer and tetrachloroethylene
exposure and 12 epidemiologic studies reporting data on cervical cancer and tetrachloroethylene
exposure were identified. This set of studies includes 10 cohort studies on breast and cervical
cancers (Andersen et al.. 1999; Blair et al.. 2003; Boice et al.. 1999; Calvert et al.. In Press;
Chang et al.. 2005; Lynge and Thygesen. 1990; Pukkala et al.. 2009; Ruder et al.. 2001; S el den
and Ahlborg. 2011; Sung et al.. 2007). one study reporting on breast cancer but not cervical
cancer (Radican et al.. 2008). two studies reporting on cervical cancer but not breast cancer
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-329 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
(Anttila et al.. 1995; Travier et al.. 2002). two breast cancer case-control studies of occupational
exposures (Band et al.. 2000; Peplonska et al.. 2007). one cervical cancer case-control study of
occupational exposure (Lynge et al.. 2006). and one breast cancer case-control study of
residential exposure through contaminated drinking water (Aschengrau et al.. 2003).
Aschengrau et al. (2003) extended Aschengrau et al. (1998). adding additional breast cancer
cases from 1987-1993, and presenting odds ratios for the combined 10-year study period,
1983-1993. Most breast cancer studies examined females (Aschengrau et al.. 2003; Band et al..
2000; Blair et al.. 2003; Peplonska et al.. 2007; Radican et al.. 2008; Sung et al.. 2007) or males
and females combined (Boice et al.. 1999; Calvert et al.. In Press). Five studies, mostly of
Nordic subjects, presented risk estimates for male subjects separately (Andersen et al.. 1999;
Chang et al.. 2005; Lynge and Thygesen. 1990; Pukkala et al.. 2009; Selden and Ahlborg. 2011).
These studies represent the core studies evaluated by EPA, as described in more detail below.
Appendix B reviews the design, exposure-assessment approach, and statistical methodology for
each study. Most studies were of the inhalation route of exposure, of occupational exposure, and
lacked quantitative exposure information.Nine studies reporting risk estimates for breast or
cervical cancer examine occupational titles such as dry cleaner, launderer, and presser as
surrogates for tetrachloroethylene, given its widespread use from 1960 onward in the United
States and Europe (Andersen et al.. 1999; Band et al.. 2000; Blair et al.. 2003; Lynge et al.. 2006;
Lynge and Thygesen. 1990; Peplonska et al.. 2007; Pukkala et al.. 2009; Ruder et al.. 2001)
(Calvert et al.. In Press; Selden and Ahlborg. 2011). Five studies conducted in Nordic countries
are either based on either the entire Swedish population or on combined populations of several
Nordic countries; strengths of these studies are their use of job title as recorded in census
databases and ascertainment of cancer incidence using national cancer registries (Andersen et al..
1999; Lynge et al.. 2006; Lynge and Thygesen. 1990; Pukkala et al.. 2009); (Selden and
Ahlborg. 2011). Subjects in the multi-Nordic country study of Pukkala et al. (2009) overlapped
those of Lynge and Thygesen (1990). Andersen et al. (1999). Lynge et al. (2006). and Selden and
Ahlborg(2011). Studies examining mortality among U.S. dry-cleaner and laundry workers (Blair
et al.. 2003; Ruder et al.. 2001) are of smaller cohorts than the Nordic studies, with fewer
observed lung cancer events.
The exposure surrogate in studies of dry-cleaners and laundry workers is a broad
category containing jobs of differing potential for tetrachloroethylene exposure. Thus, these
studies have a greater potential for exposure misclassification bias compared to studies with
exposure potential to tetrachloroethylene assigned by exposure matrix approaches applied to
individual subjects. Calvert et al. studied unionized dry cleaners in the United States in
California, Illinois, Michigan, and New York who worked for one or more years before 1960 in
one or more shops known to use tetrachloroethylene as the primary solvent (Calvert et al.. In
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-330 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Pressin press; Ruder etal.. 1994. 2001). The cohort was stratified into two groups based on the
level of certainty that the worker was employed only in facilities using tetrachloroethylene as the
primary solvent; tetrachloroethylene-only and tetrachloroethylene plus. Lynge et al. (2006).
using job titles reported in the 1970 Census, identified subjects as dry cleaners (defined as dry
cleaners and supporting staff if employed in a business of <10 workers), other job titles in dry
cleaning (launderers and pressers), unexposed (job title reported on 1970 Census was other than
in dry cleaning), or unclassifiable (information was lacking to identify job title of subject).
Selden and Ahlborg (2011) identified subjects as either dry cleaners or laundry workers and
presented risk estimates separately by job title.
Four other cohorts with potential tetrachloroethylene exposure in industrial settings have
been examined. These studies include aerospace or aircraft maintenance workers in the United
States (Boice et al.. 1999; Radican et al.. 2008). workers, in Finland, primarily in the metal
industry (Anttila et al.. 1995) and electronic factory workers in Taiwan (Chang et al.. 2005; Sung
et al.. 2007). Boice et al. (1999) and Radican et al. (2008) used an exposure assessment based on
a job-exposure matrix, and Anttila et al. (1995) used biological monitoring in blood to assign
potential tetrachloroethylene exposure to individual subjects. In contrast and less sensitive, the
exposures in the Taiwan studies included multiple solvents and tetrachloroethylene exposure was
not linked to individual workers. Additionally, cohorts included white-collar workers, who had
an expected lower potential for exposure (Chang et al.. 2005; Sung et al.. 2007).
Aschengrau et al. (2003) is a case-control study that examined residential proximity to
drinking water sources contaminated with tetrachloroethylene in Cape Cod, MA, and used an
exposure model incorporating leaching and characteristics of the community water distribution
system to assign quantitative estimates of a household relative dose of tetrachloroethylene.
In summary, with respect to exposure-assessment methodologies, four studies with breast
or cervical cancer data assigned tetrachloroethylene exposure to individuals within the study
using a job exposure matrix (Anttila et al.. 1995; Boice et al.. 1999). an exposure model
(Aschengrau et al.. 2003). a classification of the cohort by certainty of tetrachloroethylene
exposure (Calvert et al.. In Press), or restricting analyses to subjects identified as dry cleaners
(Lynge et al.. 2006; Selden and Ahlborg. 2011). The relative specificity of these exposure-
assessment approaches strengthens their ability to identify cancer hazards compared to studies
with broader and less sensitive exposure-assessment approaches. The least sensitive exposure
assessments are those using very broad definitions such as working in a plant or a factory (Chang
et al.. 2003; Sung et al.. 2007).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-331 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Five1 of the nine breast cancer studies evaluated by EPA with exposure assessment to
tetrachloroethylene or employment as dry-cleaner or laundry worker reported estimated relative
risks based on 50 or more observed events (Aschengrau et al.. 2003; Blair et al.. 2003; Lynge
and Thygesen. 1990; Pukkala et al.. 2009; Selden and Ahlborg. 2011); the observed number of
breast cancer cases or deaths ranged from 56 (Blair et al.. 2003) to 1.757(Pukkala et al.. 2009).
The largest cohort of breast cancer cases in female dry-cleaners and laundry workers (n = 1,757)
observed a standardized incidence ratio of 0.89 (95% CI: 0.85, 0.94)(Pukkala et al.. 2010).
Three other studies of dry-cleaners and laundry workers with findings based on between 68 and
219 cases or deaths observed a standardized incidence ratio or SMR estimate of 0.88 (95% CI:
0.77, 1.01) (Selden and Ahlborg. 2011). 1.0 (95% CI: 0.8, 1.3) (Blair et al.. 2003). and 1.11 (95%
CI: 0.90, 1.34) (Lynge and Thygesen. 1990) for the association between breast cancer risk and
ever having a job title of dry-cleaner or laundry worker (see Table 4-36). A case-control study
with findings based on 50 or more exposed cases observed an odds ratio of 1.2 (95% CI: 0.9, 1.7)
for living in a residence receiving contaminated water with a relative delivered dose of
tetrachloroethylene above the median value (median: 2.1, range: 0.001-243.8) compared to
controls (Aschengrau et al.. 2003). SMRs or standardized incidence ratios for breast cancer were
similar for subjects identified as dry cleaners compared to laundry workers or for the subcohort
of females whose starting date of employment was after 1960 compared to the larger cohort
(Selden and Ahlborg. 2011).
In addition to the evidence from the large cohort and case-control studies, evidence is
found in five other studies whose effect estimates for breast cancer are based on fewer observed
events and that carry lesser weight in the analysis. As expected, the magnitude of the point
estimate of the association reported in these studies is more variable than in the larger studies:
0.48 (Radican et al.. 2008). 1.1 to 1.5 (Boice et al.. 1999; Calvert et al.. In Press; Peplonska et al..
2007). and >2.0 (Band et al.. 2000). Of these five studies, only risk estimates of Band et al.
(2000) excluded 1.0. Chang et al. (2005)and Sung et al. (2008). a follow-up study of the same
population, reported standardized incidence ratios of 1.19 (95% CI: 1.03, 1.36) and 1.09 (95%
CI: 0.96, 1.22). Both studies observed over 200 breast cancer incident cases; however, these
studies carry lesser weight in the analysis, given their low level of detail of the exposure
assessment.
1 Andersen et al. (19991 is not included in this summary of the data from the individual studies because it was
updated and expanded in the analysis by Pukkala et al. (20091.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-332 DRAFT-DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-36. Summary of human studies on tetrachloroethylene exposure and breast cancer
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Cohort Studies
Biologically monitored workers
Anttila et al. (1995)
All subjects
Not reported
849 Finnish men and women, blood PCE [0.4 (imol/L in females and 0.7
(imol/L in males (median)], follow-up 1974-1992, external referents (SIR)
Aerospace workers (Lockheed)
Boice et al. (1999)
Routine exposure to PCE
1.16 (0.32,2.97)
4
77,965 (// = 2,631 with routine PCE exposure and n = 3,199 with
intermittent-routine PCE exposure), besan work durins or after 1960.
worked at least 1 yr, follow-up 1960-1996, job exposure matrix without
quantitative estimate of PCE intensity, 1987-1988 8-h TWA PCE
concentration (atmospheric monitoring) 3 ppm [mean] and 9.5 ppm
[median], external reference for routine exposure (SMR) and internal
references (workers with no chemical exposures) for routine-intermittent
PCE exposure (RR), male (ICD-9, 175) and female breast cancer (ICD-9,
174)
Routine-Intermittent exposure duration to PCE
Not reported
Electronic factory workers (Taiwan)
Chans et al. (2005); Suns et al. (2007)
All Subjects
86,868 (n = 70,735 female), follow-up 1979-1997, multiple solvents
exposure, does not identify PCE exposure to individual subjects, cancer
mortality, external referents (SIR) (Chans et al.. 2005);
63,982 females, follow-up 1979-2001, factory employment proxy for
exposure, multiple solvents exposures and PCE not identified to individual
subjects, cancer incidence, external referents, analyses lagged 15 yr (SIR)
(Suns et al.. 2007)
Males
0.90 (0.48, 1.53)
0
0.11 exp
Females
1. 19 (1.03, 1.36)
215
Females
1.09 (0.96, 1.22)
286
Aircraft maintenance workers from Hill Air Force Base
Radican et al.. (2008)
Any PCE exposure
0.48 (0.07, 3.50)
1
10,461 men and 3,605 women (total n = 14,066, n = 10,256 ever exposed to
mixed solvents, 851 ever-exposed to PCE)), employed at least 1 yr from
1952 to 1956, follow-up 1973-2000, job exposure matrix (intensity),
internal referent (workers with no chemical exposures [RR]), female breast
cancer (ICD-A8, -9, 174; ICD-10, C50)
a -r
>1
to 0\
2 ^
3
LtJ
LtJ
H
O
o
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-36. Summary of human studies on tetrachloroethylene exposure and breast cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Dry-cleaner and laundry workers
Andersen et al.. (1999)
All laundry worker and dry cleaners
29,333 men and women identified in 1960 Census (Sweden) or 1970
Census (Denmark, Finland, Norway), follow-up 1971-1987 or 1991, PCE
not identified to individual subjects, external referents (SIR), ICD-7, 170
Males
(0, 3.41)
0
Females
0.89 (0.83, 0.97)
634
Blair et al. (2003)
All subjects
1.0(0.8, 1.3)
68
5,369 U.S. men and women laundry and dry-cleaning union members
(1945-1978), follow-up 1979-1993, semiquantitative cumulative exposure
surrogate to dry clean solvents, cancer mortality, external referents (SMR),
female breast (ICDA-8, 174).
Semiquantitative exposure score
Little to no exposure
0.8 (0.6, 1.2)
30
Medium to high exposure
1.2 (0.8, 1.7)
29
Ji et al. (7005^
Laundry workers and dry cleaners in 1960 Census
9,255 Swedish men and 14,974 Swedish women employed in 1960 (men)
or 1970 (women) as laundry worker or dry cleaner, follow-up
1961/1970-2000, PCE not identified to individual subjects, external
referent (SIR) and adjusted for age, period and socioeconomic status.
Males
Not reported
Females
Not reported
Lvnge and Thvgsen C1990)
All laundry worker and dry cleaners
10,600 Danish men and women, 20-64 yr old, employed in 1970 as
laundry worker, dry cleaners and textile dye workers, follow-up
1970-1980, external referents (SIR), ICD-7, 170.
Males
0
0.2 exp
Females
1.11(0.90, 1.34)
94
Pukkala et al. (2009)
Launderer and dry cleaner
Men and women participating in national census on or before 1990, 5
Nordic countries (Denmark, Finland, Iceland, Norway, Sweden), 30-64 yr,
follow-up 2005, occupational title of launderer and dry cleaner in any
census, external referents (SIR), ICD-7, 170.
Male
0.86 (0.18,2.50)
3
Female
0.89 (0.85, 0.94)
1,757
Calvert et al.
a -r
>1
to 0\
2 ^
3
LtJ
-t±
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
0
1
a -r
>1
to 0\
2 ^
All subjects
1.05 (0.70, 1.52)
28
1,704 U.S. men and women dry-cleaning union member in CA, IL, MI, NY
follow-up 1940-2004 (618 subjects worked for one or more years prior to
1960 only at shops where PCE was the primary cleaning solvent, identified
as PCE-only exposure), cancer mortality (SMR), female and male breast
cancer (ICD-9, 174, 175)
Exposure duration/time since 1st employment
Not reported
PCE-only subjects
1.06 (0.51, 1.94)
10
Selden and
Ahlbore (2011)
Dry-cleaners and laundry workers
9,440 Swedish men (n = 2,810) and women (n = 9,440) in 461 washing and
dry-cleaning establishments, identified by employer in mid-1980s,
employed 1973-1983, follow-up 1985-2000, exposure assigned using
company self-reported information on PCE usage—PCE (dry cleaners and
laundries with a proportion of PCE dry cleaning), laundry (no PCE use),
and other (mixed exposures to PCE, CFCs, TCE, etc.), external referents
(SIR), ICD-7, 170
Males
(0.00, 7.68)
0
Females
0.88 (0.77, 1.01)
219
PCE
Males
0
Females
0.85 (0.72, 1.00)
140
Laundry
Males
0
Females
0.96 (0.76, 1.21)
76
Travier et al. (2002)
All subjects, 1960 or 1970 Census in laundry and
dry cleaner or related occupation and industry
Not reported
Swedish men and women identified as laundry worker, dry cleaner, or
presser (occupational title), in the laundry, ironing, or dyeing industry or
related industry in 1960 or 1970 (543,036 person-years); or, as laundry
worker, dry cleaner, or presser (occupational and job title) (46,933 person-
years) in both censuses, follow-up 1971-1989, external referents (SIR)
All subjects in 1960 and 1970 in laundry and dry
cleaner occupation and industry
Not reported
l>j
LtJ
Lfl
H
O
o
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
0
1
a -r
>1
to 0\
2 ^
Case-control studies
British Columbia, Canada
Band et al.. (2000)
Laundry and dry cleaning occupation
995 breast cancer cases, females ,75 yr, 1988-1989, identified from British
Pre- and postmenopausal
Columbia Cancer Registry, Canadian citizens and British Columbia
residents, English speaking, 1,020 population controls matched on age and
sex, self-administered questionnaire, job title and industry coded to
Canadian SOC and Canadian SIC as exposure surrogate, OR for
postmenopausal subjects, adjusted for body weight in 1986, family history
Usual occupation
5.24(1.41, 19.5)
9
Postmenopausal
Usual occupation
4.85 (1.26, 18.7)
8
of breast cancer, history of benign breast disease, cumulative alcohol score.
OR for pre- and postmenopausal subjects also adjusted for smoking pack-
years
Power laundries and dry cleaners industry
Pre- and postmenopausal
Usual occupation
2.00 (0.78,5.13)
9
Postmenopausal
Usual occupation
1.57 (0.68,3.61)
10
Poland, 2 regions (Warsaw and Lodz)
Peplonska et al.. (2007)
Laundry, cleaning and garment services industry
1.2 (0.7, 1.9)
28
2,275 histologically confirmed in situ or invasive breast cancers in female
Exposure duration
residents of Warsaw and Lodz, 20-74 yr, 2000-2003, population controls ,
identified from the Polish Electronic System of Population Evidence and
matched to cases by city of residence and age within 5-yr age groups, in-
person interview, structured questionnaire, lifetime occupational history,
employed >6 mo in relevant industry exposure surrogate, OR adjusted for
age, education, age of menarche, menopausal status, age at menopause,
number of full-time births, MBI, family breast cancer history, and previous
screening mammography
<10 yr
1.5 (0.8, 2.8)
23
>10 yr
0.5 (0.2, 16)
5
l>j
LtJ
On
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
0
1
a -r
>1
to 0\
2 ^
Geographic-based studies
Cape Cod, MA
Aschensrau et al.. (1998).(2003)
PCC RDD < median
1.0 (0.7, 1.3)a
0.9 (0.6, 1.3)b
91
59
334 histologically confirmed breast cancer cases in males and females,
1983-1986, 2,236 population controls identified by random digit dialing,
vital records for deceased controls, and HCFA records if >65 yr
(Aschensrau et al.. 19981; 672 histologically confirmed Drimarv or
recurrent breast cancer cases in females, 1987-1993, 616 population
controls identified by random digit dialing, vital records for deceased
controls, and HCFA records if >65 vr (Aschensrau et al.. 20031; MA
Cancer Reeistrv. telephone interview, algorithm of (19931 to estimate mass
of PCE in drinking water entering residence was surrogate exposure metric
[90th percentile, 53.4], OR adjusted for age of diagnosis or index year, vital
status at interview, family history of breast cancer, age at first live birth,
personal history of prior breast cancer and benign breast disease, and
occupational exposure to solvents (PCE, benzene, other solvents),
statistically analyses also explored effect of different latent periods
(0,5,7,9,11,13, and 15 yr)
PCE RDD > median
1.2 (0.9, 1.7)a
1.3 (0.9, 1.9)b
100
69
PCE RDD >90lh percentile
1.3 (0.7, 2.6)a
1.7 (0.8, 4.4)b
4
16
a In Aschengrau et al. (20031. odds ratios for breast cancer are presented for combined data from Aschengrau et al. (19981.
bOdds ratios considering a 7-yr latent period.
HCFA = Health Care Financing Administration, ISCO = International Standard Classification of Occupation, ISIC = International Standard Industry
Classification, JEM = job-exposure-matrix, RDD = relative delivered dose, TWA = time-weighted-average.
LtJ
LtJ
^1
~n
H
O
O
-------
No male breast cancer cases were observed in four of the five studies reporting risk
estimates for males separately from that of females (Anderson et al.. 1990; Chang et al.. 2005;
1990; Selden and Ahlborg. 2011). Not surprising given the low background rate of male breast
cancer, less than one case was expected in each study. Pukkala et al. (2010) reported three
observed cases among a cohort of 8,744 male dry-cleaners and laundry workers.
Two1 of the eight cervical cancer studies evaluated by EPA with exposure assessment to
tetrachloroethylene or employment as dry-cleaner or laundry worker reported estimated relative
risks based on 50 or more observed events. Estimates of the standardized incidence ratio or
SMR in these studies were 1.34 (95% CI: 1.12, 1.60) and 1.20 (95% CI: 1.08, 1.34) in Travier et
al. (2002) and Pukkala et al. (2009), respectively. In addition to the evidence from the two
large cohort studies, additional evidence is found in six other studies whose effect estimates are
based on fewer observed events and that carry lesser weight in the analysis. As expected, the
magnitude of the point estimate of the association reported in these studies is more variable than
in the larger studies: 0.40 to 0.98 (Lynge et al.. 2006; Lynge and Thygesen. 1990). 1.1 to 1.5
(Selden and Ahlborg. 2011). 1.6 to 2.0 (Blair et al.. 2003; Calvert et al.. In Press; Ruder et al..
2001). and >3.0 (Anttila et al.. 1995). Chang et al. (2005) and Sung et al. (2008). a follow-up
study of the same population, observed over 200 cervical cancer incident cases and reported
standardized incidence ratios of 1.06 (95% CI: 0.95, 1.18) and 0.69 (95% CI: 0.87, 1.06).
Although based on a large number of observed events, these studies carry lesser weight in the
analysis given their lower level exposure-assessment approach. SMRs or standardized incidence
ratios for cervical cancer were lower for subjects identified as dry cleaners compared to laundry
workers or for the subcohort of females whose starting date of employment was after 1960
compared to the larger cohort (Lynge et al.. 2006; Selden and Ahlborg. 2011) (see Table 4-37).
Establishment of an exposure or concentration-response relationship can add to the
weight of evidence for identifying a cancer hazard, but only limited data pertaining to exposure-
response relationships for lung cancer and tetrachloroethylene exposure are available. Three
studies of breast cancer presented risk estimates for increasing exposure categories; one study
using exposure duration as a proxy (Peplonska et al.. 2007) and two studies with a
semiquantitative or quantitative exposure surrogate (Aschengrau et al.. 2003; Blair et al.. 2003).
1 In addition to Andersen et al. (19991. Boice et al. (19991 is not counted because no cervical deaths are observed
among tetrachloroethylene-exposed female subjects.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-338 DRAFT-DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-37. Summary of human studies on tetrachloroethylene exposure and cervical cancer
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Cohort studies
Biologically monitored workers
Anttila et al. (1995)
All subjects
3.20 (0.39, 11.6)
2
849 Finnish men and women, blood PCE [0.4 (imol/L in females
and 0.7 (imol/L in males (median)], follow-up 1974-1992, external
referents (SIR)
Aerospace workers (Lockheed)
Boice et al. (1999)
Routine exposure to PCE
(0.00, 7.77)
0
0.47 exp
77,965 (// = 2,631 with routine PCE exposure and n = 3,199 with
intermittent-routine PCE exposure), besan work durins or after
1960, worked at least 1 yr, follow-up 1960-1996, job exposure
matrix without quantitative estimate of PCE intensity, 1987-1988
8-h TWA PCE concentration (atmospheric monitoring) 3 ppm
[mean] and 9.5 ppm [median], external reference for routine
exposure (SMR) and internal references (workers with no chemical
exposures) for routine-intermittent PCE exposure (RR)
Routine-Intermittent exposure duration to PCE
Not reported
Electronic factory workers (Taiwan)
Change et al. (2005); Sung et al. (2007)
All Subjects
86,868 (n = 70,735 female), follow-up 1979-1997, multiple
solvents exposure, does not identify PCE exposure to individual
subjects, cancer mortality, external referents (SIR); female genital
orsans (Chans et al.. 2005);
63,982 females, follow-up 1979-2001, factory employment proxy
for exposure, multiple solvents exposures and PCE not identified to
individual subjects, cancer incidence, external referents, analyses
lassed 15 vr (SIR) (Suns et al.. 2007)
Females
1.06 (0.95, 1.18)
337
Females
0.96 (0.87, 1.06)
337
Aircraft maintenance workers from Hill Air Force Base
Radican et al. (2008)
Any PCE exposure
Not reported
10,461 men and 3,605 women (total n = 14,066, n = 10,256 ever
exposed to mixed solvents, 851 ever-exposed to PCE), employed at
least 1 yrfrom 1952 to 1956, follow-up 1973-2000, job exposure
matrix (intensity), internal referent (workers with no chemical
exposures) (RR)
a -r
>1
to 0\
2 ^
3
LtJ
^o
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-37. Summary of human studies on tetrachloroethylene exposure and cervical cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Dry-cleaner and laundry workers
Andersen et al. (1999)
All laundry worker and dry cleaners
1.18(1.01, 1.38)
155
29,333 men and women identified in 1960 Census (Sweden) or
1970 Census (Denmark, Finland, Norway), follow-up 1971-1987
or 1991, PCE not identified to individual subjects, external referents
(SIR)
Blair et al. (2003)
All subjects
1.6(1.0, 2.3)
27
5,369 U.S. men and women laundry and dry-cleaning union
members (1945-1978), follow-up 1979-1993, semiquantitative
cumulative exposure surrogate to dry clean solvents, cancer
mortality, external referents (SMR)
Semiquantitative exposure score
Little to no exposure
1.5 (0.8, 2.7)
12
Medium to high exposure
1.4 (0.7, 1.7)
11
Ji et al. (2005a. 2005b: 2005a. 2005b. 2005c)
Laundry workers and dry cleaners in 1960 Census
Not reported
9,255 Swedish men and 14,974 Swedish women employed in 1960
(men) or 1970 (women) as laundry worker or dry cleaner, follow-up
1961/1970-2000, PCE not identified to individual subjects, external
referent (SIR) and adjusted for age, period and socioeconomic
status
Lvnse and Thvsesen (1990)
Laundry worker and dry cleaners
0.40 (0.28, 0.52)
34
10,600 Danish men and women, 20-64 yr old, employed in 1970 as
laundry worker, dry cleaners and textile dye workers, follow-up
1970-1980, external referents (SIR)
Pukkala et al. (2009)
Launderer and dry cleaner
1.20 (1.08, 1.34)
332
Men and women participating in national census on or before 1990,
5 Nordic countries (Denmark, Finland, Iceland, Norway, Sweden),
30-64 yr, follow-up 2005, occupational title of launderer and dry
cleaner in any census, external referents (SIR)
a -r
>1
to 0\
2 ^
3
o
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-37. Summary of human studies on tetrachloroethylene exposure and cervical cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Ruder et al. (2001): Calvert et al.
All subjects
1.84 (0.98,3.14)
13
1,704 U.S. men and women dry-cleaning union member in CA, IL,
MI, NY follow-up 1940-2004 (618 subjects worked for one or
more years prior to 1960 only at shops where PCE was the primary
Exposure duration/time since 1st employment
Not reported
<5 yr/<20 yr
0.84 (0.15,2.66)
2
cleaning solvent, identified as PCE-only exposure), cancer
mortality (SMR), female and male breast cancer (ICD-9, 174, 175)
>5 yr/<20 yr
2.63 (0.90, 6.03)
4
<5 yr/>20 yr
2.75 (0.94, 6.30)
4
>5 yr/>20 yr
2.08 (0.57, 5.38)
3
PCE subcohort
2.10(0.68,4.90)
5
Selden and Ahlborg (2011)
Dry-cleaners and laundry workers
1.25 (0.81, 1.85)
25
9,440 Swedish men (n = 2,810) and women (n = 9,440) in 461
washing and dry-cleaning establishments, identified by employer in
mid-1980s, employed 1973-1983, follow-up 1985-200, exposure
PCE
1.19 (0.64, 1.93)
16
Duration of employment
assigned using company self-reported information on PCE usage—
PCE (dry cleaners and laundries with a proportion of PCE dry
cleaning), laundry (no PCE use), and other (mixed exposures to
PCE, CFCs, TCE, etc.), external referents (SIR)
<1 yr
0.32 (0.01, 1.78)
1
1-4 yr
1,72 (0.7, 3.40)
8
5-11yr
1.24 (0.50, 2.56)
7
Laundry
1.45 (0.66, 2.75)
9
Travier et al. (2002)
All subjects, 1960 or 1970 Census in laundry and
dry cleaner or related occupation and industry
1.34 (1.12, 1.60)
129
Swedish men and women identified as laundry worker, dry cleaner,
or presser (occupational title), in the laundry, ironing, or dyeing
industry or related industry in 1960 or 1970 (543,036 person-years);
or, as laundry worker, dry cleaner, or presser (occupational and job
title) (46,933 person-years) in both censuses, follow-up 1971-1989,
external referents (SIR)
All subjects in 1960 and 1970 in laundry and dry
cleaner occupation and industry
1.09 (0.57, 2.09)
9
a -r
>1
to 0\
2 ^
3
~n
H
O
O
-------
a
o
<
s:
cs
>!
o
^5
§•
rs
st
3
>!
sT
St
o
>1
>!
Oq
o
>!
Si
Si
s
O
Sfc
I
§•
K) On
2
Table 4-37. Summary of human studies on tetrachloroethylene exposure and cervical cancer (continued)
Exposure group
Relative risk
(95% CI)
No. obs.
events
Reference
Case-control studies
Nordic Countries (Denmark, Finland, Norway, Sweden)
Lvnge et al. (2006)
Unexposed
1.00
105
Case-control study among 46,768 Danish, Finnish, Norwegian, and
Swedish men and women employed in 1960 as laundry worker or
dry cleaner, follow-up 1970-1971 to 1997-2001, 102 cervical
cancer cases, 3 controls per case randomly selected from cohort
matched on country, sex, age, calendar period at diagnosis time,
occupational task at 1970 Census proxy for exposure, cervical
cancer incidence, RR adjusted for matching criteria
Dry cleaner
0.98 (0.65, 1.47)
36
Other in dry-cleaning
1.72 (1.00, 2.97)
22
Unclassifiable
1.11 (0.72, 1.71)
44
Dry cleaner, employment duration, 1964-1979
<1 yr
2.68 (0.89, 8.11)
7
2-4 yr
0.78 (0.31, 1.94)
6
5-9 yr
0.47 (0.20, 1.13)
6
>10 yr
1.18(0.64,2.15)
16
Unknown
1.14(0.12, 11.00)
1
HCFA = Health Care Financing Administration, ISCO = International Standard Classification of Occupation, ISIC = International Standard Industry
Classification, JEM = job-exposure-matrix, TWA = time-weighted-average.
to
~n
H
O
O
-------
1
2
3
4
5
6
7
8
9
10
11
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
Risk estimates are larger for highest exposure groups compared to overall exposure or to a no or
low exposed group in one cohort study that use a semiquantitative or quantitative exposure-
assessment approach (Blair et al.. 2003). and in one study when latent periods are considered
(Aschengrau and Seage. 2003). One other study with an exposure assessment based on exposure
duration reported a lower risk estimate with >10 years longer exposure duration than the risk
estimate for <10 years (Peplonska et al.. 2007).
With respect to cervical cancer, five studies presented risk estimates for increasing
exposure categories using exposure duration (Blair et al.. 2003; Lynge et al.. 2006; Ruder et al..
2001; Selden and Ahlborg. 2011; Travier et al.. 2002). Ruder et al. (2001) was the only study to
report a higher risk estimate for cervical cancer for the group with longest exposure duration (<5
years versus 5+ years).
All three case-control studies of breast cancer controlled for associated risk factors
(Aschengrau and Seage. 2003; Band et al.. 2000; Peplonska et al.. 2007). Direct examination of
possible confounders is less common in cohort studies examining breast cancer compared to
case-control studies where information is obtained from study subjects or their proxies. None of
the cohort studies of cervical cancer considered socioeconomic or lifestyle factors such as
smoking or exposure to the human papilloma virus (HPV), a known risk factor for cervical
cancer and correlated with socioeconomic status, particularly with the squamous cell subtype
(NCI. 2010; Pukkala et al.. 2010). The case-control study of Lynge et al. (2006) included
controls similar in socioeconomic status as cases, and the odds ratio estimate in this study for dry
cleaners did not support an association with tetrachloroethylene.
In conclusion, most studies examined breast cancer in females (Aschengrau et al.. 2003;
Aschengrau and Seage. 2003; Band et al.. 2000; Blair et al.. 2003; Peplonska et al.. 2007;
Radican et al.. 2008; Sung et al.. 2007); or males and females combined (Boice et al.. 1999;
Calvert et al.. In Press; Ruder et al.. 2001). Five studies, mostly of Nordic subjects, presented
risk estimates for male subjects separately (Anderson et al.. 1990; Chang et al.. 2005; Lynge and
Thygesen. 1990; Pukkala et al.. 2009; Selden and Ahlborg. 2011). The results from the large
studies of breast cancer risk in women in relation to tetrachloroethylene exposure are mixed.
The largest, based on 1,757 breast cancer cases in female dry-cleaners and laundry workers,
reported a statistically significant deficit in the risk of breast cancer incidence compared to the
populations of Nordic countries (Pukkala et al.. 2009). Findings in the other six studies were
based on fewer events or exposed cases; two of four studies with nonspecific exposure-
assessment methodology provided evidence for association between breast cancer in females and
tetrachloroethylene exposure (Anderson et al.. 1990; Aschengrau et al.. 2003; Chang et al.. 2005;
Lynge and Thygesen. 1990; Sung et al.. 2007) but effects were not seen in two other large cohort
studies with a relatively high quality exposure-assessment methodology to tetrachloroethylene
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-343 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
(Blair et al.. 2003; Selden and Ahlborg. 2011). Small studies observed mixed findings
(Aschengrau and Seage. 2003; Band et al.. 2000; Boice et al.. 1999; Chang et al.. 2005;
Peplonska et al.. 2007; Radican et al.. 2008; Ruder et al.. 2001; Sung et al.. 2007). Band et al.
(2000). but not other less-weighted studies, excluded chance as an alternative explanation.
Although cohort studies were unable to control for potential confounding from reproductive
history or menopausal status, observations in case-control studies controlled for these potential
confounders in statistical analyses and provided support of an association between female breast
cancer and tetrachloroethylene compared to controls (Aschengrau et al.. 2003; Band et al.. 2000;
Peplonska et al.. 2007). Three studies examined exposure response, with risk estimates in
females monotonically increased in higher exposure groups in two studies with semiquantitative
or quantitative exposure-assessment approaches (Aschengrau et al.. 2003; Blair et al.. 2003). A
third study examining exposure duration observed an inverse relation (Peplonska et al.. 2007).
Exposure duration is more uncertain than use of a semiquantitative surrogate given increased
potential for bias associated with exposure misclassification. Because of the limitation in
statistical power, none of the five studies reporting on male breast cancer is adequate to examine
tetrachloroethylene exposure. All studies of male breast cancer are sufficiently underpowered;
no male breast cancer cases were observed in four of the five studies (Anderson et al.. 1990;
Chang et al.. 2005; Lynge and Thygesen. 1990; Selden and Ahlborg. 2011).
For cervical cancer, the results from the two large cohort studies of dry cleaners are
consistent with an elevated cervical cancer risk of 20-30% (Pukkala et al.. 2009; Travier et al..
2002). Results from four smaller cohort and case-control studies with a relatively high quality
exposure-assessment methodology presented a pattern of more variable results, with relative
risks of 0.98 (95% CI: 0.65, 1.47), 1.19 (95% CI: 0.64, 1.93), 2.10 (95% CI: 0.68, 4.90), and 3.20
(95% CI: 0.39, 11.6) (Anttila et al.. 1995; Blair et al.. 2003; Ruder et al.. 2001; Selden and
Ahlborg. 2011). A fourth study with higher quality exposure-assessment specific to
tetrachloroethylene did not observe any cervical cancer deaths among women, but less than one
death was expected (Boice et al., 1999). Calvert et al. (in press) was the only study to report an
exposure response gradient with employment duration. Dry cleaning or workers with
employment after 1960 when tetrachloroethylene use was more prevalent did not have higher
cervical cancer risk estimates than laundry workers (Lynge et al.. 2006; Selden and Ahlborg.
2011). Lack of data on socioeconomic status—a proxy for exposure to the human papilloma
virus, a known risk factor for cervical cancer—indicates great uncertainty for asserting this
association with tetrachloroethylene exposure. Potential confounding by socioeconomic status is
an alternative explanation, with some support provided by Lynge et al. (2006). a case-control
study with controls of similar socioeconomic status as cases, and who did not observe an
association between cervical cancer and dry cleaning.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-344 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
4.7.3. Summary of Human and Animal Developmental/Reproductive Studies
4.7.3.1. Summary of Human Data
Studies of tetrachloroethylene exposure have evaluated several reproductive outcomes
including effects on menstrual disorders (Zielhuis et al.. 1989). semen quality (Eskenazi et al..
1991a; Eskenazi et al.. 1991b). fertility (Eskenazi etal.. 1991b; Rachootin and Olsen. 1983).
time to pregnancy (Sallmen et al.. 1998; Sallmen et al.. 1995). and risk of adverse pregnancy
outcomes including spontaneous abortion (Ahlborg. 1990a; Aschengrau et al.. 2009a; Bosco et
al.. 1987; Doyle et al.. 1997; Kyyronen et al.. 1989; Lindbohm et al.. 1991; Lindbohm et al..
1990; McDonald et al.. 1986; McDonald et al.. 1987; Olsen et al.. 1990; Taskinen et al.. 1989;
Windham et al.. 1991). low birth weight or gestational age (Aschengrau et al.. 2008; Bosco et al..
1987; McDonald et al.. 1987; Olsen et al.. 1990). birth anomalies (Ahlborg. 1990a; Aschengrau
et al.. 2009b; Bosco et al.. 1987; McDonald et al.. 1987; Olsen et al.. 1990). and stillbirth
(McDonald et al.. 1987; Olsen et al.. 1990). A few studies evaluated effects of prenatal exposure
to tetrachloroethylene on postnatal development including learning and behavior, and
schizophrenia (Janulewicz et al.. 2008; Perrin et al.. 2007). Many of the studies evaluated
exposure during a specific critical window relevant to the health endpoint under study, for
example, the period before conception or during the first trimester.
Some studies that relied on detailed work histories and monitoring data to classify
exposure were suggestive that maternal or paternal exposure to tetrachloroethylene or work in
dry cleaning reduces fertility or delays conception (Sallmen et al.. 1998; Sallmen et al..
1995){Eskenazi, 1988, 701886}. However, the risk estimates were imprecise because the
number of participants reporting exposure to tetrachloroethylene was small. One small study of
primarily unionized workers in the dry-cleaning and laundry industries in California observed
subtle deficits in sperm quality in relation to tetrachloroethylene exposure {Eskenazi, 1988,
701886}. However, three clinically recognized measures of sperm quality were not associated
with exposure in the study population. A study of occupational exposures among a group of
infertile couples who sought treatment found no association between either a diagnosis of sperm
abnormalities among male partners, or a diagnosis of hormonal disturbances among female
partners with self-reported exposure to dry-cleaning chemicals (Rachootin and Olsen. 1983).
The results of Eskenazi et al. {, 1988, 701886} are compelling, but more studies are
needed to conclude if exposure to tetrachloroethylene is associated with adverse effects on male
and female reproduction.
Results of several studies of maternal occupational exposure to tetrachloroethylene
suggest an increased risk of spontaneous abortion, particularly at higher levels (Doyle et al..
1997; Kyyronen et al.. 1989; Lindbohm et al.. 1990; Olsen et al.. 1990; Windham et al.. 1991).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-345 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Most of the studies evaluated exposure during the first trimester of pregnancy. Some of the
studies observed an increased odds ratio ranging between 1.4 to 4.7, but had low statistical power
because the cohort contained small numbers of exposed cases and controls, and were limited in
their ability to evaluate potential confounding (Bosco et al.. 1987; Lindbohm et al.. 1990; 01 sen
et al.. 1990; Windham et al.. 1991). In general, the studies that used a more precise definition of
exposure, or categorized exposure into levels of increasing dose or intensity, observed higher
risk estimates. For example, two reports of occupational exposure in the dry-cleaning and
laundry industries in Finland observed a dose-related increase in risk among employees
classified into risk levels based on whether or not their work tasks involved dry cleaning
(Kyyronen et al.. 1989; 01 sen et al.. 1990). Odds ratios for low and high exposure compared to
no exposure were 1.18 (95% CI: 0.71-1.97) and 4.53 (95% CI: 1.11-18.5), respectively. The
Finnish studies controlled for reported exposure to other substances in the workplace as well as
for several potential confounders. They also found agreement between self-reported exposures
and biological measurements taken close to the time of pregnancy for a small subset of the
cohorts. A relatively large study of workers in the United Kingdom classified exposure among
current and former employees at dry-cleaning and laundry establishments by job tasks (machine
operator versus other tasks) and analyzed risk of spontaneous abortions among all pregnancies
reported between 1980 and 1995 (Doyle et al.. 1997). Machine operators had a 63% higher risk
of spontaneous abortion compared to nonoperators adjusting for several potential confounders
(OR: 1.63, 95% CI: 1.09-3.05). These findings are consistent with breathing zone
measurements of tetrachloroethylene in dry-cleaning establishments, indicating that machine
operators have the highest exposures (Gold et al.. 2008).
Increased risks were not found among dry cleaners in Sweden using a comparable study
design (Ahlborg. 1990b; 01 sen et al.. 1990). Further, three studies of paternal occupational
exposure prior to the beginning of the pregnancy did not observe an association (Eskenazi et al..
1991b; Lindbohm et al.. 1991; Taskinen et al.. 1989). Two of these surveyed occupational
exposure to a broad array of substances and, consequently, had low statistical power for
chemical-specific analyses (Lindbohm et al.. 1991; Taskinen et al.. 1989). Although there is no
evidence of an increased risk associated with paternal exposure, the studies were not of sufficient
size, nor did they provide adequate detail regarding exposure estimates to allow definitive
conclusions. Finally, no associations with incidence of spontaneous abortion were observed
among two populations exposed to tetrachl or ethylene in drinking water (Aschengrau et al.. 2008;
Lagakos et al.. 1986)schengrau et al.„ 2009). The studies of drinking water contamination
evaluated populations with much lower exposures compared to the occupational cohorts.
Studies of tetrachloroethylene in drinking water have reported that exposure during
pregnancy is associated with low birth weight (Bove et al.. 1995; Lagakos et al.. 1986;
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-346 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Sonnenfeld et al.. 20011 eye/ear anomalies (Lagakos et al.. 1986). and oral clefts (Aschengrau et
al.. 2009b; Bove et al.. 1995; Lagakos et al.. 1986). However, the number of cases with birth
anomalies in specific diagnostic groups was very small, and CIs often included one. In addition,
imprecise exposure estimates likely resulted in nondifferential misclassification, biasing risk
estimates toward the null. Participants in the studies were exposed to multiple contaminants, and
it was not possible to disentangle substance-specific risks.
Aschengrau et al. (2008) evaluated a unique exposure event in a population in eight Cape
Cod towns exposed to a wide range of tetrachloroethylene concentrations in an irregular pattern
throughout the region (1.5-7,750 (J,g/L). It is less likely that the population was exposed to
sizable concentrations of other halogenated substances. A detailed exposure model was used to
estimate the distribution of contaminated water to the homes of residents. Birth weight and
gestational age were not associated with exposure to tetrachl or ethylene. Effect estimates for
some congenital anomalies were increased, although the number of infants with anomalies was
very small, and statistical power was low. The small increased risk is consistent with the other
studies of drinking water exposure to mixtures of halogenated pollutants in drinking water.
Diagnoses of attention deficit disorder, hyperactive disorder or educational histories reported by
the mothers about their children were not increased in relation to the amount of
tetrachloroethylene delivered to the homes during pregnancy or childhood (Janulewicz et al..
2008). On the other hand, a more than threefold risk of schizophrenia was associated with dry
cleaning as a surrogate for prenatal tetrachloroethylene exposure (Perrin et al.. 2007). The
longitudinal design and use of a national registry to identify psychiatric diagnoses were strengths
of the study, but tetrachloroethylene exposure was not directly analyzed. In conclusion, the
literature is insufficient to draw conclusions regarding effects of tetrachloroethylene exposure on
development in infants and children.
Most epidemiologic studies examined breast cancer in females (Aschengrau et al.. 2003;
Band et al.. 2000; Blair et al.. 2003; Peplonska et al.. 2007; Radican et al.. 2008; Sung et al..
2007) or males and females combined (Boice et al.. 1999; Calvert et al.. In Press; Ruder et al..
2001); five studies presented risk estimates for male subjects separately (Anderson et al.. 1990;
Chang et al.. 2005; Lynge and Thygesen. 1990; Pukkala et al.. 2009; Selden and Ahlborg. 2011).
The largest study, based on 1,757 breast cancer cases in female dry-cleaners and laundry
workers, reported a statistically significant deficit in the risk of breast cancer incidence compared
to the populations of Nordic countries. Findings in the other four large studies were based on
fewer events or exposed cases with mixed findings (Aschengrau et al.. 2003; Blair et al.. 2003;
Lynge and Thygesen. 1990; Selden and Ahlborg. 2011). Additional studies carrying less weight
also observed mixed findings (Band et al.. 2000; Boice et al.. 1999; Chang et al.. 2005;
Peplonska et al.. 2007; Radican et al.. 2008; Ruder et al.. 2001; Sung et al.. 2007). Three studies
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-347 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
examined exposure-response, with risk estimates in females monotonically increased in higher
exposure groups in two studies with semiquantitative or quantitative exposure-assessment
approaches (Aschengrau et al.. 2003; Blair et al.. 2003) and a negative direction in a third study
examining exposure duration. Exposure duration is an inferior metric compared to a
semiquantitative approach because there is increased potential for bias associated with exposure
misclassification (Peplonska et al.. 2007). None of the five studies reporting on male breast
cancer is adequate to examine tetrachloroethylene exposure. All studies of male breast cancer
are statistically underpowered; no male breast cancer cases were observed in four of the five
studies (Anderson etal.. 1990; Chang et al.. 2005; Lynge and Thygesen. 1990; S el den and
Ahlborg. 2011). less than one case was expected in each study, and Pukkala et al. (2010)
observed three cases among a cohort of 8,744 male dry-cleaners and laundry workers.
For cervical cancer, the results from the two large cohort studies with broad exposure
assessment is consistent with an elevated cervical cancer risk of 20-30% (Pukkala et al.. 2009;
Travier et al.. 2002). Results from four smaller cohort and case-control studies with a higher
quality exposure-assessment methodology presented a pattern of more variable results, with
relative risks of 0.98, 1.19, 1.89, and 3.20 in Lynge et al. (2006). Selden and Ahlborg (2011).
Ruder et al. (2001). and Anttila et al. (1995). respectively. A fourth study with high quality
exposure assessment specific to tetrachloroethylene did not observe any cervical cancer deaths
among women and was insensitive, as less than one death was expected. Ruder et al. (2001) was
the only study to report an exposure response gradient. Dry cleaning or workers employed after
1960 when tetrachloroethylene use was more prevalent did not have higher cervical cancer risk
estimates than laundry workers (Lynge et al.. 2006; Ruder et al.. 2001; Selden and Ahlborg.
2011). Lack of data on socioeconomic status—a proxy for exposure to the human papilloma
virus, a known risk factor for cervical cancer—indicates great uncertainty for asserting this
association with tetrachloroethylene exposure. Potential confounding by socioeconomic status is
an alternative explanation with some support provided by Lynge et al. (2006). a case-control
study with controls of similar socioeconomic status as cases, and who did not observe an
association between cervical cancer and dry cleaning.
4.7.3.2. Summary of Animal Data
Table 4-38 summarizes the findings of the animal developmental and reproductive
toxicity studies described in Sections 4.7.2.1 to 4.7.2.3. The inhalation study database includes
assessments of developmental toxicity in rats, mice, and rabbits following exposures during
gestation, assessments of developmental neurotoxicity in rats following pre- and/or postnatal
exposures of the offspring, and evaluation of reproductive and fertility outcomes in rats and
mice. Additional supportive studies include in vitro assays of embryo development and oocyte
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-348 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-38. Summary of mammalian developmental and reproductive
toxicity studies for tetrachloroethylene
Subjects
Effects
Concentration
Authors
Developmental toxicity studies
Rat (whole
embryo culture)
Mortality, malformations, delayed
growth and differentiation
No effect at 2.5 mM, effects at 3.5
mM and higher
Saillenfait et al.
(1995)
Japanese
medaka
Decreased egg viability at 96-h
(LC50 = 27 mg/L); at 10 d: decreased
hatchability and larval survival,
increased developmental abnormalities
10-d: 0, 1.5,3,6, 12, 25 mg/L
LOAEL =1.5 mg/L
Spencer et al.,
(2002)
SW Mice
Maternal toxicity, decreased fetal
weight, delayed ossification, 9%
decrease in birth weight
Inhalation: 0, 300 ppm on GDs 6-15
Schwetz et al.
(1975)
S-D Rats
Maternal toxicity, increased resorptions
(fetal death)
Inhalation: 0, 300 ppm on Days 6-15
Schwetz et al.
(1975)
S-D Rats, NZW
Rabbits
No developmental toxicity
Inhalation: Exposures throughout
gestation
NOAEL = 500 ppm
Hardin et al.
(1981)
F344 Rats
100% mortality at 1,200 mg/kg-day,
increased mortality and micro-
/anophthalmia at 900 mg/kg-day; soft
tissues not examined
Gavage, 0, 900, 1,200 mg/kg-day on
GDs 6-19
Narotsky and
Kavlock (1995)
CFY Rats
Maternal toxicity (decreased body
weight gain, increased liver weight and
serum enzymes); increased pre- and
postimplantation loss, skeletal
retardation, and total malformations;
decreased fetal weight
Inhalation: 0, 1,500, 4,500, 8,500
mg/m3 on GDs 1-20
LOAEL = 1,500 mg/m3
Szakmary et al.
(1997)a
C57B1 Mice
Maternal toxicity (increased liver
weight); visceral malformations
Inhalation: 0, 1,500 mg/m3 on GDs
7-15
LOAEL = 1,500 mg/m3
Szakmary et al.
(1997)
NZW Rabbits
Maternal toxicity (decreased body
weight gain, increased liver weight);
abortions, total litter resorptions,
increased postimplantation loss,
malformations
Inhalation: 0, 4,500 mg/m3 on GDs
7-20
LOAEL = 4,500 mg/m3
Szakmary et al.
(1997)
S-D Rats
Maternal toxicity (decreased body
weight gain; decreased gravid uterine
weight); fetal body weight and
placental weight decrements, increased
delays in thoracic vertebral ossification
Inhalation: 0, 75, 250, or 600 ppm
(actual concentrations: 0, 66, 249,
600 ppm), 6 h/d, 7 d/wk, on GDs
0-19
Maternal LOAEL = 600 ppm
Fetal LOAEL = 250 ppm
Carney et al.,
(2006)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-349 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-38. Summary of mammalian developmental and reproductive
toxicity studies for tetrachloroethylene (continued)
Subjects
Effects
Concentration
Authors
Developmental neurotoxicity assessments
CFY Rats
Decreased postnatal survival, minimal
transient decreases in exploratory
activity and muscular strength, and
increased motor activity in females on
PND 100
Inhalation: 0, 1,500, 4,500 mg/m3 on
GDs 1-20 (and perhaps postnatally
to PND 100)
LOAEL = 1,500 mg/m3
Szakmary et al.
(1997)a
S-D Rats
Decreased weight gain, behavioral
changes (more extensive for late
pregnancy exposure), decreased brain
acetylcholine
Inhalation: 0, 100, 900 ppm on Days
7-13 or on Days 14-20
NOAEL =100 ppm
LOAEL = 900 ppm
Nelson et
al.(1980)
S-D Rats, two-
generation study
Behavioral effects (decreased activity;
reduced response to sound) in F1 pups
Inhalation: 0, 100, 300, 1,000 ppm
NOAEL = 300 ppm
LOAEL = 1,000 ppm
Tinston, (1994)b
NMRI Mice
Alterations in spontaneous motor
activity (locomotion, rearing, and total
activity) at PND 60
Gavage: 0, 5, 320 mg/kg-day on
PNDs 10-16
LOAEL = 5 mg/kg-day
Fredriksson et al.
(1993)
Reproductive toxicity studies
Rat (in vitro)
Reduced fertilizability of extracted
oocytes
12,000 mg/m3, 2 hours/d, 5 d/wk for
2 wk
Berger and Horner
(2003)
CD-I Mice
Abnormal sperm heads at 500 ppm but
not at 100 ppm, spermatogonia or
spermatocyte stage affected
Inhalation: 0, 100, 500 ppm for 5 d
LOAEL = 500 ppm
Beliles et al.,
(1980)
S-D Rats, two-
generation study
Increased death of F1A and F2A and
F2B pups, decreased body weight
Inhalation: 0, 100, 300, 1,000 ppm
NOAEL =100 ppm for
body weight reduction
Tinston (1994)b
a The Szakmary et al. (19971 study in CFY rats assessed both developmental toxicity and developmental
neurotoxicity outcomes.
b The Tinston (19941 study in S-D rats demonstrated both developmental neurotoxicity and reproductive
toxicity outcomes.
1 fertilizability, a developmental assay in Japanese medaka, and two oral gavage studies that
2 assessed developmental toxicity in rats and developmental neurotoxicity in mice.
3 Limitations of the inhalation developmental and reproductive toxicity studies are
4 described in the individual study summaries above. These limitations include the lack of dose-
5 response information due to the use of a single treatment level in the prenatal developmental
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-350 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
toxicity assessment by Schwetz et al.(1975); the lack of either maternal or developmental
toxicity in Hardin et al.(1981); absence of methodological details in study reporting (Szakmary et
al.. 1997); and a concern about a short peri-parturition exposure gap in Tinston (1994).
Additionally, the studies were conducted in accordance with standard EPA and OECD
toxicological study guidelines in place at the time but did not assess endpoints that are included
in the guidelines that were revised and harmonized in 1998 (e.g.. see Tinston. 1994). Maternal
toxicity, when observed, did not compromise the evaluation or interpretation of treatment-related
findings in the offspring.
The tetrachloroethylene database included assessments of the various potential
manifestations of developmental toxicity, i.e., alterations in survival, growth, morphology, and
functional development. Indications of effects on prenatal survival following in utero exposure
included increased pre- and/or postimplantation loss in rats, mice, and rabbits (Schwetz et al..
1975; Szakmary et al.. 1997). These findings were supported by evidence of embryo mortality in
a rat whole embryo culture (WEC) assay (Saillenfait et al.. 1995) and decreased viability in a
Japanese medaka assay (Spencer et al.. 2002). Decreased prenatal growth was observed in mice
(Schwetz et al.. 1975) and rats (Szakmary et al.. 1997). Morphological alterations associated
with prenatal exposures to tetrachloroethylene included delays in skeletal ossification in mice
(Schwetz et al.. 1975) and rats (Carney et al.. 2006; Szakmary et al.. 1997). which were often
associated with fetal weight decrements, and increased incidences of malformations in mice, rats,
and rabbits (Szakmary et al.. 1997). Evidence of tetrachloroethylene exposure-related
malformations was also observed in the rat WEC and Japanese medaka assays (Spencer et al..
2002)(Saillenfait et al., 1975) and in a gavage prenatal developmental toxicity screening study in
rats (Narotsky and Kavlock. 1995). Alterations in neurological function following pre- and/or
postnatal inhalation exposures to tetrachloroethylene were observed in rats by Szakmary et al.
(1997). Nelson et al. (1980). and Tinston (1994). These findings were supported by a study that
found altered spontaneous motor activity in young adult rats that had been treated orally with
tetrachloroethylene postnatally during a critical period of nervous system development
(Fredriksson et al.. 1993). Additionally, reductions in brain acetylcholine and dopamine were
observed in rat offspring following gestational tetrachloroethylene exposures (Nelson et al..
1980).
An assessment of fertility and reproductive function in rats exposed to
tetrachloroethylene via inhalation over the course of two generations was conducted by Tinston
(1994). Effects on offspring included decreased pup weights and postnatal survival in both
generations, as well as behavioral alterations in the F1 pups. Decreased mean testes weight was
observed in Fla males; however, no effects on male or female fertility or other evidence of
alterations in reproductive function were observed. For males, this finding is supported by the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-351 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
results of a study by Beliles et al. (1980). who found no sperm abnormalities in rats following up
to 10 weeks of tetrachloroethylene inhalation exposures. While the Beliles et al. (1980) study
identified an increase in abnormal sperm heads in mice after 4 weeks of exposure, no other
reproductive toxicity data in mice were available to aid in the interpretation of this finding.
In conclusion, based upon a consideration of the entire available database of animal
developmental and reproductive toxicity studies for tetrachloroethylene, the overall inhalation
NOAEL is 100 ppm, based on Tinston (1994). The overall inhalation LOAEL is 300 ppm, based
on Tinston (1994) and Schwetz et al. (1975). in which increased mortality and decreased body
weight of the offspring were observed.
Overall, the developmental and reproductive toxicity database for tetrachloroethylene
was judged to include a range of data from appropriate well-conducted studies in several
laboratory animal species plus limited human data and was considered sufficient for hazard
characterization and dose-response assessment, based upon EPA risk assessment guidelines
(U.S. EPA. 1991a. 2006b).
4.7.4. Mode of Action for Developmental Effects
Because of its lipid solubility, tetrachloroethylene can cross both the blood:brain barrier
and the placental barrier and, therefore, it can be present in all tissues, including the brain, during
development.
Peroxidation of the lipids of the cell membranes (Coiocel et al.. 1989). alteration of
regulation of fatty acid composition of the membrane (Kyrklund and Haglid. 1991). disturbances
in the properties of the nerve membrane (Juntunen. 1986). and progressively increased activity in
one or more of the phosphoinositide-linked neurotransmitters (Subramoniam et al.. 1989) have
all been suggested as MO As for neurotoxic effects. These mechanisms could be involved during
development phases, as well as in adults.
The metabolite TCA may be a causative agent or contribute to developmental toxicity
expressed as morphological changes, lethality, or growth reductions. Evidence in support of this
speculative position is presented in the following discussion. TCA is a weak organic acid, as are
many developmental toxicants, such as ethylhexanoic acid and valproic acid. These materials
accumulate to a greater extent in the embryo/fetal compartment than in the mother, based on the
pKa of the acid and the pH gradient between the maternal plasma and the embryo compartments
(O'Flahertv et al.. 1992). TCA could induce developmental toxicity by changing the intracellular
pH or through peroxisome proliferation. Ghantous et al. (1986) detected TCA in the amniotic
fluid of pregnant mice exposed to tetrachloroethylene via inhalation.
Smith et al. (1989) found that oral gavage doses of TCA (330, 800, 1,200, and
1,800 mg/kg-day) delivered on GDs 6-15 to pregnant Long-Evans rats produced soft tissue
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-352 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
malformations, principally in the cardiovascular system. Johnson et al. (1998) found cardiac
defects in rat fetuses whose mothers received 2,730-ppm TCA in drinking water during the
period of cardiac development. Saillenfait et al. (1995). using the rat whole embryo (Day 10)
culture system, found that both tetrachloroethylene and TCA induced embryo toxicity, including
mortality, malformations, and delayed growth and differentiation. TCA produced a reduction in
the first branchial arch as well as other morphological changes at a lower concentration (2.5 mM)
than that at which tetrachloroethylene induced no adverse effect (3.5 mM). TCA also induced a
reduction of the yolk sac diameter at 1 mM.
Arguments counter to the involvement of TCA in the MOA for tetrachloroethylene
developmental toxicity include that the types of malformations associated with TCA [i.e., cardiac
malformations reported by Smith et al. (1989) and Johnson et al. (1998)1 or other weak acid
exposures [e.g., valproic acid and ethylhexanoic acid; Scott et al. (1994)1 are not consistent with
those observed in tetrachloroethylene studies. Additionally, relatively high concentrations of
TCA are required to cause developmental toxicity compared with the concentration expected to
result from metabolism of tetrachloroethylene in vivo, which may account for the differences in
the type of developmental effects resulting from tetrachloroethylene exposure. There is also a
lack of information on the availability of metabolized TCA to the developing fetus and the
potential differences related to oral-versus-inhalation exposure in tetrachloroethylene studies.
48 GENOTOXICITY
Tetrachloroethylene and its metabolites have been extensively studied for genotoxic
activity in a variety of in vitro assay systems such as bacteria, yeast, and mammalian cells (See
reviews by AT SDR. 1997a: Breslow and Day. 1994; IARC. 1995; U.S. EPA. 1985a; U.S. EPA.
1991b; WHO. 2006). This section discusses the genotoxic potential of tetrachloroethylene and
its known or postulated metabolites (TCA, DCA, CH, TCVC, TCVG, NAcTCVC,
tetrachloroethylene epoxide), with a summary provided at the end of each section for
tetrachloroethylene or its metabolite for their mutagenic potential, in addition to an overall
synthesis summary at the end of this section. TCVC sulfoxide does not appear to have been
investigated for genotoxicity.
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 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, which includes proteins associated with DNA or RNA, can also
cause transmissible changes. Genetic alterations can occur through a variety of mechanisms
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-353 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
including gene mutations, deletions, translocations, or amplifications; evidence of mutagenesis
provides mechanistic support for the inference of potential for carcinogenicity in humans.
Evaluation of genotoxicity data entails a weight-of-evidence approach that includes
consideration of the various types of genetic damage that can occur. In acknowledging that
genotoxicity tests are, by design, complementary evaluations of different mechanisms of
genotoxicity, a recent IPCS publication (Eastmond et al.. 2009) notes that —raltiple negative
results may not be sufficient to remove concern for mutagenicity raised by a clear positive result
in a single mutagenicity assay." These considerations inform the present approach. In addition,
consistent with EPA's Guidelines on Carcinogenic Risk Assessment and Supplemental Guidance
for Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA. 2005a. c), the
approach does not address relative potency (e.g., among tetrachloroethylene metabolites, or of
such metabolites with other known genotoxic carcinogens) per se, nor does it consider
quantitative issues related to the probable production of these metabolites in vivo. Instead, the
analysis of genetic toxicity data presented here focuses on the identification of a genotoxic
hazard of these metabolites; a quantitative analysis of tetrachloroethylene metabolism to reactive
intermediates, via PBPK modeling, is presented in Section 3.
Below, the genotoxicity data for tetrachloroethylene and its metabolites are briefly
reviewed, with detailed study information in the corresponding tables. The contributions of
these data are twofold. First, to the extent that these metabolites may be formed in the in vitro
and in vivo test systems for tetrachloroethylene, these data provide insight into what agent or
agents may contribute to the limited activity observed with tetrachloroethylene in these
genotoxicity assays. Second, because the in vitro systems do not necessarily fully recapitulate in
vivo metabolism, the demonstration of in vitro genotoxicity by the known in vivo metabolites
themselves provides information regarding the expected genotoxicity of tetrachloroethylene
following in vivo exposure.
4.8.1. Tetrachloroethylene (PCE)
Limited studies have been performed examining tetrachloroethylene genotoxicity in vivo.
These and in vitro genotoxicity studies of tetrachloroethylene are described below and listed in
Tables 4-39 and 4-40.
4.8.1.1. Mammalian Systems (Including Human Studies)
4.8.1.1.1. Gene mutation
Tetrachloroethylene was negative for increased frequency of mutations of thymidine
kinase locus in L5178Y/TK +/- mouse lymphoma cells both with and without S9 activation
(F344 rat liver) (NTP. 1986b). Experiments were performed twice, with replicates of all doses.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-354 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
L5178Y/TK +/- mouse lymphoma cells were exposed to tetrachloroethylene in 1%
dimethylsufoxide for 4 hours at 37°C in medium; cells were then washed and resuspended in
fresh medium for 48 hours at 37°C. TK mutation frequency was determined by plating cells in
medium supplemented with trifluorothymidine. Overall cell viability was determined by plating
cells in nonselective medium. Mutation frequency was not above background for any dose
tested (6.25, 12.50, 25, 50, 100 nL/mL in the presence of S9; 12.5, 25, 50, 75, and 150 nL/mL in
the absence of S9). Positive controls in both the presence and absence of S9 activation
[3-methylcholanthrene (2.5 |ig/mL) and ethyl methanesulfonate (250 |ig/mL), respectively]
showed significant increases in mutation frequencies (p < 0.001, /-test) (NTP. 1986b).
Gene mutations were induced in a host-mediated assay, using S. typhimurium strain
TA98 implanted into the peritoneal cavity of male and female CD-I mice that were previously
exposed to tetrachloroethylene by inhalation (100 or 500 ppm, 7 hours/day, for 5 days) (Bellies
et al.. 1980). Positive results were observed in male mice at 100 (but not 500) ppm, and in female
mice at 500 (but not 100) ppm. Although no explanation was given for the variability in the dose
response, the authors conclude that tetrachloroethylene is an active frameshift mutagen using in
vivo activation.
In summary, the in vitro thymidine kinase gene mutation assay in mammalian cells was
negative for gene mutations in the presence and absence of S9 (F344 rat liver) metabolic
activation (NTP. 1986b). Positive results for frameshift mutagenicity were observed in a host-
mediated assay by implanting S. typhimurium into mice exposed to tetrachloroethylene, but
without a clear dose-response effect (Beliles et al.. 1980).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-355 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-39. Genotoxicity of tetrachloroethylene—mammalian systems (in
vitro and in vivo)a
Test system/endpoint
Doses
(LED or HID)b
Results0
Reference
With
activation
Without
activation
Unscheduled DNA synthesis, rat primary
hepatocytes in vitro
166 (vapor)
NT
d
Shimada et al.
(1985)
Unscheduled DNA synthesis, Osborne Mendel
rat primary hepatocytes in vitro
NA
NT
—
Milman et al.,
(1988)
Unscheduled DNA synthesis, B6C3Fi mouse
primary hepatocytes in vitro
NA
NT
—
Milman et al.,
(1988)
Gene mutation, mouse lymphoma L5178Y
cells, tk locus
245
-
-
NTP (1986b)
Sister chromatid exchange, Chinese hamster
ovary (CHO) cells in vitro
164
-
-
Galloway et al.,
(1987)
Chromosomal aberrations, Chinese hamster
lung (CHL) cells in vitro
500
-
-
Sofuni et al.,
(1985)
Chromosomal aberrations, Chinese hamster
ovary (CHO) cells in vitro
136
-
-
Galloway et al.,
(1987)
Cell transformation, RLV/Fischer rat embryo
F1706 cells in vitro
16
NT
+
Price et al.,
(1978)
BALB/c-3T3 mouse cells, cell transformation
in vitro
250
NT
-
Tu et al., (1985)
Rat and mouse hepatocyte, DNA damage
(unscheduled DNA synthesis)
2.5mM
NT
Costa and
Ivanetich,
(1984)
Human fibroblast cells, DNA damage
(unscheduled DNA synthesis)
0.1 nL/mL
(+/-)
(+/")
Beliles et al.
(1980)
Host mediated assay—S. typhimurium
implanted in CD-I mice
100 ppm (male mice)
500 ppm (female mice)
+
NT
Beliles et al.
(1980)
Chinese hamster ovary cells, sister chromatid
exchange
164 (ig/mL
-
-
NTP (1986b)
Chinese hamster ovary (CHO-K1) cells,
increased frequency of micronuclei
~63 ppm
NT
+
Wang et al.,
(2001)
Cytochalasin B-blocked
micronucleus assay using human
lymphoblastoid cell lines with
enhanced metabolic activity,
increased frequency of
micronuclei
AHH-1
5 mM
NT
+
Doherty et al.,
(1996)
H2E1
1 mM
NT
+
Doherty et al.,
(1996)
MCL-5
1 mM
NT
+
Doherty et al.,
(1996)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-356 DRAFT-DO NOT CITE OR QUOTE
-------
Human white blood cells, length of DNA
5 x l(T3 M
-
-
Hartmann and
migration
Speit r 19951
Human lymphocytes, sister chromatid
-
-
exchange
Table 4-39. Genotoxicity of tetrachloroethylene—mammalian systems (in
vitro and in vivo)a (continued)
Test system/endpoint
Doses
(LED or HID)b
Results0
Reference
With
activation
Without
activation
Gene conversion and reverse mutation in
S. cerevisiae D7 recovered from liver, lungs,
and kidneys of CD-I mice
11,000 p.o. x 1
NT
Bronzetti et al.
(1983)
Gene conversion and reverse mutation in
S. cerevisiae D7 recovered from liver, lungs,
and kidneys of CD-I mice
2,000 p.o. x 12
NT
Bronzetti et al.
(1983)
DNA single-strand breaks (alkaline unwinding)
in liver and kidney of male NMRI mice in vivo
660 i.p. x 1
NT
+e
Walles (1986)
Sister chromatid exchange, human
lymphocytes in vivo
1,500 mg/m3 inhaled
NT
-
Ikeda et al.
(1980)
Chromosomal aberrations, human lymphocytes
in vivo
92 ppm inhaled
NT
-
Ikeda et al.
(1980)
Binding (covalent) to calf thymus DNA in
vitro
2.5 jxCi 14C-PCE
+
Data not
shown
Mazzullo et al.
(1987)
Binding (covalent) to DNA in male B6C3Fi
mouse liver in vivo
1,400 inhaled 6 h
600 ppm
NT
-
Schumann et al.
(1980)
Binding (covalent) to DNA in male B6C3F,
mouse liver in vivo
500 p.o. x 1
NT
-
Schumann et al.
(1980)
Binding (covalent) to DNA in male BALB/c
mouse and Wistar rat liver, kidney, lung, and
stomach in vivo
1.4 i.p. x l
22 h
NT
+
Mazzullo et al.
(1987)
Binding (covalent) to RNA and protein in male
BALB/c mouse and Wistar rat liver, kidney,
lung, and stomach in vivo
1.4 i.p. x l
22 h
NT
+
Mazzullo et al.
(1987)
Human lymphocytes, sister chromatid
exchange
10 ppm (geometric
mean)
NT
-
Seiji et al.
(1990)
Mouse, reticulocytes, micronucleus
2,000 mg/kg
NT
Murakami and
Horikawa
(1995)
Mouse, hepatocytes, micronucleus
Before partial hepatectomy
After partial hepatectomy
1,000 mg/kg
NT
+
Murakami and
Horikawa
(1995)
Mouse, induction of DNA damage in
hepatocytes (alkaline Comet assay)
1,000 mg/kg-day
2,000 mg/kg-day
NT
+/-
+/-
Cederberg et
al.. (2010)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-357 DRAFT-DO NOT CITE OR QUOTE
-------
Mouse, induction of DNA damage in kidney
(alkaline Comet assay)
1,000 mg/kg-day
2,000 mg/kg-day
NT
-
Cederberg et
al.. (2010)
Rat bone marrow cells, chromosomal
aberrations
100 and 500 ppm
NT
-
Beliles et al.
(1980)
Enzyme-altered foci in male Osborne Mendel
rat liver in vivo, promotion protocol, with or
without N-nitrosodiethylamine as an initiator
1,000, 5 d/wk for 7 wk
NT
+
Milman et al.,
(1988)
Enzyme-altered foci in male Osborne Mendel
rat liver in vivo, initiation protocol,
phenobarbital as a promoter
1,000
NT
Milman et al.,
(1988)
Micronucleus induction (Chinese hamster lung
cell line)
250 (ig/mL
-
-
Matsushima et
al. (1999s)
Gap Junction Intercellular Communication (rat
liver cells)
0.1 mM
NT
+
Benane et al.
(1996)
DNA damage (80HdG) in urine and
leukocytes of dry cleaners (female only)
3.8 ±5.3 ppm (TWA)
NT
-
Toraason et al.
(2003)
DNA damage (80HdG) in Fischer rats
measured in urine, lymphocytes, and liver
100-1,000 mg/kg
NT
(Substantial
morbidity at
all doses limits
interpretation.)
Toraason et al.
(1999)
Human lymphocytes in vitro
(unscheduled DNA synthesis)
1 mM
-
-
Perocco et al.
(1983)
Human lymphocytes in vivo
(Chromosomal aberrations)
144 mg/m3 (but
contaminated with
trichloroethylene)
NT
+
Fender(1993)
DNA single-strand breaks
1,000 mg/kg p.o.
NT
—
Potter et al.
(1996)
aTable adapted from ATSDR (1997a) and IARC monograph (19951 and modified/updated for newer references.
bLED, lowest effective dose; HID, highest ineffective dose; doses are in ng/mL for in vitro tests; mg/kg for in vivo
tests unless otherwise specified; i.p. = intraperitoneal; p.o. = oral; NA = not available.
"Results: + = positive; (+) = weakly positive; (+/-) = mixed results; - = negative; NT = not tested.
dPCE with stabilizers was positive with and without metabolic activation.
"Negative in lung.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-358 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-40. Genotoxicity of tetrachloroethylene—bacterial, yeast, and fungal
systems"
Test system/endpoint
Doses
(LED or HID)b
Results0
Reference
With
activation
Without
activation
SOS chromotest, E. coli PQ37
8,150
-
-
Mersch-Sundermann et
al. (1994)
SOS chromotest, E. coli PQ37
NA
-
-
von der Hude et al.,
(1988)
X Prophage induction, E. coli WP2
10,000
-
-
DeMarini et al.. C1994)
S. typhimurium BAL13, forward mutation
(ara test)
76
-
-
Roldan-Arjona et al.,
(1991s)
S. typhimurium TA100, reverse mutation
660
-
-
Bartsch et al.. (1979)
S. typhimurium TA100, reverse mutation
167
-
-
Haworth et al.. (1983)
S. typhimurium TA100, reverse mutation
1,000
-
-
Connor et al.. (1985)
S. typhimurium TA100, reverse mutation
166 (vapor)
-
d
Shimada et al.. (1985)
S. typhimurium TA100, reverse mutation
NA
-
—
Milman et al.. (1988)
S. typhimurium TA100, reverse mutation
332
+e
—
Vamvakas et al.,
(1989c)
S. typhimurium TA100, reverse mutation
1.3 (vapor)
-
-
DeMarini et al.. (1994)
S. typhimurium TA1535, reverse mutation
50
NT
-
Krinestad et al.. (1981)
S. typhimurium TA1535, reverse mutation
167
-
-
Haworth et al.. (1983)
S. typhimurium TA1535, reverse mutation
66 (vapor)
(+)
d
Shimada et al.. C1985)
S. typhimurium TA1535, reverse mutation
NA
-
—
Milman et al.. C1988)
S. typhimurium TA1537, reverse mutation
167
-
—
Haworth et al.. C1983)
S. typhimurium TA1537, reverse mutation
NA
-
-
Milman et al..(1988)
S. typhimurium, gene mutation TA100,
TA1535, TA1537, TA98
333 |ig/platc
-
-
NTP (1986b)
S. typhimurium TA98, reverse mutation
167
-
-
Haworth et al.. (1983)
S. typhimurium TA98, reverse mutation
1,000
-
-
Connor et al.. (1985)
S. typhimurium TA98, reverse mutation
NA
-
-
Milman et al.. (1988)
S. typhimurium UTH8413, reverse mutation
1,000
-
-
Connor et al.. (1985)
S. typhimurium UTH8414, reverse mutation
1,000
-
-
Connor et al.. (1985)
S. typhimurium TA102, TA2638
E. coli WP2/pKM101, WP2 uvrA/pKMlOl,
gene mutation
1,250 ng/plate
NT
Watanabe et al.. CI998)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-359 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-40. Genotoxicity of tetrachloroethylene—bacterial, yeast, and fungal
systems" (continued)
Test system/endpoint
Doses
(LED or HID)b
Results0
Reference
With
activation
Without
activation
S. typhimurium, YG7108pin3ERb5, gene
mutation (strain is methyltransferase
deficient and stably expresses complete
electron transport chain including P450
reductase, cytochrome b5 and CYP2E1)
200 ng/plate
NT
Emmert et al.. (2006)
E. coli K12, forward mutation
150
-
-
Greim et al.. (1975)
E. coli K12, reverse mutation (arg*)
150
-
-
Greim et al.. (1975)
E. coli K12, reverse mutation (gal*)
150
-
-
Greim et al.. (1975)
E. coli K12, reverse mutation (nad*)
150
-
-
Greim et al.. (1975)
S. cerevisiae D7, log-phase cultures, gene
conversion
1,100
NT
+
Callen et al., (1980)
S. cerevisiae D7, gene conversion
9,960
-
-
Bronzetti et al.,
(1983)
S. cerevisiae D7, log-phase and stationary
cultures, gene conversion
2,440
-
-
Koch et al., (1988)
S. cerevisiae D7, log-phase cultures,
mitotic recombination or other genetic
alterations (ade2)
1,100
NT
+
Callen etal., (1980)
S. cerevisiae D7, mitotic recombination
9,960
-
-
Bronzetti et al.,
(1983)
S. cerevisiae D7, log-phase cultures,
reverse mutation
810
NT
(+)
Callen etal., (1980)
S. cerevisiae D7, reverse mutation
9,960
-
-
Bronzetti et al.,
(1983)
S. cerevisiae D7, log-phase and stationary
cultures, reverse mutation
2,440
-
-
Koch etal., (1988)
S. cerevisiae D61.M, growing cells,
aneuploidy
810
(+)
(+)
Koch etal., (1988)
D. melanogaster, sex-linked recessive
lethal mutation
4,000 ppm p.o.
1,000 ppm injection
NT
-
NTP (1986b)
D. melanogaster, sex-linked recessive
lethal mutation
3,400 mg/m3, 7 h
NT
—
Beliles et al. (1980)
aTable adapted from ATSDR (1997a) and I ARC monograph (1995) and modified/updated for newer
references.
bLED, lowest effective dose; HID, highest ineffective dose; doses are in ng/mL for in vitro tests unless
otherwise specified; NA = not available.
°Results: + = positive; (+) = weakly positive; - = negative; NT = not tested.
dPCE with stabilizers was positive with and without metabolic activation.
eWeak increase in activity with rat liver S9, rat kidney microsomes and glutathione (GSH): fourfold
increase with rat kidney microsomes, GSH and GSH ^-transferase.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-360 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
4.8.1.1.2. DNA binding
Schumann et al. (19801 assessed hepatic macromolecular binding in both rats and mice
exposed to radiolabeled tetrachloroethylene by inhalation (10 or 600 ppm, 6 hours; binding
measured at 6, 24, 48, and 72 hours postexposure) or a single oral gavage (500 mg/kg in corn oil;
binding measured at 1, 6, 12, 24, 48, and 72 hours). In mice, tetrachloroethylene binding to
macromolecules in liver peaked at the termination of the inhalation exposure or 6 hours postoral
exposure. In rats, hepatic macromolecular binding peaked 24 hours after either oral or inhalation
exposure. At these peak times, no DNA binding was observed in the mouse (rat data not
reported). Using a more sensitive assay, Mazzullo et al. (1987) reported low levels of DNA
binding (2.9 pmol/mg) in mouse liver 22 hours after i.p. injection (1.4 mg/kg bw). Levels of
DNA binding were 6- to 10-fold lower in rat liver and in the kidney, lung, and stomach of mice
and rats. Binding to RNA or protein was considerably higher than binding to DNA in both mice
and rats. This raises the concern that possible contamination with RNA or protein might have
contributed to the DNA results. Protein binding levels were highest in mouse liver and rat
kidney. In a companion in vitro study, binding to calf thymus DNA was increased by
microsomal fractions from rat or mouse liver, but not kidney, lung, or stomach. Cytosolic
fractions from rat or mouse liver, kidney, lung, or stomach also enhanced DNA binding in vitro,
with mouse and rat liver and mouse lung fractions being the most efficient. Cytosolic and
microsomal fractions, when combined, enhanced DNA binding to a comparable extent as
cytosolic fractions alone. Phenobarbital pretreatment of animals increased cytosol-mediated
binding but minimally affected microsomal-mediated binding. DNA binding by rat liver
microsomal fraction was enhanced 17-fold by GSH but decreased by superoxide dismutase or
mannitol (Mazzullo et al.. 1987).
In summary, DNA binding was not observed in one assay in mice exposed to
tetrachloroethylene by inhalation and oral routes, while protein and RNA binding was observed
(Schumann et al.. 1980). Low levels of DNA binding in mouse liver, and yet lower levels in
mouse kidney or rat and mouse stomach, were observed after i.p. injection using a more sensitive
assay (Mazzullo et al.. 1987). In vitro binding to calf thymus DNA was enhanced by
microsomal and cytosolic fractions from various mouse and rat tissues. These results suggest a
role for metabolic activation of the parent compound in DNA binding in vitro.
4.8.1.1.3. Chromosomal aberrations
NIOSH (1980) assessed bone marrow chromosomal aberrations and aneuploidy in male
and female Sprague-Dawley rats after acute (sacrificed 6, 24, or 48 hours after dosing) and
subchronic (7 hours a day, for 5 days; sacrificed 6 hours after last exposure) exposures to
tetrachloroethylene by inhalation (100 and 500 ppm). The only effect reported with acute
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-361 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
exposure was a slight increase in the percentage of cells with aberrations and aneuploidy (peak
of 3.3% compared to 0.7% in controls with 500-ppm tetrachloroethylene) in male, but not
female, rats. No significant effects were observed in any subchronically exposed groups, but
female rats showed a nonsignificant increase in cells with aberrations (Beliles et al.. 1980). NTP
(1986b) did not observe chromosomal aberrations in Chinese hamster ovary cells exposed to
tetrachloroethylene (17, 34.1, 68.1, and 136.3 [j,g/mL without activation or 17, 34.1, and
68.1 [j,g/mL with activation by Sprague-Dawley rat liver S9).
4.8.1.1.3.1. Micronucleus induction
Tetrachloroethylene exposure increased the frequency of micronuclei in hepatocytes, but
not peripheral blood reticulocytes, of ddY mice given single i.p. injections of 1,000- or 2,000-
mg/kg tetrachloroethylene after, but not prior to, partial hepatectomy (Murakami and Horikawa.
1995). This twofold increase in micronuclei in hepatocytes after partial hepatectomy was
statistically significant but was not evident at the lower dose of 500 mg/kg. Conflicting results
of other studies of tetrachloroethylene micronuclei induction have also been reported in cultured
Chinese hamster cells (Matsushima et al.. 1999; Wang et al.. 2001) and in human cells (Dohertv
et al.. 1996; White et al.. 2001). Micronucleus induction was not observed in a Chinese hamster
lung cell line (CHL/IU) following exposure to high doses of tetrachloroethylene
(125-250 |ig/inL) as part of a test validation assay, but some induction (not statistically
significant) was observed at the lower dose (75 jig/mL) in the presence of S9 fraction
(Matsushima et al.. 1999). Details from this study are limited. Wang et al. (2001) examined
micronuclei induction following in vitro exposure to tetrachloroethylene (-63 ppm in culture
medium at peak) in a closed system. Chinese hamster ovary (CHO-K1) cells were plated in a
petri dish surrounding a glass dish of tetrachloroethylene and incubated for 24 hours.
Tetrachloroethylene exposure led to a dose-dependent significant increase in micronuclei
induction (p < 0.001) (Wang et al.. 2001). Similar results were also observed in human cell lines
in other studies.
Micronucleus induction was enhanced by tetrachloroethylene exposure in AHH-1
parental human lymphoblastoid cells, and in two daughter cell lines (h2El and MCL-5) stably
expressing human metabolic enzymes lines (Dohertv et al.. 1996). Parental AHH-1 cells possess
native, albeit low, CYP1A1 activity but considerable glutathiones-transferase activity; h2El
cells stably express human CYP2E1; and MCL-5 cells stably express human CYP1A2, 2A6,
3A4, 2E1, and microsomal epoxide hydrolase. Tetrachloroethylene (5 mM) induced a threefold
increase in micronuclei in AHH-1 cells and ninefold increases in h2El and MCL-5 cells,
respectively (Doherty et al., 1996). White et al. (2001) similarly observed dose-dependent
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-362 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
increases in micronuclei induction after 24 hours incubation (p < 0.05) with tetrachloroethylene
(0, 0.01, 0.05, 0.1, 0.25, 0.5, 1.0, 2.0 mM) intheMCL-5 cell line.
4.8.1.1.3.2. Sister chromatid exchanges (SCEs)
Limited studies of sister chromatid exchanges demonstrate conflicting results. No
differences were observed in the frequency of chromosomal aberrations and SCE between
unexposed workers, workers exposed to moderate levels of tetrachloroethylene (70-280mg/m3),
and those exposed to high doses (200-1,500 mg/m3) (Ikeda et al.. 1980). Although an exposure
assessment was performed in this study, the results are limited by the small number of subjects
(total n= 19). Another study from this group had similar limitations (total n = 10) and also
found no sister chromatid exchanges in lymphocytes in workers occupationally exposed to either
high-dose tetrachloroethylene (92 ppm, geometric mean) or low-dose tetrachloroethylene (10-40
ppm range) (Ikeda et al.. 1980). Similarly, no differences were observed between exposed and
controls in a larger Japanese study, which examined SCE in 27 occupationally exposed workers
(Seiji et al.. 1990). or a German study on dry-cleaning workers (Bottger and Elstermeier. 1989).
Increased chromosomal aberrations were observed in another occupational study following
exposure to tetrachloroethylene (144-348 mg/m3); however, exposure also included a small
amount of trichloroethylene (0.11-0.43% by wt), so interpretation of the results relative to
tetrachloroethylene alone may be limited (Fender. 1993)
Tetrachloroethylene-induced damage was also not observed in the sister chromatid
exchange (SCE) assay or in the single-cell gel test (i.e., the Comet assay) in cultured human
blood exposed to up to 5-mM (~830-mg/L) tetrachloroethylene, a dose that reduced viability by
40% due to cytotoxicity (Hartmann and Speit. 1995). Neither chromosome aberrations nor SCE
were induced in Chinese hamster ovary cells following in vitro exposure to tetrachloroethylene
(Galloway et al.. 1987; Sofuni et al.. 1985) as summarized in (NRC. 2010). Chinese hamster
ovary cells exposed to tetrachloroethylene (16.4, 54.5, or 164 (j,g/mL) in the presence and
absence of S9 activation (Sprague-Dawley rat livers) showed no increase in frequency of sister
chromatid exchanges following exposure to tetrachloroethylene (NTP. 1986b).
In summary, the majority of studies of chromosomal aberrations, micronuclei induction,
and sister chromatid exchange following exposure to tetrachloroethylene are negative. Positive
micronuclei induction was observed following partial hepatectomy at high doses
(2,000 mg/kg-day i.p.) in ddY mice (Murakami and Horikawa. 1995). Increased micronuclei
induction was observed in CHO cells in vitro when exposed to tetrachloroethylene in a closed
system (Wang et al.. 2001) but not in CHL cells when exposed in an open system (Matsushima
et al.. 1999). suggesting the need to control for loss of tetrachloroethylene via vaporization in in
vitro assays. Dose-dependent increases in micronuclei were observed in human lymphoblastoid
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-363 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
cell lines, an effect enhanced by stable expression of CYP450 enzymes (Dohertv et al.. 1996;
White et al.. 2001); however, these cell lines are not generally considered part of the standard
genotoxicity testing battery. No in vitro studies of tetrachloroethylene (Galloway et al.. 1987;
Hartmann and Speit. 1995; NTP. 1986b) and only one occupational exposure study of exposures
to tetrachloroethylene and trichloroethylene (Fender. 1993) reported sister chromatid exchanges.
4.8.1.1.4. Unscheduled DNA synthesis
Human fibroblasts (WI-38 cells) were assayed for unscheduled DNA synthesis following
exposure to tetrachloroethylene (0.1 to 5.0 [xL/mL), but the results were equivocal, with results at
low doses similar to the positive controls and negative results at high doses, but it is noted that
the high doses yielded considerable cytotoxicity (Beliles et al.. 1980). The positive controls
were only weakly positive, as described based on the laboratory criteria (criteria details not
given). No evidence of unscheduled DNA synthesis was observed in human lymphocytes,
human fibroblasts, or rat and mouse hepatocytes (Costa and Ivanetich. 1984; Milman et al..
1988; Perocco et al.. 1983; Shimada et al.. 1985). In summary, UDS was not statistically
significantly increased in any published studies, although some increases were observed in one
study (Beliles et al.. 1980).
4.8.1.1.5. DNA strand breaks
An increased level of DNA single-strand breaks (SSB), as assessed by a DNA unwinding
technique, was seen in liver and kidney tissues but not in the lung tissue of male NMRI mice
1 hour after single i.p. injections in Tween 80 of 4-8 mmol/kg (663-1,326 mg/kg) of
tetrachloroethylene (Walles. 1986). This effect was reversible as early as 24 hours postexposure,
presumably by DNA repair. Limitations of i.p. injection include the potential inflammatory
effect at the site of injection, which could, in turn, lead to production of reactive oxygen species
and other inflammatory mediators. These could lead to an increase in DNA damage unrelated to
the specific exposure. Potter et al. (1996) found no increases in DNA strand breaks, when
assessed by an alkaline unwinding procedure, in kidneys of male F344 rats assessed after daily
oral gavage treatment with 1,000-mg/kg tetrachloroethylene for 7 days. A more recent study
(Cederberg et al.. 2010) found oral gavage exposure to tetrachloroethylene (1,000 or
2,000 mg/kg-day given as two administrations, 24 hours apart, in corn oil) led to slight increases
(1.3- and 1.4-fold as compared to control) in DNA damage in liver (but not kidney) of CD1 mice
as measured by the alkaline Comet assay when tissues were sampled 3 hours after the last
administration. Others have interpreted these data to demonstrate a lack of DNA damage in the
liver and kidney of CD1 mice after oral tetrachloroethylene exposure (presented in 2003; and in
Lillford et al.. 20101. Cederberg et al. (2010) reported a statistically significant dose-related
increase in tail intensity (p = 0.041; one-sided Jonckheere-Terpstra test using exact permutation)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-364 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
in the liver following exposure to PCE. The authors note that 8 of 12 tetrachloroethylene-
exposed animals had higher tail intensity values than the highest value in the controls, a finding
significant by the Fisher exact probability test (p = 0.013). No statistically significant effects
were observed for tail moment in the liver, or for either tail intensity or moment in the kidney.
The alternative interpretation is that the variability between mice in the treatment groups and the
low magnitude of the response in the tetrachloroethylene-dosed animals does not support the
conclusion that tetrachloroethylene induced DNA damage in this study. This interpretation is
supported by the lack of statistical significance when the results are analyzed by Dunnett's test
for pairwise comparisons. Cederberg et al. (2010) argue that the interindividual animal
variability is not exceptionally large, and that the Dunnett's pairwise test has less power than the
trend test of Jonckheere-Terpstra (Cederberg et al.. 2010). Further discussion of this publication
in the literature is ongoing (Lillford et al.. 2010; Lovell. 2010). Lillford et al. (2010) give
additional details on the alternative interpretation described in the original paper, stating also that
the limited biological significance of these slight increases in tail intensity needs to be taken into
account. This paper states that the results described in the original study are within the range of
historical controls in the study laboratory. Lillford et al. (2010) endorse the use of the parametric
test for statistical analysis (Dunnett's), which showed no statistical significance for the results
reported in Cederberg. The third publication discusses the use of various statistical analyses
used in the two interpretations (Lovell. 2010). Overall, Lovell (2010) states that it is not a
question of one statistical analysis being right and the other wrong; it is more a question of using
the best statistical analysis for the hypothesis being tested. The different approaches show a
contrast between a powerful trend test and a more conservative pairwise comparison. Lovell
(2010) also commented on the magnitude of the response as it relates to biological relevance.
Further studies, as suggested by Cederberg et al. (2010),may or may not address this issue if
carried out the same way as the original study. Finally, both Lillford et al. (2010) and Lovell
(2010) agree that the statistical analysis utilized should not be used as the sole determinant of
how the results of this, or any study, are interpreted.
In summary, the results of the limited DNA strand break assays following exposure to
tetrachloroethylene are equivocal. Walles (1986)demonstrated DNA single-strand breaks in the
liver and kidney of male mice exposed by i.p. injection, but this was reversible within 24 hours.
A second study examined DNA strand breaks after 1 week oral exposure to tetrachloroethylene
and demonstrated no DNA damage (Potter et al.. 1996). A recently published report on DNA
strand breaks showed a marginal increase in only one parameter from the Comet assay (tail
length) following oral exposure to tetrachloroethylene in mice (Cederberg et al.. 2010). but the
statistical and biological significance of this result has been disputed (Cederberg et al.. 2010;
Lillford et al.. 2010; Lovell. 2010).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-365 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
4.8.1.1.6. DNA damage related to oxidative stress
Toraason et al. (2003) reported no increase in leukocyte 8-OHdG in 18 dry-cleaner
workers compared with 20 launderers, and reported no increase in urinary 8-OHdG among the
dry-cleaner workers sampled pre- and postshift work (time-weighted average [TWA]
concentration of tetrachloroethylene was 3.8 ± 5.3 ppm). Under the conditions of this study, no
evidence of oxidative DNA damage was found. Toraason et al. (1999) measured 8-OHdG and a
—freeadical-catalyzed isomer of arachidonic acid and marker of oxidative damage to cell
membranes, 8-Epi-prostaglandin F2a (8-epiPGF)," 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 of tetrachloroethylene in Alkamuls vehicle. Male
Fischer rats sacrificed 24 hours after a single i.p. injection of tetrachloroethylene (0, 100, 500, or
1,000 mg/kg) showed no significant increases in 8-OHdG in liver, lymphocytes, or urine
(Toraason et al.. 1999). Lipid peroxidation of the liver (as measured by TBARS) was also not
observed following a single exposure to tetrachloroethylene. However, the authors reported
morbidity and mortality with a single 500-mg/kg tetrachloroethylene exposure inducing Stage II
anesthesia (loss of righting reflex but maintained reflex response) and a single 1,000-mg/kg
tetrachloroethylene exposure inducing Level III or IV (absence of reflex response) anesthesia
and burgundy-colored urine during the first 12 hours of collection. Although none of the rats
exposed to 1,000-mg/kg tetrachloroethylene died from treatment, the authors state that some in
this high-dose group would not have survived another 24 hours. Thus, using this paradigm, there
was significant 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.
Although the authors suggest that evidence of oxidative damage was equivocal, the effects on
urine volume and water consumption, as well as the profound toxicity induced by this exposure
paradigm, limit interpretation of these data. In summary, the limited studies examining DNA
adduct formation related to oxidative stress are inconclusive, with no results in the urine or
leukocytes of occupationally exposed individuals and limited utility of the animal study due to
significant toxicity in the exposed animals.
4.8.1.1.7. Cell transformation
Tetrachloroethylene exposure did not lead to cell transformation in BALB/c-3T3 cells
after 3-day exposure (0, 1, 10, 100, and 250 (j,g/mL) followed by a 30-day incubation period (Tu
et al.. 1985). Exposure to tetrachloroethylene (study details not given) was also negative for cell
transformation in BALB/c-3T3 cells (Milman et al.. 1988). However, Fischer rat embryo cells
were transformed in the absence of metabolic activation (Price et al.. 1978).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-366 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
4.8.1.1.8. Gap junction intercellular communication
One assay examined gap junction intercellular communication following exposure to
tetrachloroethylene in rat liver cells (0, 0.01, 0.1, and 1 mM at 0, 1, 4, 6, 24, 48, and 168 hours)
(Benane et al.. 1996). Communication was inhibited following exposure to 0.1-mM
tetrachloroethylene at 48 hours and continued at the final time point tested (168 hours). This
study also examined tetrachloroethylene metabolites, including DCA, TCA, CH, and
trichloroethanol. These metabolites also led to decreases in intercellular communication, but to
varying levels.
4.8.1.1.9. Tumor initiation
Milman et al. (1988) reported a statistically significant increase (p < 0.01) in
y-glutamyltranspeptidase-positive liver foci in a promotion, but not in an initiation, test protocol
in male Osborne-Mendel rats. Initiation capacity was tested by exposing 10 rats to 1,000-mg/kg
tetrachloroethylene after partial hepatectomy, followed by phenobarbital promotion for 7 weeks.
In the promotion test, rats were initiated with DEN after partial hepatectomy, followed by
promotion with tetrachloroethylene for 7 weeks. In a separate initiation study of neonatal female
Wistar rats exposed to 2,000 ppm, 8 hours/day, 5 days/week, for 10 weeks (described in Bolt et
al. (1982). as reported in (NRC. 2010). preneoplastic liver foci were reportedly not observed.
4.8.1.2. Drosophila melanogaster
Limited tetrachloroethylene genotoxicity studies have been performed in Drosophila
melanogaster. One study was negative for both the induction of sex-linked recessive lethal
mutations and chromosomal aberrations following inhalation exposure to tetrachloroethylene in
D. melanogaster (up to 3,400 mg/m3 for 7 hours) (Beliles et al.. 1980). The frequencies of the
sex-linked recessive lethal mutations were 0 and 0.10% for the low- and high-dose exposures,
respectively, which was not significantly different from the negative control (0.11%). This study
also showed no chromosomal aberrations, as there were no significant loses of the long arm of
the Y chromosome for either the low (0.11%) or high (0.02%) doses as compare to the negative
control (0.02%). A second study, also negative for sex-linked recessive lethal mutations,
exposed male Drosophila by feeding tetrachloroethylene (4,000 ppm) or by injection (1,000
ppm) before successive mating with untreated females for 3 days (NTP. 1986b) (also reported in
Valencia et al.. 1985). F1 heterozygous daughters were mated to their siblings. Analysis of the
data after 17 days demonstrated no significant increase in sex-linked recessive lethal mutations
following exposure to tetrachloroethylene.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-367 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
4.8.1.3. Bacterial and Fungal Systems
Cells of Saccharomyces cerevisiae contain cytochrome P450 monooxygenase system and
are capable of metabolizing promutagens to genetically active products. Tetrachloroethylene
alone was positive for mitotic recombination in yeast following 1-hour exposure to 6.6-mM
tetrachloroethylene (Callen etal.. 1980)but negative in yeast exposed in suspension with
metabolic activation or in the intrasinguineous hose-mediated assay (Bronzetti et al.. 1983; Koch
et al.. 1988). Results were negative in the same assay for tetrachloroethylene, but the high level
of cytotoxicity in this assay at the dose used (9.8 mM) limits the interpretation of these results
(Koch et al.. 1988). Bronzetti et al. (1983) also demonstrated negative results both in vitro (0, 5,
10, 20, 60, and 85 mM) with and without S9 activation. There also appeared to be high
cytotoxicity in yeast cells exposed to high dose tetrachloroethylene based on decreasing
percentage survival in this study, which may also limit the interpretation of these data.
A number of in vitro genotoxicity assays have been performed using prokaryotic cells.
Studies of mutagenicity on Escherichia coli have been negative (Greim et al.. 1975)and also
reported in Henscher (1977). Most Ames tests using S. typhimurium have indicated that
tetrachloroethylene in the absence of metabolic activation or in the presence of the standard S9
fraction is not a mutagen (Bartsch et al.. 1979; Connor et al.. 1985; DeMarini et al.. 1994; Greim
et al.. 1975; Hardin et al.. 1981; Haworth et al.. 1983; Kringstad etal.. 1981; Milman et al.. 1988;
NTP. 1986b; Roldan-Ariona et al.. 1991; Shimada et al.. 1985; Warner et al.. 1988; Watanabe et
al.. 1998). However when incubated with rat liver GST, GSH, and a rat kidney fraction,
tetrachloroethylene exhibited a clear dose response (Vamvakas et al.. 1989c)Specificallv. this
study demonstrated the mutagenicity in S. typhimurium (primarily strain TA100) of
tetrachloroethylene that had been preincubated with rat liver GST, GSH, and rat kidney
microsomes, and of TCVG that had been preincubated with rat kidney microsomes.
Additionally, the bacterial mutagenicity of bile from liver perfusate following
tetrachloroethylene exposure in rats was demonstrated (Vamvakas et al.. 1989c). These results
support a role for GSH conjugation in the genotoxicity of tetrachloroethylene.
A more recent study examined genotoxicity of tetrachloroethylene in an S. typhimurium
strain (YG7108pin3ERb5) with enhanced metabolic activity (transformed with CYP2E1,
cytochrome P450 reductase, and cytochrome b5) ,which led to microcolony formation believed
to be from toxicity of tetrachloroethylene metabolites formed at 200- and l,000-[^g doses (but not
at the higher doses of 2,000 or 3,000 |ig) (Emmert et al.. 2006). Tetrachloroethylene was
negative in the parent strain (YG7108) at all doses in the presence of S9. These results support a
role for CYP2E1-derived metabolites in the toxicity of tetrachloroethylene, but not the
mutagenicity of tetrachloroethylene.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-368 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
In summary, gene mutations were not observed following exposure to tetrachloroethylene
in E. coli or S. typhimurium cells in the absence of metabolic activation. Addition of standard S9
fraction also did not lead to mutagenicity, but exposure to bacterial cells with enhanced
metabolic activity (CYP2E1 GSH) led to positive Ames test results. These support a role of
metabolic activation of tetrachloroethylene in its genotoxicity. Results in yeast cells are
conflicting, with one positive study (Callen et al.. 1980) and two negative studies (Bronzetti et
al.. 1983; Koch et al.. 1988) . However, tetrachloroethylene led to cytotoxicity of S. cerevisiae at
the doses tested, making interpretation of these results difficult. These results, although limited,
suggest tetrachloroethylene exposure can lead to genotoxicity in the presence of appropriate
metabolic activation.
4.8.1.4. Summary
The in vitro thymidine kinase gene mutation assay in mammalian cells was negative in
the presence and absence of S9 (F344 rat liver) metabolic activation (NTP. 1986b). Positive
results for frameshift mutation were observed in a host-mediated assay by implanting
S. typhimurium into mice exposed to tetrachloroethylene, but without a clear dose-response
effect (Beliles et al.. 1980). Studies of mutagenicity on E. coli have been negative (Greim et al..
1975) and also reported in Henschler, 1977). A number of mutagenicity studies in S.
typhimurium indicate that, in the absence of metabolic activation or in the presence of the
standard S9 fraction, tetrachloroethylene is not a mutagen (Bartsch et al.. 1979; Connor et al..
1985; DeMarini et al.. 1994; Emmert et al.. 2006; Greim et al.. 1975; Hardin et al.. 1981;
Haworth et al.. 1983; Kringstad et al.. 1981; Milman et al.. 1988; NTP. 1986b; Roldan-Ariona et
al.. 1991; Shimada et al.. 1985; Warner et al.. 1988; Watanabe et al.. 1998). However, when
tetrachloroethylene was activated with rat liver GST, GSH, and a rat kidney fraction,
tetrachloroethylene exhibited a clear dose-response (Vamvakas et al.. 1989c). These findings
support a role of metabolic activation of tetrachloroethylene in its in vitro genotoxicity. Results
in yeast cells are conflicting, with one positive study (Callen et al.. 1980) and two negative
studies (Bronzetti et al.. 1983; Koch et al.. 1988). However, tetrachloroethylene led to
cytotoxicity of S. cerevisiae at the doses tested, making interpretation of these results difficult.
These results, although limited, suggest tetrachloroethylene exposure can lead to genotoxicity in
the presence of appropriate metabolic activation.
DNA binding was not observed in one assay in mice exposed to tetrachloroethylene by
inhalation and oral routes, while protein and RNA binding was observed (Schumann et al..
1980). With a more sensitive assay, low levels of DNA binding were observed in mouse liver,
and even lower levels in mouse kidney and rat and mouse stomach after i.p. injection exposure
(Mazzullo et al.. 1987). In vitro binding to calf thymus DNA occurred in the presence of various
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-369 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
microsomal fractions, as well as in the presence of cytosolic fractions from mice and rats. These
results suggest a role for metabolic activation of the parent compound in DNA binding.
The majority of studies of chromosomal aberrations, micronuclei induction, and sister
chromatid exchange following exposure to tetrachloroethylene are negative. Positive
micronuclei induction was observed following partial hepatectomy at high doses
(2,000 mg/kg-day) in mice (Murakami and Horikawa. 1995). Increased micronuclei induction
was observed in CHO cells in vitro when exposed to tetrachloroethylene in a closed system
(Wang et al.. 2001) but not in CHL cells when exposed in an open system (Matsushima et al..
1999). Dose-dependent increases were observed in human lymphoblastoid cell lines that were
enhanced by stable expression of CYP450 enzymes (Dohertv et al.. 1996; White et al.. 2001) .
Sister chromatid exchanges were not observed in any in vitro studies (Galloway et al.. 1987;
Hartmann and Speit. 1995; NTP. 1986b) and were observed in only one occupational exposure
study, but exposures were contaminated with trichloroethylene, so the interpretation of these
results is limited (Fender. 1993)
Although some increases were observed in UDS following exposure, these were not
statistically significant (NTP. 1986b). The results of DNA strand break assays following
exposure to tetrachloroethylene are equivocal. Walles (1986)demonstrated DNA single-strand
breaks in the liver and kidney of male mice exposed by i.p. injection, but this was reversible
within 24 hours. A second study examined DNA strand breaks after 1 week oral exposure to
tetrachloroethylene, and demonstrated no DNA damage (Potter et al.. 1996). A study of DNA
strand breaks showed a marginal increase in only one parameter from the alkaline Comet assay
(tail intensity) in the liver but not the kidney following oral exposure to tetrachloroethylene in
mice (Cederberg et al.. 2010). but the statistical and biological significance of this result has been
disputed (Cederberg et al.. 2010; Lillford et al.. 2010; Lovell. 2010).
Studies examining DNA adduct formation related to oxidative stress are inconclusive,
with no results in the urine or leukocytes of occupationally exposed individuals (Toraason et al..
2003) and limited utility of the animal study due to significant toxicity in the exposed animals
(Toraason et al.. 1999). Tumor initiation was not observed in Milman et al. (1988) or Bolt et al.
(1982). but the former study reported significant increases in liver foci in a tumor promotion
study. A study examining inhibition of gap junction intercellular communication was positive
(Benane et al.. 1996). Negative results were found for a limited number of other genotoxicity
endpoints including cell transformations (Tu et al.. 1985) and sex-linked recessive lethal
mutation assay in Drosophila (Beliles et al.. 1980; NTP. 1986b) [also reported in Valencia
(1985)].
Overall, evidence from a number of different analyses with various genetic endpoints
indicates that tetrachloroethylene has the potential to induce damage to the structure of the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-370 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
chromosome in a number of targets but has little-to-no ability to induce mutation in bacterial
systems in the absence of metabolic activation or with the standard S9 fraction. However,
metabolic activation via GSH conjugation or cytochrome P450s yields positive results in
bacterial mutagenicity assays.
4.8.2. Trichloroacetic Acid (TCA)
The tetrachloroethylene metabolite TCA has been studied using a variety of genotoxicity
assays for its genotoxic potential (see International Agency for Research on Cancer (I ARC.
2004) for additional information). Evaluation of in vitro studies of TCA must consider toxicity
and acidification of medium resulting in precipitation of proteins, as TCA is commonly used as a
reagent to precipitate proteins. These studies are summarized in Tables 4-41 and 4-42.
4.8.2.1. Mammalian Systems (Including Human Studies)
4.8.2.1.1. Gene mutations
The mutagenicity of TCA has also been tested in cultured mammalian cells (see
Table 4-41). 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 a culture medium treated with TCA concentrations up to 2,150 |ag/m L in the
presence of S9 metabolic activation and up to 3,400 |ag/mL 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 twofold 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.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-371 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-41. Genotoxicity of trichloroacetic acid (TCA)—mammalian systems
(in vitro and in vivo)a
Test system/endpoint
Doses
(LED or
HID)b
Results0
Reference
With
activation
Without
activation
Gene mutation, mouse lymphoma L5178Y/TK+/-
cells, in vitro
3,000
(+)
?
Harrington-Brock
etal.. (1998s)
DNA strand breaks, B6C3Fi mouse and Fischer 344
rat hepatocytes, in vitro
1,630
NT
-
Chang et al.,
Q 992)
DNA strand breaks, human CCRF-CEM
lymphoblastic cells, in vitro
1,630
NT
-
Chang et al.,
(1992)
DNA damage, Chinese hamster ovary cells, in vitro,
comet assay
3 mM
NT
-
Plewa et al.. (2002)
DNA strand breaks, B6C3F, mouse liver, in vivo
1.0, p.o., xl
NT
+
Nelson and Bull
(1988)
DNA strand breaks, B6C3F, mouse liver, in vivo
500, p.o., xl
NT
+
Nelson et al.
(1989)
DNA strand breaks, B6C3Fi mouse liver, in vivo
500, p.o., 10
repeats
NT
-
Nelson et al.
(1989)
DNA strand breaks, B6C3Fi mouse liver and epithelial
cells from stomach and duodenum, in vivo
1,630, p.o., xl
NT
-
Chang et al.,
(1992)
DNA strand breaks, male B6C3Fi mice, in vivo
500
(neutralized)
NT
-
Styles et al.. (1991)
DNA strand breaks, male B6C3Fi mouse liver, in vivo
300, p.o.
NT
+
Hassoun and Dey,
(2008)
Micronucleus formation, Swiss mice, in vivo
125, i.p., x2
NT
+
Bhunya and
Behera. (1987)
Micronucleus formation, female
C57BL/6JfBL10/Alpk mouse bone-marrow
erythrocytes, in vivo
1,300, i.p., x2
NT
Mackay et al.,
(1995)
Micronucleus formation, male C57BL/6JfBL10/Alpk
mouse bone-marrow erythrocytes, in vivo
1,080, i.p., x2
NT
-
Mackay et al.,
(1995)
Micronucleus formation, Pleurodeles waltl larvae
peripheral erythrocytes, in vivo
80
NT
+
Giller et al, (1997)
Chromosomal aberrations, Swiss mouse bone-marrow
cells in vivo
125, i.p., xl
NT
+
Bhunya and
Behera. (1987)
Chromosomal aberrations, Swiss mouse bone-marrow
cells in vivo
100, i.p., x5
NT
+
Bhunya and
Behera. (1987)
Chromosomal aberrations, Swiss mouse bone-marrow
cells in vivo
500, p.o., X1
NT
+
Bhunya and
Behera. (1987)
Chromosomal aberrations, chicken Gallus domesticus
bone marrow, in vivo
200, i.p., xl
NT
+
Bhunya and Jena,
(1996)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-372 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-41. Genotoxicity of trichloroacetic acid (TCA)—mammalian
systems (in vitro and in vivo)a (continued)
Test system/endpoint
Doses
(LED or
HID)b
Results0
Reference
With
activation
Without
activation
Chromosomal aberrations, human lymphocytes, in
vitro
5,000
(neutralized)
NT
—
Mackay et al.,
(1995s)
Sperm morphology, Swiss mouse, in vivo
125, i.p., x5
NT
+
Bhunya and
Behera. (1987s)
Increased detection of MiG and 8-OHdG adducts,
B6C3Fi neonatal mouse liver DNA, in vivo
2,000 nmol
NT
+
Von Tungeln et al.,
(2002)
aTable adapted from ATSDR (1997a) and IARC monograph (19951 and modified/updated for newer references.
bLED, lowest effective dose; HID, highest ineffective dose; doses are in ng/mL for in vitro tests; mg/kg for in vivo
tests unless specified.
°Results: + = positive; (+) = weakly positive; - = negative; NT = not tested; ? = inconclusive.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-373 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-42. Genotoxicity of trichloroacetic acid (TCA)—bacterial systems"
Test system/endpoint
Doses
(LED or HID)b
Results0
Reference
With
activation
Without
activation
1 Prophage induction, E. coli WP2s
10,000
-
-
DeMarini et al.,
(1994s)
SOS chromotest, E. coli PQ37
10,000
-
-
Giller et al.. (1997)
S. typhimurium TA1535, 1536, 1537, 1538,
reverse mutation
20 ng/plate
NT
-
Shirasu et al.. (1976s)
S. typhimurium TA100, 98, reverse mutation
450 ng/plate
-
-
Waskell. (1978s)
S. typhimurium TA100, 1535, reverse
mutation
4,000 ng/plate
-
-
Nestmann et al.,
(1980s)
S. typhimurium TA1537, 1538, 98, reverse
mutation
2,000 ng/plate
-
-
Nestmann et al.,
(1980s)
S. typhimurium TA100, reverse mutation
520 ng/plate
NT
-
Rapsonetal.. (1980s)
S. typhimurium TA100, 98, reverse mutation
5,000 ng/plate
-
-
Morivaetal.. (1983s)
S. typhimurium TA100, reverse mutation
600 ppm
-
-
DeMarini et al.,
(1994s)
S. typhimurium TA100, reverse mutation,
liquid medium
1,750
+
+
Giller etal., (1997)
S. typhimurium TA104, reverse mutation,
microsuspension
250 ng/plate
-
-
Nelson et al. (2001s)
S. typhimurium TA100, RSJ100, reverse
mutation
16,300
-
-
Kargalioglu et al.,
(2002s)
S. typhimurium TA98, reverse mutation
13,100
-
-
Kargalioglu et al.,
(2002s)
S. typhimurium TA1535, SOS DNA repair
NA
+
-
Ono et al.. (1991)
aTable adapted from IARC monograph (20041 and modified/updated for newer references.
bLED, lowest effective dose; HID, highest ineffective dose; doses are in |ig/mL for in vitro tests, unless otherwise
specified.
°Results: + = positive; - = negative; NT = not tested.
1
4.8.2.1.2. Chromosomal aberrations
2 Mackay et al. (1995) investigated the ability of TCA to induce chromosomal damage in
3 an in vitro chromosomal aberration assay using cultured human cells. The authors treated the
4 cells with TCA as free acid, both in the presence and absence of metabolic activation. TCA
5 induced chromosomal damage in cultured human peripheral lymphocytes at concentrations
6 (2,000 and 3,500 |ig/mL) that significantly reduced the pH of the medium. However, exposure
7 of cells to neutralized TCA did not have any effect, even at a cytotoxic concentration of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-374 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
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 the
isolates 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 unneutralized TCA. In another in vitro study, Plewa et al. (2002) evaluated
the induction of DNA strand breaks by TCA (1-25 mM) in CHO cells and did not observe any
genotoxicity.
4.8.2.1.2.1. Micronucleus induction
Genotoxicity of TCA was tested in a mouse in vivo system using three different
cytogenetic assays (bone marrow chromosomal aberrations, micronucleus and sperm-head
abnormalities) (Bhunya and Behera. 1987) and for chromosomal aberrations in chicken (Bhunya
and Jena. 1996). TCA induced a variety of anomalies including micronucleus in the bone
marrow of mice and chicken. A small increase in the frequency of micronucleated erythrocytes
at 80 |ig/mL in a newt (Pleurodeles waltl larvae) micronucleus test was observed in 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 i.p. 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 2 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.
4.8.2.1.2.2. DNA damage studies
DNA unwinding assays have been used as indicators of single-strand breaks. Studies
were conducted on the ability of TCA to induce DNA single-strand breaks (Chang et al.. 1992;
Table 4 12; Nelson and Bull. 1988; Nelson et al.. 1989; Styles et al.. 1991). Nelson and Bull
(1988) evaluated the ability of TCA and other compounds to induce DNA single-strand 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 DNA single-strand breaks by the
alkaline unwinding assay. Dose-dependent increases in DNA single-strand breaks were induced
in both rats and mice, with mice being more susceptible than rats. The lowest dose of TCA that
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-375 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
produced significant SSBs was 0.6 mmol/kg (98 mg/kg) in rats but 0.006 mmol/kg (0.98 mg/kg)
in mice.
However, in a follow-up study (Nelson et al.. 1989). no significant differences from
controls in DNA single-strand breaks in whole liver homogenates were seen in male B6C3Fi
mice exposed to 500-mg/kg TCA. Moreover, DCA increased single-strand breaks but with no
dose response between 10 and 500 mg/kg, raising concerns about the reliability of the DNA
unwinding assay used in these studies. In an additional follow-up experiment with a similar
experimental paradigm, 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 1 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).
In summary, Nelson and Bull (1988) reported that DCA and TCA enhance DNA
unwinding in mice, with DCA having the highest activity and TCA the lowest. However, Nelson
et al. (1989) reported no effect for TCA and a lack of dose response for the effect of DCA (with
10- and 500-mg/kg DCA inducing the same magnitude of effect). Moreover, Styles et al. (1991)
did not report a positive result for TCA using the same paradigm as Nelson and Bull (1988) and
Nelson et al. (1989). Furthermore, Chang et al. (1992) also did not find increased DNA single-
strand breaks for TCA exposure in rats.
4.8.2.2. Bacterial Systems
4.8.2.2.1. Gene mutations
TCA has been evaluated in a number of in vitro test systems including the bacterial
assays (Ames) using different S. typhimurium strains such as TA98, TA100, TA104, TA1535,
and RSJ100 (see Table 4-42). The majority of these studies did not report positive findings for
genotoxicity (DeMarini et al.. 1994; Kargalioglu et al.. 2002; Moriya et al.. 1983; Nelson et al..
2001; Nestmann et al.. 1980; Rapson et al.. 1980; Shirasu et al.. 1976; Waskell. 1978). Waskell
(1978) studied the effect of TCA (0.45 mg/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 nontoxic
dose tested. Following exposure to TCA, Rapson et al. (1980) reported no change in mutagenic
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-376 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
activity in strain TA100 in the absence of 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 S. typhimurium TA100 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 S. typhimurium
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 S. typhimurium 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 E. coli PQ37, with and without S9
(Giller et al.. 1997). evaluated the genotoxic activity of TCA ranging from 10 to 10,000 |ig/mL,
and no response was reported. 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.
4.8.2.3. Summary
TCA, an oxidative metabolite of tetrachloroethylene, exhibits little, if any genotoxic
activity in vitro. TCA did not induce mutations in S. typhimurium strains in the absence of
metabolic activation or in an alternative protocol using a closed system (DeMarini et al.. 1994:
Giller et al.. 1997: Kargalioglu et al.. 2002: Nelson et al.. 2001: Rapson et al.. 1980: Waskell.
1978). but a mutagenic response was induced in TA100 in the Ames fluctuation test (Giller et al..
1997). 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 (Mackav et al.. 1995).
Measures of DNA-repair responses in bacterial systems have shown induction of DNA repair
reported in S. typhimurium but not in E. coli. Mutagenicity in mouse lymphoma cells was only
induced at cytotoxic concentrations (Harrington-Brock et al.. 1998). TCA was positive in some
genotoxicity studies in vivo mouse, newt, and chick test systems (Bhunya and Behera. 1987:
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-377 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Bhunya and Jena. 1996; Birner et al.. 1994; Giller et al.. 1997). DNA unwinding assays have
either shown TCA to be much less potent than DCA (Nelson and Bull 1988) or negative (Nelson
et al.. 1989; Styles et al.. 1991). Due to limitations in the genotoxicity database, the possible
contribution of TCA to tetrachloroethylene genotoxicity is unclear.
4.8.3. Dichloroacetic Acid (DCA)
DCA is another metabolite of tetrachloroethylene that has been studied using a variety of
genotoxicity assays for its genotoxic potential (see Tables 4-43 and 4-44; see (IARC. 2004) for
additional information).
4.8.3.1. Mammalian Systems
4.8.3.1.1. Gene mutations
The mutagenicity of DCA has been tested in mammalian systems, particularly, mouse
lymphoma cell lines in vitro (Fox et al.. 1996; Harrington-Brock et al.. 1998) and lacl transgenic
mice in vivo (Leavitt et al.. 1997). Harrington-Brock et al. (1998) evaluated DCA for 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). Different pH levels were tested in induction of
mutant frequencies, and it was determined that the mutagenic effect observed was due to the
chemical and not pH effects.
Mutation frequencies were studied in male transgenic B6C3Fi mice harboring the
bacterial lacl gene administered DCA at either 1.0 or 3.5 g/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 1.0-g/L DCA showed a slight increase (1.3-fold) in the mutant frequency over the control,
but mice treated with 3.5-g/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
T:A sites in Codon 61 of the H-ras gene was increased in DCA-induced liver tumors in B6C3Fi
mice (Leavitt et al.. 1997).
Table 4-43. Genotoxicity of dichloroacetic acid (DCA)—mammalian systems
(in vitro and in vivo)a
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-378 DRAFT-DO NOT CITE OR QUOTE
-------
Test system/endpoint
Doses
(LED or
HID)b
Results0
Reference
With
activation
Without
activation
Gene mutation, mouse lymphoma cell line
L5178Y/TK+/- in vitro
5,000
-
-
Fox et al. (1996)
Gene mutation, mouse lymphoma cell line
L5178Y/TK+/- -3.7.2C in vitro
400
NT
+
Harrington-Brock
et al. ri998,l
DNA strand breaks and alkali-labile damage, Chinese
hamster ovary cells in vitro (single-cell gel
electrophoresis assay)
3,225 ng/mL
NT
Plewa et al. (2002)
DNA strand breaks, B6C3Fi mouse hepatocytes in
vitro
2,580
NT
-
Chans et al. (1992)
DNA strand breaks, Fischer 344 rat hepatocytes in
vitro
1,290
NT
-
Chans et al. (1992)
Micronucleus formation, mouse lymphoma
L5178Y/TK+/- -3.7.2C cell line in vitro
800
NT
-
Harrington-Brock
et al. ri998,l
Micronucleus induction, peripheral blood
erythrocytes, Tg.AC hemizygous mouse, dermal
application in vivo
500 mg/kg
NT
NTP (2007)
Micronucleus induction, peripheral blood
erythrocytes, Tg.AC hemizygous mouse, drinking
water, in vivo
2,000 mg/L
NT
NTP (2007)
Micronucleus induction, peripheral blood
erythrocytes, p53 haploinsufficient mouse, drinking
water, in vivo
2,000 mg/L
NT
NTP (2007)
Micronucleus induction, peripheral blood
erythrocytes, B6C3F, mouse, drinking water, in vivo
67 mg/L
NT
- (male)
equivocal
(female)
NTP (2007)
Chromosomal aberrations, Chinese hamster ovary in
vitro
5,000
-
-
Fox et al. (1996)
Chromosomal aberrations, mouse lymphoma
L5178Y/Tk+/- -3.7.2C cell line in vitro
600
NT
+
Harrington-Brock
et al. ri998^
Aneuploidy, mouse lymphoma L5178Y/Tk+/- -3.7.2C
cell line in vitro
800
NT
-
Harrington-Brock
et al. Q998N)
DNA strand breaks, human CCRF-CEM
lymphoblastoid cells in vitro
1,290
NT
-
Chans et al. (1992)
DNA strand breaks, male B6C3Fi mouse liver in vivo
13, p.o., xl
NT
+
Nelson and Bull
(1988)
DNA strand breaks, male B6C3Fi mouse liver in vivo
©
O
X
NT
+
Nelson et al.
(1989)
DNA strand breaks, male B6C3Fi mouse liver in vivo
1,290, p.o., xl
NT
-
Chans et al. (1992)
Table 4-43. Genotoxicity of dichloroacetic acid (DCA)—mammalian systems
(in vitro and in vivo)a (continued)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-379 DRAFT-DO NOT CITE OR QUOTE
-------
Test system/endpoint
Doses
(LED or
HID)b
Results0
Reference
With
activation
Without
activation
DNA strand breaks, male B6C3Fi mouse splenocytes
in vivo
1,290, p.o., xl
NT
—
Chans et al. (1992)
DNA strand breaks, male B6C3Fi mouse epithelial
cells from stomach and duodenum in vivo
1,290, p.o., xl
NT
-
Chans et al. (1992)
DNA strand breaks, male B6C3Fi mouse liver in vivo
5,000, dw,
x7-14d
NT
-
Chans et al. (1992)
DNA strand breaks, male B6C3Fi mouse liver, in
vivo
300, p.o.
NT
+
Hassoun and Dey
(2008)
DNA strand breaks, alkali-labile sites, cross linking,
male B6C3Fi mouse blood leukocytes in vivo (single-
cell gel electrophoresis assay)
3,500, dw,
x28d
NT
+
Fuscoe et al.
(1996)
DNA strand breaks, male Sprague-Dawley rat liver in
vivo
30, p.o., X1
NT
+
Nelson and Bull
(1988)
DNA strand breaks, male Fischer 344 rat liver in vivo
645, p.o., X1
NT
-
Chans et al. (1992)
DNA strand breaks, male Fischer 344 rat liver in vivo
2,000, dw,
x30 wk
NT
-
Chans et al. (1992)
Gene mutation, lacl transgenic male B6C3Fi mouse
liver assay in vivo
1,000, dw,
x60 wk
NT
+
Leavitt et al.
(1997)
Altered gene expression, male B6C3Fi mouse liver
assay in vivo
2,000, dw,
x4 wk
NT
+
Thai et al. (2003)
Micronucleus formation, male B6C3Fi mouse
peripheral erythrocytes in vivo
3,500, dw, x9 d
NT
+
Fuscoe et al.
(1996)
Micronucleus formation, male B6C3Fi mouse
peripheral erythrocytes in vivo
3,500, dw, x28
d
NT
-
Fuscoe et al.
(1996)
Micronucleus formation, male B6C3Fi mouse
peripheral erythrocytes in vivo
3,500, dw, xlO
wk
NT
+
Fuscoe et al.
(1996)
Micronucleus formation, male and female Crl:CD
(S-D) BR rat bone-marrow erythrocytes in vivo
1,100, i.v., x3
NT
-
Fox et al. (1996)
Micronucleus formation, Pleurodeles waltl larvae
peripheral erythrocytes in vivo
80 d
NT
—
Giller et al. (1997)
aTable adapted from IARC monograph (20041 and modified/updated for newer references.
bLED, lowest effective dose; HID, highest ineffective dose; doses are in ng/mL for in vitro tests; mg/kg for in vivo
tests unless specified; dw = drinking-water (in mg/L); i.v. = intravenous.
°Results: + = positive; - = negative; NT = not tested.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-380 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-44. Genotoxicity of dichloroacetic acid (DCA)—bacterial systems"
Test system/endpoint
Doses
(LED or HID)b
Results0
Reference
With
activation
Without
activation
1 Prophage induction, E. coli WP2s
2,500
+
-
DeMarini et al.,
(1994)
SOS chromotest, E. coli PQ37
500
-
(+)
Giller et al. (19971
S. typhimurium, DNA repair-deficient strains
TS24, TA2322, TA1950
31,000
-
-
Waskell (1978)
S. typhimurium TA100, TA1535, TA1537,
TA1538, reverse mutation
NA
-
-
Herbert et al.. (19801
S. typhimurium TA100, reverse mutation
50
+
+
DeMarini et al.,
(19941
S. typhimurium TA100, TA1535, TA1537,
TA98, reverse mutation
5,000
-
-
Fox et al. (19961
S. typhimurium TA100, reverse mutation, liquid
medium
100
+
+
Giller etal. (1997)
S. typhimurium RSJ100, reverse mutation
1,935
-
+
Kargalioglu et al.
(2002)
S. typhimurium TA104, reverse mutation,
microsuspension
150 ng/plate
-
-
Nelson etal. (2001)
S. typhimurium TA98, reverse mutation
10 ng/plate
(+)
-
Herbert et al.. (1980)
S. typhimurium TA98, reverse mutation
5,160
-
+
Kargalioglu et al.
(2002)
S. typhimurium TA100, reverse mutation
1,935
+
+
Kargalioglu et al.
(2002)
S. typhimurium TA98, gene mutation
3 ng/plate
-
-
NTP (2007)
S. typhimurium TA100, gene mutation
333 ng/plate
-
+
NTP (2007)
S. typhimurium TA1535, gene mutation
333 ng/plate
-
+
NTP (2007)
E. coli WP2uvrA, reverse mutation
5,000
-
-
Fox et al. (1996)
"Table adapted from IARC monograph (20041 and modified/updated for newer references.
bLED, lowest effective dose; HID, highest ineffective dose; doses are in ng/mL for in vitro tests, unless otherwise
specified; NA = not available.
°Results: + = positive; (+) = weakly positive; - = negative.
1
4.8.3.1.2. Chromosomal aberrations and micronucleus induction
2 Harrington-Brock et al. (1998) evaluated DCA for its potential to induce chromosomal
3 aberrations in DCA-treated (0, 600, and 800 |ig/mL) mouse lymphoma cells. A clearly positive
4 induction of aberrations was observed at both concentrations tested. No significant increase in
5 micronucleus was observed in DCA-treated (0, 600, and 800 |ig/mL) mouse lymphoma cells
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-381 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
(Harrington-Brock et al.. 1998). However, no chromosomal aberrations were found in Chinese
hamster ovary cells exposed to DCA (Fox et al.. 1996).
Fuscoe et al. (1996) investigated in vivo genotoxic potential of DCA in bone marrow and
blood leukocytes using the peripheral-blood-erythrocyte micronucleus assay (to detect
chromosome breakage and/or malsegregation) and the alkaline single cell gel electrophoresis
(comet) assay, respectively. Mice were exposed to DCA in drinking water, available ad libitum,
for up to 31 weeks. A statistically significant dose-related increase in the frequency of
micronucleated PCEs was observed following subchronic exposure to DCA for 9 days.
Similarly, a significant increase was also observed when mice were exposed for >10 weeks,
particularly at the highest dose of DCA tested (3.5 g/L). DNA cross-linking was observed 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 are exposed to
concentrations similar to those used in the rodent bioassay.
4.8.3.1.3. DNA damage studies
Nelson and Bull (1988) and Nelson et al. (1989) have been described above in
Sections 4.2.2.4 and 4.2.3. Nelson and Bull (1988) reported positive results for DNA unwinding
for DCA, although Nelson et al. (1989) reported the same response at 10 and 500 mg/kg in mice,
raising concerns about the reliability of the assay in these studies. Chang et al. (1992) conducted
both in vitro and in vivo studies to determine the ability of DCA to cause DNA damage. Primary
rat (Fischer 344) hepatocytes and primary mouse hepatocytes treated with DCA for 4 hours did
not induce DNA single-strand breaks as detected by the alkaline DNA unwinding assay. No
DNA single-strand breaks were observed in human CCRF-CEM lymphoblastoid cells in vitro
exposed to DCA. Similarly, analysis of the DNA single-strand breaks in mice killed 1 hour after
a single dose of 1, 5, or 10-mM/kg DCA did not cause DNA damage. None of the Fischer 344
rats killed 4 hours after a single gavage treatment (1-10 mM/kg) produced any detectable DNA
damage.
4.8.3.2. Bacterial Systems
4.8.3.2.1. Gene mutations
Studies were conducted to evaluate mutagenicity of DCA in different S. typhimurium and
E. coli strains (DeMarini etal.. 1994; Fox et al.. 1996; Gilleretal.. 1997; Herbert et al.. 1980;
Kargalioglu et al.. 2002; Nelson et al.. 2001; Waskell. 1978) summarized in Table 4-44). 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 (TA104, TA1535, TA1537, and TA1538) or in E.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-3 82 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
coli strain WP2uvrA. In one study, DCA caused a weak induction of SOS repair in E. coli strain
PQ37 (Giller et al.. 1997V
DeMarini et al. (1994). in the same study as described in the TCA section (see Section
4.8.2), 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 threefold 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 S. typhimurium TA100 strain, DCA was mutagenic both in the
presence and absence of S9, producing three- to fivefold increases in the revertants/plate
compared to the background. The lowest effective concentration for DCA without S9 was
100 ppm and 50 ppm 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 S. typhimurium
TA100. DCA-induced revertants were chosen for further molecular analysis at concentrations
that produced mutant yields that were two- to fivefold greater than the background. The
mutation spectra of DCA were significantly different from the background mutation spectrum.
Thus, despite the modest increase in the mutant yields (3-5 times) produced by DCA, 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 including DCA in S. typhimurium strains TA98, TA100, and
RSJ100 +/- S9. DCA was mutagenic in this test, although the response was low when compared
to other disinfection by-products 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 S. typhimurium 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 the data are limited, it appears that DCA has mutagenic
activity in the S. typhimurium strains, particularly TA100.
4.8.3.3. Summary
DCA, a chloroacid metabolite of tetrachloroethylene, has also been studied using
different types of genotoxicity assays. Although studies are limited for different genetic
endpoints, DCA has been demonstrated to be mutagenic in some strains in S. typhimurium assays
(DeMarini et al.. 1994; Kargalioglu et al.. 2002; Plewa et al.. 2002). a mouse lymphoma assay
(Harrington-Brock et al.. 1998). in vivo cytogenetic tests (Fuscoe et al.. 1996; Leavitt et al..
1997). the micronucleus induction test, the Big Blue mouse system, and other tests (Chang et al..
1989; DeMarini et al.. 1994; Fuscoe et al.. 1996; Harrington-Brock et al.. 1998; Leavitt et al..
1997; Nelson and Bull. 1988; Nelson et al.. 1989). DCA can cause DNA strand breaks in mouse
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-383 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
and rat liver cells following in vivo exposures (Fuscoe et al.. 1996). Because of uncertainties as
to the extent of DC A formed from tetrachloroethylene exposure, inferences as to the possible
contribution from DCA genotoxicity to tetrachloroethylene toxicity are difficult to make.
4.8.4. Chloral Hydrate
Although chloral hydrate is postulated as a metabolite of tetrachloroethylene, this is not
widely accepted. However, to be inclusive of all known genotoxicity information, chloral
hydrate genotoxicity studies have been reviewed in the following section. Chloral hydrate has
been evaluated for its genotoxic potential using a variety of genotoxicity assays (see Tables 4-45,
4-46, and 4-47).
4.8.4.1. Mammalian Systems (Including Human Studies)
4.8.4.1.1. Gene mutations
Harrington-Brock (1998) noted that chloral hydrate-induced concentration related
cytotoxicity in TK+/- mouse lymphoma cell lines without S9 activation. A nonstatistical
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 |ig/mL), cytotoxicity was observed. Percentage cell
survival ranged from 96 to 4%.
4.8.4.1.2. DNA binding studies
Limited analysis has been performed examining the DNA binding potential of chloral
hydrate (Keller and Heck. 1988; Ni et al.. 1995; Von Tungeln et al.. 2002). Keller and Heck
(1988) conducted both in vitro and in vivo experiments using the B6C3Fi mouse strain. The
mice were pretreated with 1,500-mg/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. Another
study with in vivo exposures to nonradioactive chloral hydrate at a concentration of 1,000 and
2,000 nmol in B6C3Fi mice demonstrated an increase in malondialdehyde-derived and
Table 4-45. Genotoxicity of chloral hydrate—mammalian systems (in vitro)3
Test system/endpoint
Doses
(LED or
HID)b
Results0
Reference
With
activation
Without
activation
DNA-protein cross-links, rat nuclei in vitro
41,250
NT
-
Keller and Heck (1988)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-3 84 DRAFT-DO NOT CITE OR QUOTE
-------
DNA single-strand breaks, rat primary hepatocytes in
vitro
1,650
NT
-
Chans et al. (1992)
Gene mutation, mouse lymphoma L5178Y/TK+/-, in
vitro
1,000
NT
(+)
Harrington-Brock et al.
(1998)
Sister chromatid exchange, CHO cells, in vitro
100
+
+
Beland. (1999)
Micronucleus formation (kinetochore-positive),
Chinese hamster CI cells, in vitro
165
NT
+
Degrassi and Tanzarella
(1988)
Micronucleus formation (kinetochore-negative),
Chinese hamster CI cells, in vitro
250
NT
-
Degrassi and Tanzarella
(1988)
Micronucleus formation (kinetochore-positive),
Chinese hamster LUC2 cells, in vitro
400
NT
+
Parrv etal. (1990)
Micronucleus formation (kinetochore-positive),
Chinese hamster LUC2 cells, in vitro
400
NT
+
Lvnch and Parrv. (1993)
Micronucleus formation, Chinese hamster V79 cells,
in vitro
316
NT
+
Seelbach et al. (1993)
Micronucleus formation, mouse lymphoma
L5178Y/TK+/-, invito
1,300
NT
-
Harrington-Brock et al.
(1998)
Micronucleus formation, mouse lymphoma
L5178Y/TK+/-, invito
500
NT
+
Nesslany and Marzin
(1999)
Chromosomal aberrations, Chinese Hamster CHED
cells, in vitro
20
NT
+
Furnus et al. (1990)
Chromosomal aberrations, Chinese Hamster ovary
cells, in vitro
1,000
+
+
Beland, (1999)
Chromosomal aberrations, mouse lymphoma
L5178Y/TK +/- cells line, invito
1,250
NT
(+)
Harrington-Brock et al.
(1998)
Aneuploidy, Chinese hamster CHED cells, in vitro
10
NT
+
Furnus et al. (1990)
Aneuploidy, primary Chinese hamster embryonic
cells, in vitro
250
NT
+
Nataraian et al. (1993)
Aneuploidy, Chinese hamsterLUC2p4 cells, invito
250
NT
+
Warretal. M993^
Aneuploidy, mouse lymphoma L5178Y/TK+/-, in
vitro
1,300
NT
-
Harrington-Brock et al.
(1998)
Tetraploidy and endoredupliation, Chinese hamster
LUC2p4cells, invito
500
NT
+
Warretal. (1993)
Cell transformation, Syrian hamster embryo cells
(24-h treatment)
350
NT
+
Gibson etal. (1995)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-385 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-45. Genotoxicity of chloral hydrate—mammalian systems (in vitro)3
(continued)
Test system/endpoint
Doses
(LED or
HID)b
Results0
Reference
With
activation
Without
activation
Cell transformation, Syrian hamster dermal cell line
(24-h treatment)
50
NT
+
Parrv et al.. (1996)
DNA single-strand breaks, human lymphoblastoid
cells, in vitro
1,650
NT
-
Chans et al. (1992)
Gene mutation, tk and hprt locus, human
lymphoblastoid
1,000
NT
+
Beland, (1999)
Sister chromatid exchanges, human lymphocytes, in
vitro
54
NT
(+)
Gu et al. (1981)
Micronucleus formation, human lymphocytes, in vitro
100
+
Van Hummelen and
Kirsch-Volders,
(1992)
Micronucleus formation, human lymphoblastoid
AHH-1 cell line, in vitro
100
NT
+
Parrv et al.. (1996)
Micronucleus formation, human lymphoblastoid
MCL-5 cell line, in vitro
500
NT
-
Parrv et al.. (1996)
Micronucleus formation (kinetochore-positive),
human diploid LEO fibroblasts, in vitro
120
NT
+
Bonatti et al. (1992)
Aneuploidy (double Y induction), human
lymphocytes, in vitro
250
NT
+
Vagnarelli et al.,
(1990)
Aneuploidy (hyperdiploidy and hypodiploidy), human
lymphocytes in vitro
50
NT
+
Sbrana et al.. (1993)
Polyploidy, human lymphocytes, in vitro
137
NT
+
Sbrana et al.. (1993)
C-Mitosis, human lymphocytes, in vitro
75
NT
+
Sbrana et al.. (1993)
aTable adapted from IARC monograph (20041 and modified/updated for newer references.
bLED, lowest effective dose; HID, highest ineffective dose; doses are in |ig/mL for in vitro tests unless otherwise
specified.
°Results: + = positive; (+) = weakly positive; - = negative; NT = not tested.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-386 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-46. Genotoxicity of chloral hydrate—mammalian systems (in vivo)"
Test system/endpoint
Doses
(LED or HID)b
Results0
Reference
DNA single-strand breaks, male Sprague-Dawley rat liver
300, p.o.
+
Nelson and Bull ^1988")
DNA single-strand breaks, male Fischer 344 rat liver
1,650, p.o.
-
Chang et al. (1992)
DNA single-strand breaks, male B6C3Fi mouse liver
o
©
O
+
Nelson and Bull (1988s)
DNA single-strand breaks, male B6C3Fi mouse liver
825, p.o.
-
Chang et al. (1992)
Increased detection of MiG and 8-OHdG adducts, B6C3Fi
neonatal mouse liver DNA, in vivo, i.p. injection
2,000 nmol
+
Von Tunseln et al. (2002)
Micronucleus formation, male and female NMRI mice, bone-
marrow erythrocytes
500, i.p.
-
Leuschner and Leuschner,
(1991)
Micronucleus formation, BALB/c mouse spermatids
83, i.p.
-
Russo and Levis. (1992a)
Micronucleus formation, male BALB/c mouse bone-marrow
erythrocytes and early spermatids
83, i.p.
+
Russo and Levis. (1992a)
Micronucleus formation, male BALB/c mouse bone-marrow
erythrocytes
200, i.p.
+
Russo et al.. (1992a)
Micronucleus formation, male F1 mouse bone-marrow
erythrocytes
400, i.p.
-
Leooardi et al.. (1993)
Micronucleus formation, C57B1 mouse spermatids
41, i.p.
+
Allen etal.. (1994)
Micronucleus formation, male Swiss CD-I mouse bone-
marrow erythrocytes
200, i.p.
+
Marrazzini et al. (1994)
Micronucleus formation, B6C3Fi mouse spermatids after
spermatogonia! stem-cell treatment
165, i.p.
+
Nutlev et al.. (1996)
Micronucleus formation, B6C3Fi mouse spermatids after
meiotic cell treatment
413, i.p.
-
Nutlev et al.. (1996)
Micronucleus formation, male Fl, BALB/c mouse peripheral-
blood erythrocytes
200, i.p.
-
Grawe et al. (1997)
Micronucleus formation, male B6C3Fi mouse bone-marrow
erythrocytes
500, i.p., x3
+
Beland, (1999)
Micronucleus formation, infants, peripheral lymphocytes
50, p.o.
+
Ikbal et al.. (2004)
Chromosomal aberrations, male and female Fl mouse bone
marrow cells
600, i.p.
-
Xu and Alder, (1990)
Chromosomal aberrations, male and female Sprague-Dawley
rat bone-marrow cells
1,000, p.o.
-
Leuschner and Leuschner,
(1991)
Chromosomal aberrations, BALB/c mouse spermatogonia
treated
83, i.p.
-
Russo and Levis. (1992a)
Chromosomal aberrations, Fl mouse secondary
spermatocytes
82.7, i.p.
+
Russo et al. (1984)
Chromosomal aberrations, male Swiss CD-I mouse bone-
marrow erythrocytes
400, i.p.
-
Marrazzini et al. (1994)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-387 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-46. Genotoxicity of chloral hydrate—mammalian systems (in vivo)a
(continued)
Test system/endpoint
Doses
(LED or HID)b
Results0
Reference
Chromosomal aberrations, ICR mouse oocytes
600, i.p.
Mailhes et al.. (1993)
Micronucleus formation, infants, peripheral lymphocytes
50, p.o.
+
Ikbal et al.. (2004)
Polyploidy, male and female Fl, mouse bone-marrow cells
600, i.p.
Xu and Adler. (1990)
Aneuploidy Fl mouse secondary spermatocytes
200, i.p.
+
Miller and Adler. (1992)
Aneuploidy, male Fl mouse secondary spermatocytes
400, i.p.
Leopardi et al.. (1993)
Hyperploidy, male Swiss CD-I mouse bone-marrow
erythrocytes
200, i.p.
+
Marrazzini et al. (1994)
"Table adapted from IARC monograph (20041 and modified/updated for newer references.
bLED, lowest effective dose; HID, highest ineffective dose; doses are in mg/kg bw for in vivo tests unless otherwise
specified; i.p. = intraperitoneal^, p.o. = orally.
°Results: + = positive; - = negative
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-388 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-47. Genotoxicity of chloral hydrate—bacterial, yeast, and fungal
systems"
Test system/endpoint
Doses
(LED or HID)b
Results0
Reference
With
activation
Without
activatio
n
SOS chromotest, E. coli PQ37
10,000
-
-
Gilleretal.. M995^
S. typhimurium TA100, TA1535, TA98, reverse
mutation
10,000
-
-
Waskell. (1978)
S. typhimurium TA100, TA1537, TA1538, TA98,
reverse mutation
1,000
+
+
Haworth et al. (1983)
S. typhimurium TA100, reverse mutation
5,000 ng/plate
-
-
Leuschner and Leuschner,
(1991)
S. typhimurium TA100, reverse mutation
2,000 ng/plate
+
+
Ni et al.. (1994)
S. typhimurium TA100, reverse mutation, liquid
medium
300
+
-
Gilleretal., (1995)
S. typhimurium TA100, TA104, reverse mutation
1,000 ng/plate
+
+
Beland. (1999)
S. typhimurium TA104, reverse mutation
1,000 ng/plate
+
+
Ni et al.. CI994)
S. typhimurium TA1535, reverse mutation
1,850
-
-
Leuschner and Leuschner,
(1991)
S. typhimurium TA1535, TA1537 reverse
mutation
6,667
-
-
Haworth et al. (1983)
S. typhimurium TA1535, reverse mutation
10,000
-
-
Beland. (1999)
S. typhimurium TA98, reverse mutation
7,500
-
-
Haworth etal. (1983)
S. typhimurium TA98, reverse mutation
10,000 ng/plate
-
+
Beland. (1999)
A. nidulans, diploid strain 35X17, mitotic
crossovers
1,650
NT
-
Crebelli et al. (1985)
A. nidulans, diploid strain 30, mitotic crossovers
6,600
NT
-
Kafer (1986)
A. nidulans, diploid strain NH, mitotic crossovers
1,000
NT
-
Kappas. (1989)
A. nidulans, diploid strain PI, mitotic crossovers
990
NT
-
Crebelli et al.. (1991)
A. nidulans, diploid strain 35X17,
nondisjunctions
825
NT
+
Crebelli et al. (1985)
A. nidulans, diploid strain 30, aneuploidy
825
NT
+
Kafer (1986)
A. nidulans, haploid conidia, aneuploidy,
polyploidy
1,650
NT
+
Kafer (1986)
A. nidulans, diploid strain NH, nondisjunctions
450
NT
+
Kappas. (1989)
A. nidulans, diploid strain PI, nondisjunctions
660
NT
+
Crebelli et al.. (1991)
A. nidulans, haploid strain 35, hyperploidy
2,640
NT
+
Crebelli et al.. (1991)
S. cerevisiae, meiotic recombination
3,300
NT
?
Sora and Agostini Carbone
(1987)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-389 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Table 4-47. Genotoxicity of chloral hydrate—bacterial, yeast, and fungal
systems" (continued)
Test system/endpoint
Doses
(LED or HID)b
Results0
Reference
With
activation
Without
activation
S. cerevisiae, disomy in meiosis
2,500
NT
+
Sora and Agostini
Carbone (1987)
S. cerevisiae, disomy in meiosis
3,300
NT
+
Sora and Agostini
Carbone (1987)
S. cerevisiae, D61.M, mitotic chr. malsegregation
1,000
NT
+
Albertini. (1990)
D. melanogaster, somatic mutation wing spot test
825
NT
+
Zordanetal.. (1994)
D. melanogaster, induction of sex-linked lethal
mutation
37.2 feed
NT
?
Beland, (1999)
D. melanogaster, induction of sex-linked lethal
mutation
67.5 injection
NT
—
Beland, (1999)
"Table adapted from IARC monograph (20041 and modified/updated for newer references.
bLED, lowest effective dose; HID, highest ineffective dose; doses are in ng/mL for in vitro tests; inj =
injection.
°Results: + = positive; - = negative; NT = not tested; ? = inconclusive.
8-oxo-2'-deoxyguanosine adducts in liver DNA (Von Tungeln et al.. 2002). Ni et al. (1995)
observed malondialdehyde adducts in calf thymus DNA when exposed to chloral hydrate and
microsomes from male B6C3Fi mouse liver.
Keller and Heck (1988) investigated the potential of chloral to form DNA-protein cross-
links in rat liver nuclei using concentrations of 25, 100, or 250 mM. No statistically significant
increase in 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.
4.8.4.1.3. Chromosomal aberrations
Chloral hydrate induced aneuploidy in vitro in multiple Chinese hamster cell lines
(Furnus et al.. 1990; Nataraian et al.. 1993; Warr et al.. 1993) and human lymphocytes (Sbrana et
al.. 1993; 1990) but not mouse lymphoma cells (Harrington-Brock et al.. 1998). In vivo studies
performed in various mouse strains led to increased aneuploidy in spermatocytes (Liang and
Pacchierotti. 1988; Miller and Adler. 1992; Russo et al.. 1984) but not oocytes (Mailhes et al..
1988) or bone marrow cells (Leopardi et al.. 1993; Xu and Adler. 1990).
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
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-390 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
male mice results in a significant increase in frequencies of hyperploidy in metaphase II cells.
This hyperploidy was thought to have arisen from chromosomal nondisjunction 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 Brinklev. 1985). Chloral
hydrate also causes meiotic delay, which may be associated with aneuploidy (Miller and Adler.
1992). Chloral hydrate has been shown to induce micronuclei but not structural chromosomal
aberrations in mouse bone-marrow cells. Micronuclei induced by nonclastogenic agents are
generally believed to represent intact chromosomes that failed to segregate into either daughter-
cell nucleus at cell division (Russo and Levis. 1992a; Xu and Adler. 1990). Furthermore, chloral
hydrate-induced micronuclei in mouse bone-marrow cells (Russo and Levis. 1992a) and in
cultured mammalian cells (Bonatti et al.. 1992; Degrassi and Tanzarella. 1988) 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;
Eastmond and Tucker. 1989; Hennig et al.. 1988). Therefore, both TCE and chloral hydrate
appear to increase the frequency of micronuclei.
Allen et al. (1994) exposed male C57B1/6J mice to a single i.p. injection of 0-, 41-, 83-,
or 165-mg/kg chloral hydrate. Spermatids were harvested at 22 hours and 11, 13.5, and 49 days
following exposure (Allen et al.. 1994). Harvested spermatids were processed to identify both
kinetochore-positive micronucleus (aneugenicity) and kinetochore-negative micronucleus
(clastogenicity). 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. This study is in contrast with other
studies (Bonatti et al.. 1992; Degrassi and Tanzarella. 1988) that 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 nonsmoking
donors (Sbrana et al.. 1993). Cells were exposed for 72 and 96 hours at doses between 50 and
250 |ig/mL, No increase in percentage hyperdiploid, tetraploid, or endoreduplicated cells was
observed when cells were exposed for 72 hours at any doses tested. However, at 96 hours of
exposure, a significant increase in hyperdiploid was observed at one dose (150 (j.g/mL) and was
not dose dependent. Tetraploidy was significantly increased at 137 mg/mL, again without dose-
dependency.
Ikbal et al. (2004) assessed genotoxicity (i.e., induction of micronuclei) in cultured
peripheral blood lymphocytes of 18 infants (age range of 31-55 days) before and after
administration of a single dose of chloral hydrate (50 mg/kg of body weight) for sedation before
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-391 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
a hearing test. A significant increase in micronuclei frequency was observed after administration
of chloral hydrate.
Analysis of chloral hydrate treated mouse lymphoma cell lines for chromosomal
aberrations resulted in a nonsignificant 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 1.5 hours
(Turnus et al.. 1990). A nonstatistically significant increase in frequency of chromosomal
aberrations was observed only at 0.002 and 0.003% concentrations, with the increase not being
dose dependent. In this study, it should be noted that the cells were only exposed for 1.5 hours to
chloral hydrate and cells were allowed to grow for 48 hours (two cell cycles) to obtain similar
mitotic indices before analyzing for chromosomal aberrations. No information on cytotoxicity
was provided except that higher doses decreased the frequency of mitotic cells at the time of
fixation.
In vivo chromosome aberration studies have mostly reported negative or null results
(Leuschner and Leuschner. 1991; Liang and Pacchierotti. 1988; Mailhes et al.. 1993; Russo and
Levis. 1992a; Xu and Adler. 1990) with the exception of one study (Russo et al.. 1984) in an F1
cross of mouse strain between C57B1/Cne x C3H/Cne.
4.8.4.1.3.1. 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 (Allen et
al.. 1994; Beland. 1999; Bonatti et al.. 1992; Degrassi and Tanzarella. 1988; Giller et al.. 1995;
Grawe et al.. 1997; Harrington-Brock et al.. 1998; Ikbal et al.. 2004; Leopardi etal.. 1993;
Leuschner and Leuschner. 1991; Lynch and Parry. 1993; Marrazzini etal.. 1994; Nesslany and
Marzin. 1999; Nutlev et al.. 1996; Parry et al.. 1996; Russo and Levis. 1992a. b; Seelbach et al..
1993; Van Hummel en and Kirsch-Volders. 1992). Some studies have attempted to make
inferences regarding aneuploidy induction or clastogenicity as an effect of chloral hydrate.
Aneuploidy results from defects in chromosome segregation during mitosis and is a common
cytogenetic feature of cancer cells. 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
larvae.
4.8.4.1.3.2. Sister chromatid exchanges (SCEs)
SCEs were assessed by Ikbal et al. (2004) in cultured peripheral blood lymphocytes of
18 infants (age range of 31-55 days) before and after administration of a single dose of chloral
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-392 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
hydrate (50 mg/kg of body weight) for sedation before a hearing test. The authors report a
significant increase in the mean number of SCEs, from before administration (7.03 ±0.18
SCEs/cell) and after administration (7.90 ± 0.19 SCEs/cell), with each of the 18 individuals
showing an increase with treatment. Micronuclei were also significantly increased. SCEs were
also assessed by Gu et al. (1981) 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.
4.8.4.1.4. 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, which induced cellular transformation.
4.8.4.2. Bacterial and Fungal Systems
4.8.4.2.1. 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 (Beland. 1999; Giller et al.. 1995; Haworth et al.. 1983; Ni et al.. 1994). Waskell
(1978) studied the effect of chloral hydrate along with TCE and its other metabolites. Chloral
hydrate was tested at different doses (1.0-13 mg/plate) in different S. typhimurium strains
(TA98, TA100, TA1535) for gene mutations using the Ames assay. No revertant colonies were
observed in strains TA98 or TA1535 both in the presence and absence of S9 mix. Similar results
were obtained by Leuschner and Leuschner (1991). However, in TA100, 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
recrystallized from one to six times from chloroform, and the authors describe this as crude
chloral hydrate. However, this positive result is consistent with other studies in this strain as
noted above. 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 10-mM chloral hydrate (Crebelli etal.. 1985). Studies of mitotic crossing-over
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-393 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
in A. nidulans have been negative, while these same studies were positive for aneuploidy
(Crebelli et al.. 1985. 1991; Kafer. 1986; Kappas. 1989V
Two studies were conducted in S. cerevisiae to understand the chromosomal
malsegregation as a result of exposure to chloral hydrate (Albertini. 1990; Sora and Agostini
Carbone. 1987). Chloral hydrate (1-25 mM) was dissolved in sporulation medium, and the
frequencies of various meiotic events such as recombination and disomy were analyzed. Chloral
hydrate inhibited sporulation as a function of dose and increased diploid and disomic clones.
Chloral hydrate was also tested for mitotic chromosome malsegregation using S. cerevisiae
D61 M (Albertini. 1990). The tester strain was exposed to a dose range of 1-8 mg/mL. An
increase in the frequency of chromosomal malsegregation was observed as a result of exposure
to chloral hydrate.
Limited analysis of chloral hydrate mutagenicity has been performed in Drosophila
(Beland. 1999; Zordan et al.. 1994). Of these two studies, chloral hydrate was positive in the
somatic mutation wing spot test (Zordan et al.. 1994). equivocal in the induction of sex-linked
lethal mutation when administered in feed, but negative when exposed via injection (Beland.
1999).
4.8.4.3. Summary
Chloral hydrate has been reported to induce micronuclei formation, aneuploidy, and
mutations in multiple in vitro systems and in vivo. In vivo studies are limited to increased
micronuclei formation mainly in mouse spermatocytes. CH is positive in some in vitro
genotoxicity assays that detect point mutations, micronuclei induction, chromosomal aberrations,
and/or aneuploidy. The in vivo data exhibit mixed results (Allen et al.. 1994; Leuschner and
Beuscher. 1998; Mailhes et al.. 1993; Nutlev et al.. 1996; Xu and Adler. 1990). Most of the
positive studies show that chloral hydrate induces aneuploidy. 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. Some have suggested that chloral hydrate may act
through a mechanism of spindle poisoning, resulting in numerical changes in the chromosomes,
but some data also suggest induction of chromosomal aberrations. These results are consistent
with tetrachloroethylene, albeit there are more limited data on tetrachloroethylene for these
genotoxic endpoints.
4.8.5. Trichloroacetyl Chloride
Trichloroacetyl chloride results from oxidative metabolism of tetrachloroethylene. The
limited genotoxicity studies of this metabolite are described below and listed in Table 4-48.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-394 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
4.8.5.1. Bacterial Systems
4.8.5.1.1. Gene mutation
The genotoxicity of trichloroacetyl chloride has been studied in S. typhimurium with
inconsistent results. Reichert et al. (1983) found no mutagenicity of trichloroacetyl chloride
exposed in a liquid suspension to S. typhimurium TA98 and TA100 strains with and without S9
activation. A second study (DeMarini et al.. 1994) evaluated genotoxicity in S. typhimurium
TA100 in the vapor state and found trichloroacetyl chloride to be positive in the presence and
absence of S9 activation, but inducing predominantly GC-to-TA transversions (the predominant
background mutation). Trichloroacetyl chloride was negative for prophage induction in E. coli
in the same study (DeMarini et al.. 1994).
4.8.6. Tetrachloroethylene (PCE) Epoxide
Tetrachloroethylene epoxide, a hypothesized intermediate in tetrachloroethylene P450
oxidative metabolism (Henschler. 1977; Henschler and Bonse. 1977). has been investigated in
only one published study. This study is described below and listed in Table 4-48.
4.8.6.1. Bacterial Systems
4.8.6.1.1. Gene mutation
In a study examining the genotoxicity of multiple chloroepoxides, tetrachloroethylene
epoxide (0, 0.5, 1.3, 2.5, 5.0, 25.0 mM, closed system) was mutagenic in S. typhimurium
TA1535 but not in E. coli WP2 uvrA (Kline et al.. 1982). Mutagenicity was observed at the
lower doses in S. typhimurium, but not at higher doses, most likely due to cytotoxicity at the high
doses.
4.8.7. Trichloroethanol (TCOH)
4.8.7.1. Bacterial Systems
4.8.7.1.1. Gene mutation
Limited studies are available on the effect of TCOH on genotoxicity (see Table 4-47).
TCOH is negative in the S. typhimurium assay using the TA100 strain (Bignami etal.. 1980;
DeMarini et al.. 1994; Waskell. 1978). A study by Beland (1999) using S. typhimurium strain
TA104 did not induce reverse mutations without exogenous metabolic activation, however, did
increase mutant frequency in the presence of exogenous metabolic activation at a dose above
2,500 |ig/plate. TCOH has not been evaluated in other recommended screening assays.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-395 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Therefore, the database is limited for the determination of TCOH genotoxicity (summarized in
Table 4-48).
4.8.8. £-(l,2,2-Trichlorovinyl)-Z-Cysteine (1,2-TCVC), 5-Trichlorovinyl Glutathione
(TCVG), j^LAcetyl-5-(l,2,2-Trichlorovinyl)-Z-Cysteine (NAcTCVC)
Limited studies have been performed examining the genotoxicity of three metabolites
from the GSH-conjugation metabolic pathway of tetrachloroethylene. The results for all three
are described below and summarized in Table 4-48.
4.8.8.1. Bacterial Systems
4.8.8.1.1. Gene mutation
TCVG produced from tetrachloroethylene in isolated perfused rat liver and excreted into
bile, in the presence of a rat kidney fraction, was mutagenic in Salmonella, as was purified
TCVG (Vamvakas et al.. 1989c). This study performed the Ames assay in S. typhimurium
TA100, TA98, and TA2638 with tetrachloroethylene, TCVG, and bile from liver perfusate
following tetrachloroethylene exposure in rats, demonstrated that the GST-metabolites or
tetrachloroethylene in the presence of bile containing GST led to gene mutations in
S. typhimurium TA100. Dreessen et al. (2003) also demonstrated for TCVG an unequivocal
dose-dependent mutagenic response in the TA100 strain in the presence of the rat kidney
S9-protein fraction; TCVC was mutagenic without metabolic activation in this strain. In a
separate study, the tetrachloroethylene metabolite TCVC (1-10 nmol/plate) was also positive in
Salmonella strains TA98 and TA100 but not strain TA2638, and inhibition of P-lyase activity
was blocked by the addition of aminooxyacetic acid (AOAA) (Dekant et al.. 1986d). A
subsequent study from this same group indicated that Salmonella also were capable of
deacetylating the urinary metabolite NAcTCVC (50-100 nmol/plate) when TA100 showed a
clear positive response in the Ames assay without exogenous activation (Vamvakas et al.. 1987).
Addition of cytosolic protein increased this mutagenicity, while addition of a P-lyase inhibitor
(AOAA) decreased it.
4.8.8.2. Mammalian Systems
4.8.8.2.1. Unscheduled DNA synthesis
Vamvakas et al. (1989a) reported concentration-related increases in unscheduled DNA
synthesis (UDS) in LLC-PK1 (a porcine kidney cell line) exposed to TCVC, with the effect
abolished by a P-lyase inhibitor. This effect was observed at exposure to 5 x 10 6—10 5 M
TCVC for 24 hours. This study also measured LDH release to determine cytotoxicity at the
same doses, and no increases in LDH were observed at these doses.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-396 DRAFT-DO NOT CITE OR QUOTE
-------
1
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-397 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-48. Genotoxicity of additional tetrachloroethylene metabolites—all
systems
Metabolite
Test system/endpoint
Doses
(LED or
HID)a
Resultsb
Reference
With
activation
Without
activation
Chloral
S. typhimurium TA100, increased
mutation frequency
NA
+
possible
Sato et al. (1985)
Oxalic acid
Sclerotinia sclerotiorum, DNA
fragmentation
10 mM
NT
+
Kim et al. (2008)
Madin-Darby cultured canine
kidney cells, renal prothrombin
fragment-1 mRNA expression
0.09 mM
NT
+
Moryama et al.
(2005)
Crepis capillaris, chromosomal
aberrations
1.0 mM
NT
(+)
Shevchenko et al.
(1985)
Trichloroethanol
(TCOH)
S. typhimurium TA100, 98,
reverse mutation
7,500 ng/plate
-
-
Waskell (1978)
S. typhimurium TA100, reverse
mutation
0.5 |ig/cm3
vapor
-
-
DeMarini et al.,
(1994)
S. typhimurium TA104, reverse
mutation
2,500 ng/plate
+
-
Beland, (1999)
S. typhimurium TA100, 1535
reverse mutation
NA
-
-
Bignami et
al. (1980s)
Sister chromatid exchanges
NA
NA
+
Gu et al. C198D
Trichloroacetyl
chloride
PRB, X Prophage induction, E.
coli WP2
10,000
-
-
DeMarini et al.,
(1994)
SAO, S. typhimurium TA100,
reverse mutation
2.6
+
+
DeMarini et al.,
(1994)
S. typhimurium TA100, increased
mutation frequency
5 |ig/mL
-
-
Reichert et al.,
(1983)
Trichlorovinyl-
glutathione
(TCVG)
S. typhimurium TA100, reverse
mutation
100 nmol/plate
+
-
Dreessen et al.
(2003)
S. typhimurium TA100, increased
mutation frequency
25 nmol/plate
(with)
250-500
nmol/plate
(without)
+
(+)
Vamvakas et al.
(1989b)
Cultured porcine LLC-PK1
(kidney) cells, unscheduled DNA
synthesis, in vitro
7.5 x 10~6M
NT
+
Vamvakas et al.
(1989c)
Trichlorovinyl-
cysteine (TCVC)
S. typhimurium TA100, reverse
mutation
50 nmol/plate
NT
+
Dreessen et al.
(2003)
Cultured porcine LLC-PK1
(kidney) cells, unscheduled DNA
synthesis, in vitro
5 x 10~6 M
NT
+
Vamvakas et al.
(1989a)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-398 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Table 4-48. Genotoxicity of additional tetrachloroethylene metabolites—all
systems (continued)
Metabolite
Test system/endpoint
Doses
(LED or
HID)a
Resultsb
Reference
With
activation
Without
activation
NAcTCVC
S. typhimurium TA100, increased
mutation frequency
<50 nmol°
+
+
Vamvakas et al.
(1987s)
PCE oxide
S. typhimurium TA1535, reverse
mutation
2.5 mM
NT
+
Kline etal. (1982)
E. coli WP2 uvrA, reverse
mutation
25 mM
NT
—
Kline etal. (1982)
aLED, lowest effective dose; HID, highest ineffective dose; NA = not available.
bResults: + = positive; (+) = weakly positive; - = negative; NT = not tested.
°Lower-level concentrations that indicate mutagenicity are not specified in Vamvakas et al. (1987).
4.8.9. TCVC Sulfoxide
TCVC sulfoxide does not appear to have been investigated for genotoxicity.
4.8.10. Synthesis and Overall Summary
Tetrachloroethylene and its metabolites (TCA, DCA, CH, TCVC, TCVG, and
NAcTCVC) have been evaluated to varying degrees for their genotoxic activity in several of in
vitro systems such as bacteria, yeast, and mammalian cells and, also, in in vivo systems.
Genotoxicity studies of other metabolites (e.g., TCVC sulfoxide, tetrachloroethylene epoxide,
trichloroacetyl chloride, trichloroethanol) are limited or nonexistent but are discussed where
available.
The results of a large number of in vitro genotoxicity tests in which tetrachloroethylene
was the test agent do not clearly support the conclusion that tetrachloroethylene exhibits direct
mutagenic activity in the absence or presence of the standard S9 fraction (Bartsch et al.. 1979;
Connor et al.. 1985; DeMarini et al.. 1994; Greim et al.. 1975; Hardin et al.. 1981; Haworth et
al.. 1983; Kringstad et al.. 1981; Milman et al.. 1988; NTP. 1986a; Roldan-Ariona et al.. 1991;
Shimada et al.. 1985; Warner et al.. 1988; Watanabe et al.. 1998) (summarized in Table 4-40). A
more recent study demonstrated cytotoxicity but not genotoxicity of tetrachloroethylene in an
S. typhimurium strain (YG7108pin3ERb5) with enhanced metabolic activity (transformed with
CYP2E1, cytochrome P450 reductase, and cytochrome b5) (Emmert et al.. 2006). PCE was
negative in the parent strain (YG7108) at all doses in the presence of S9. However, when
tetrachloroethylene was activated with rat liver GST, GSH, and a rat kidney fraction,
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-399 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
tetrachloroethylene exhibited a clear dose response ("Vamvakas et al.. 1989c). These findings
support a role of metabolic activation of tetrachloroethylene in its in vitro genotoxicity.
Limited in vivo studies of tetrachloroethylene are inconsistent, with only negative
(Bronzetti et al.. 1983; NTP. 1986b) or equivocal (Beliles et al.. 1980; Cederberg et al.. 2010)
genotoxicity assay results demonstrated following inhalation or oral exposure to
tetrachloroethylene in animals (see Table 4-39). Intraperitoneal injection assays have
demonstrated both negative (NTP. 1986a) as well as positive results for different genotoxicity
endpoints (Walles. 1986). Assays of clastogenic effects following inhalation exposure in
humans have shown inconsistent results and are suggested to be related to coexposures (Ikeda et
al.. 1980; Seiji et al.. 1990). Studies of chromosomal aberrations following exposure to
tetrachloroethylene are mostly negative (Galloway et al.. 1987; NTP. 1986a; Sofuni et al.. 1985).
but positive results have been observed in vivo (Murakami and Horikawa. 1995) and in vitro
studies with enhanced metabolic activation (Dohertv et al.. 1996).
TCA, an oxidative metabolite of tetrachloroethylene, exhibits little, if any, genotoxic
activity in vitro (see Tables 4-41 and 4-42). TCA did not induce mutations in S. typhimurium
strains in the absence of metabolic activation or in an alternative protocol using a closed system
(DeMarini etal.. 1994; Giller et al.. 1997; Kargalioglu et al.. 2002; Nelson et al.. 2001; Rap son
et al.. 1980; Waskell. 1978). but a mutagenic response was induced in TA100 in the Ames
fluctuation test (Giller et al.. 1997). 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 (Mackav et al.. 1995). Measures of DNA-repair responses in bacterial systems have
shown induction of DNA repair reported in S. typhimurium but not in E. coli. Mutagenicity in
mouse lymphoma cells was only induced at cytotoxic concentrations (Harrington-Brock et al..
1998). TCA was positive in some genotoxicity studies in in vivo mouse, newt, and chick test
systems (Bhunya and Behera. 1987; Bhunya and Jena. 1996; Birner et al.. 1994; Giller et al..
1997). DNA unwinding assays have either shown TCA to be much less potent than DCA
(Nelson and Bull. 1988) or negative (Nelson et al.. 1989; Styles et al.. 1991). Due to limitations
in the genotoxicity database, the possible contribution of TCA to tetrachloroethylene
genotoxicity is unclear.
DCA, a chloroacid metabolite of tetrachloroethylene, has also been studied using
different types of genotoxicity assays (see Tables 4-43 and 4-44). Although limited studies are
conducted for different genetic endpoints, DCA has been demonstrated to be mutagenic in the
S. typhimurium assays, in vitro (DeMarini et al.. 1994; Kargalioglu et al.. 2002; Plewa et al..
2002) in some strains, in a mouse lymphoma assay (Harrington-Brock et al.. 1998). in vivo
cytogenetic tests (Fuscoe et al.. 1996; Leavitt et al.. 1997). in the micronucleus induction test,
using the Big Blue mouse system, and in other tests (Chang et al.. 1989; DeMarini et al.. 1994;
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-400 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Fuscoe et al.. 1996; Gu et al.. 1981; Harrington-Brock et al.. 1998; Leavitt et al.. 1997; Nelson
and Bull 1988; Nelson et al.. 1989). DCA can cause DNA strand breaks in mouse and rat liver
cells following in vivo exposure in mice and rats (Fuscoe et al.. 1996). Because of uncertainties
as to the extent of DCA formed from tetrachloroethylene exposure, inferences as to the possible
contribution from DCA genotoxicity to tetrachloroethylene toxicity are difficult to make.
Chloral hydrate is mutagenic in the standard battery of screening assays (see Tables 4-45,
4-46, and 4-47). 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 (Gualandi. 1987; Kafer. 1986; Singh and Sinha. 1976. 1979; Sora and Agostini Carbone.
1987). cultured mammalian somatic cells (Degrassi and Tanzarella. 1988). and spermatocytes of
mice (Liang and Pacchierotti. 1988; Russo et al.. 1984). 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 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, resulting in numerical changes in the chromosomes. These results are consistent with
tetrachloroethylene, albeit there are limited data on tetrachloroethylene for these genotoxic
endpoints.
The genotoxicity analysis of other metabolites (e.g., trichloroacetyl chloride,
tetrachloroethylene epoxide, trichloroethanol) is limited (see Table 4-48). Trichloroacetyl
chloride was found to be mutagenic in S. typhimurium when exposed in vapor phase (DeMarini
et al.. 1994) but not in liquid phase (Reichert et al.. 1983); tetrachloroethylene epoxide was
found to be mutagenic in S. typhimurium but not E. coli (Kline et al.. 1982); and trichloroethanol
was found to be negative in three (Bignami et al.. 1980; DeMarini et al.. 1994; Waskell. 1978) of
four mutagenicity studies (Beland. 1999). These results are limited, and further studies are
needed to make any conclusions on the genotoxicity of these metabolites.
Although also limited, genotoxicity tests for the GSH conjugation metabolites are
positive (see Table 4-48). These include 1,2-TCVC, TCVG, and NAcTCVC. In the one
mammalian study, unscheduled DNA synthesis in porcine kidney cells was observed to increase
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-401 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
in a dose-dependent manner following exposure to TCVC (Vamvakas et al.. 1989b).
Mutagenicity assays found TCVG (Dreessen et al.. 2003; Vamvakas et al.. 1989c) and
NAcTCVC (Vamvakas et al.. 1987) to be mutagenic in the presence of activation, while TCVC
was mutagenic even in the absence of activation (Dreessen et al.. 2003 )(Dekant et al., 1986).
In summary, tetrachloroethylene has been shown to induce some genotoxic effects
(micronuclei induction following in vitro exposure, DNA binding, and SSBs in tumor tissue), but
these result are inconsistent. A number of in vitro mutagenicity (Ames) tests of
tetrachloroethylene have largely been negative in the absence or presence of the standard S9
fractions. Positive results have been observed in tests of conditions where metabolites of the
GSH pathway are generated. These support a role of metabolic activation of tetrachloroethylene
in its genotoxicity. Consistent with this view, positive results have been reported when the GSH
metabolites were used as the test agent, and certain of the oxidative metabolites (especially
DCA) are also mutagenic. TCVC is the most potent bacterial mutagen of the tetrachloroethylene
metabolites and induces UDS in a porcine kidney cell line; TCVG and NAcTCVC are also
mutagenic in bacteria.
There are several challenges in interpreting the genotoxicity results obtained from
tetrachloroethylene exposure. Because of the volatile nature of tetrachloroethylene, there could
be false negative results if proper precautions are not taken to limit evaporation, such as the use
of a closed sealed system. The adequacy of the enzyme-mediated activation of
tetrachloroethylene in vitro tests is another consideration. For example, it is not clear if standard
S9 fractions can adequately recapitulate the complex in vivo metabolism of tetrachloroethylene
to reactive intermediates, which, in some cases, entails multiple sequential steps involving
multiple enzyme systems (e.g., CYP, GST, etc.). In addition, the relative potency of the
metabolites in vitro may not necessarily inform their relative contribution to the overall
mechanistic effects of the parent compound, tetrachloroethylene. Furthermore, although
different assays provided data relevant to different types of genotoxic endpoints, not all effects
that are relevant for carcinogenesis are encompassed. The standard battery of prokaryotic as
well as mammalian genotoxicity test protocols typically specify the inclusion of significantly
cytotoxic concentrations of the test compound.
In conclusion, uncertainties with regard to the characterization of tetrachloroethylene
genotoxicity remain. This is primarily because in vivo tests of tetrachloroethylene have been
equivocal, with at most, modest evidence of genotoxic effects in rodent tumor tissues examined
(including mouse liver and rat kidney) following exposure at tumorigenic doses. However, no
evidence is available regarding the potential contribution of tetrachloroethylene genotoxicity to
other rodent tumor types (particularly, MCL, testes, and brain). Ames assays of
tetrachloroethylene have yielded largely negative results. The tetrachloroethylene metabolites
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-402 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
TCVG, TCVC, NAcTCVC, tetrachloroethylene oxide, and DCA are genotoxic, but not all such
metabolites have been sufficiently tested in the standard screening battery to support clear
conclusions about their genotoxic potential. However, the predominance of positive data for
these metabolites supports their potential genotoxicity following in situ production and/or
bioactivation. This, in turn, supports the view that contribution of genotoxicity to
tetrachloroethylene carcinogenesis cannot be ruled out for one or more target organs. Additional
testing of the genotoxicity of tetrachloroethylene and its metabolites (particularly those from the
GSH conjugation pathway) using state-of-the-art methods and in a more comprehensive panel
of tumor tissues is warranted.
4 9 SUSCEPTIBLE POPULATIONS
Variation in response to tetrachloroethylene may be due to age, gender, genetics, and
race/ethnicity, as well as differences in lifestyle factors, nutrition, preexisting disease status,
socioeconomic status, and multiple exposures. These could be potential modifying risk factors
that play an important role in determining an individual's susceptibility to chemical exposures
and are discussed below.
4.9.1. Life-Stages
Individuals in one life-stage are physiologically, anatomically, and biochemically unique
from individuals in another life-stage. Early and later life-stages differ greatly from mid-life-
stages in body composition, organ function, and many other physiological parameters that can
influence the toxicokinetics of parent chemicals and their metabolites from the body (ILSL
1992). This section presents and evaluates the pertinent published literature available to assess
how individuals of early life-stages (see Section 4.9.1.1) and later life-stages (see Section
4.9.1.2) may respond differently to tetrachloroethylene than adults. The limited data on
tetrachloroethylene exposure suggest that these populations—particularly individuals in early
life-stages—may have greater susceptibility than does the general population.
4.9.1.1. Early Life-Stages
4.9.1.1.1. Early life-stage-specific exposures
Section 2.2 describes the various exposure routes of concern for tetrachloroethylene. For
all postnatal life-stages, the primary exposure routes of concern include inhalation (see
Section 2.2.1) and contaminated water (see Section 2.2.2). Ingestion of contaminated food or
soil is also a possible exposure route (see Section 2.2.3), as is direct ingestion (see Section 2.2.5).
In addition, certain exposure pathways to tetrachloroethylene are unique to early life-stages, such
as through placental transfer or via breast milk ingestion (see Section 2.2.4), or may be increased
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-403 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
during early or later life-stages. Other reviews of the reproductive and developmental effects of
tetrachloroethylene exist (Beliles. 2002; Bove et al.. 2002; Brown Dzubow et al.. 2010;
T abac ova. 1986; van der Gulden and Zielhuis. 1989) (Danielsson, 1990.
Prenatal. In utero, lipophilic substances are known to cross the placental barrier (Herrera
et al., 2006). There is biological plausibility of transfer of tetrachloroethylene across the human
placental barrier as measured in fetal blood and amniotic fluid in rodents (Ghantous et al.. 1986;
Szakmary et al.. 1997). Fetal blood concentrations have been modeled for human exposure
(Gentry et al.. 2003).
Inhalation. Inhalation exposures may be altered for early life-stages compared to adults,
because children have increased ventilation rates (both intake and exhalation) per kg body
weight compared to adults (NRC. 1993; U.S. EPA. 2008). These populations spend the majority
of their time indoors (Bateson and Schwartz. 2008; NRC. 1993; U.S. EPA. 2008). where
increased concentrations of tetrachloroethylene have been found compared to those measured
outdoors (U.S. EPA. 2001a). Increased indoor air concentrations have been measured in places
where children may spend time: inside apartments containing dry-cleaned clothing (Thomas et
al.. 1991; Tichenor et al.. 1990). in the homes of dry-cleaning employees (Aggazzotti et al..
1994a; Aggazzotti etal.. 1994b; AT SDR. 1997b). in apartments above or adjacent to dry
cleaners (Altmann et al.. 1995; Chi en. 1997; Garetano and Gochfeld. 2000; McDermott et al..
2005; NYSDOH. 2010; Schreiber. 1993; Schreiber et al.. 2002; Storm et al.. In Press; Verberk
and Scheffers. 1980). in daycare centers adjacent to dry cleaners (NYSDOH. 2005b). in a
classroom exposed to tetrachloroethylene from an air -emission from a small chemical factory"
(Monster and Smolders. 1984a). and in automobiles containing dry-cleaned clothing (Gulvas and
Hemmerling. 1990; Park et al.. 1998). Similarly, increased ambient air concentrations have been
measured in places where children may spend time: outside of a daycare center adjacent to a dry
cleaner (NYSDOH. 2005c). and on a playground near a factory (Monster and Smolders. 1984a).
Adgate and colleagues (Adgate et al.. 2004a; Adgate et al.. 2004b) measured tetrachloroethylene
in outside and indoor air at school, indoor air at home, and using personal samples on children,
and demonstrated that tetrachloroethylene levels are lower in homes with greater ventilation and
in homes in nonurban settings (Adgate et al.. 2004a; 2004b). In addition, inhalation may also
occur indoors during showering or bathing as dissolved tetrachloroethylene in the warm tap
water is volatilized, although dermal exposure is also relevant during these scenarios (Rao and
Brown. 1993).
Ingestion. Due to its lipophilicity, tetrachloroethylene has been found in human breast
milk samples (Bagnell and Ellenberger. 1977; Pellizzari et al.. 1982; Schreiber. 1993. 1997;
Schreiber et al.. 2002; Sheldon et al.. 1985; U.S. EPA. 2001a). as well as in milk from cows
(Wanner et al.. 1982). goats (Hamada and Tanaka. 1995). and rats (Byczkowski and Fisher.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-404 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
1994; Byczkowski et al.. 1994). The breast milk of one woman was found to contain 10-mg/L
tetrachloroethylene 1 hour following a visit to her spouse working at a dry-cleaning
establishment, dropping to 3 mg/L after 24 hours (Bagnell and Ellenberger. 1977).
Tetrachloroethylene has also been measured in the breast milk of two women living in
apartments colocated with a dry-cleaning facility (NYSDOH. 2005c; Schreiber et al.. 2002).
PBPK models have been used to estimate the dose a nursing infant might receive from an
exposed mother's breast milk (Byczkowski and Fisher. 1995; Byczkowski et al.. 1994; Fisher et
al.. 1997; Gentry et al.. 2003; Schreiber. 1993). A PBPK model was also developed and
validated for breast milk ingestion in nursing rats after maternal inhalation exposure (Fisher.
1994). Using different exposure scenarios, Schreiber (1993) predicted that breast milk
concentrations could range from 1.5 |ig/L for a typical residential scenario, 16-3,000 |ig/L for a
residential scenario near a dry cleaner, to 857-8,440 [j,g/L for an occupational scenario.
Assuming that a 7.2-kg infant ingests 700 mL of breast milk per day, Schreiber estimated dose to
the infant could range from 0.0001 to 0.82 mg/kg-day (Schreiber et al.. 1993). Byczkowski and
Fisher (1995) refined the approach used by Schreiber (1993) and found that with the same
residential exposure conditions, the results predicted lower doses to the infant (0.0009-0.202
mg/kg-day). Using milk production and suckling variables, Fisher et al. (1997) estimated the
dose that a human infant might receive after maternal occupational exposure to be 25 ppm/day.
Gentry et al. (2003) modeled a rapid decline in concentration of tetrachloroethylene and TCA
during lactation in humans. Although ingestion of tetrachloroethylene through breast milk may
be a significant pathway of exposure for some infants, it has been suggested that if these infants
live adjacent to or in close proximity of dry-cleaning facilities, the dose received through
ingestion of breast milk will become less important when compared with the dose resulting from
inhalation exposure (McKone and Daniels. 1991; Schreiber. 1997).
Children ingest higher amounts of water per body weight than adults (NRC. 1993; U.S.
EPA. 2008). For infants on formula, ingestion of tetrachloroethylene-contaminated water may
be of concern. Taking into account tetrachloroethylene volatilization in boiling water,
Letkiewicz et al. (1982) estimated that 22% of formula-fed infants received fluids contaminated
with tetrachloroethylene levels found in the water supply. Data showed that about 11%
(0.5 x 22%) of formula-fed infants could receive an increased exposure as compared with adults
on a mg/kg-basis through drinking contaminated water. In addition, incidental water
consumption may occur for children when swimming or bathing (U.S. EPA. 2008).
Children consume a higher quantity of food per body weight compared to adults,
specifically dairy and other foods with high fat content (U.S. EPA. 2008) that have been found to
have elevated concentrations of tetrachloroethylene (see Section 2.2.3). Assuming 100 mg/kg
represents the average tetrachloroethylene concentration in fatty foods such as butter, and using
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-405 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
daily total fat intake rates by age (U.S. EPA. 2008). the daily dose would be 0.46 mg/kg-day for
a 10-kg 1-year-old compared to the daily dose of 0.12 mg/kg-day for a 70-kg adult. Therefore,
there may be concern for ingestion of contaminated dairy products in early life-stages, although
this exposure route for tetrachloroethylene has not been well characterized for any life-stage.
Where contamination occurs, tetrachloroethylene can be measured in soil (U.S. EPA.
2001a). This pathway for ingestion of tetrachloroethylene has not been directly examined. A
clear need exists to evaluate this pathway because children, particularly those with pica, can
ingest high quantities of contaminated soil through hand-to-mouth activity, as has been shown
for lead (U.S. EPA. 2008).
Rare instances of direct ingestion of tetrachloroethylene have been documented,
including a 6-year-old boy who directly ingested 12-16 g of tetrachloroethylene (Koppel et al..
1985).
Dermal. Dermal exposures may be increased for both early life-stages, because infants
have increased surface area-per-body weight-ratio than adults (NRC. 1993; U.S. EPA. 2008).
Although an infant's skin has similar permeability to adults, a premature infant may have
increased permeability (Guzelian et al.. 1992). Dermal exposure for children may occur in a
residential setting from showering, bathing, or swimming in contaminated water, although
inhalation exposure is also relevant during these scenarios (Rao and Brown. 1993; U.S. EPA.
2001a). While dermal exposure is generally not considered a major route of exposure, this route
of exposure is not well characterized for early life-stages (prenatal or postnatal).
4.9.1.1.2. Early life-stage-specific toxicokinetics
Section 3 describes the toxicokinetics of tetrachloroethylene. However, children may
have differential exposure to tetrachloroethylene compared to adults due to age-related
physiological differences. These include body composition, organ function, and many other
physiological parameters that can influence the toxicokinetics of chemicals and their metabolites
from the body (ILSI. 1992; Renwick. 1998). Early life-stage-specific information regarding
toxicokinetics 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 (Byczkowski and Fisher. 1994; Clewell et al.. 2004; Gentry et
al.. 2003; see Section 3.5; Rao and Brown. 1993).
Absorption. As discussed in Section 3.1, exposure may occur via inhalation, ingestion,
and skin absorption. In addition, prenatal exposure may result in absorption via the
transplacental route. For lipophilic compounds such as tetrachloroethylene, percentage adipose
tissue, which varies with age (NRC. 1993). will affect absorption and retention of the absorbed
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-406 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
dose. Absorption into the lungs via inhalation is related to the ventilation rate per body weight,
which is higher in children than in adults (NRC. 1993; U.S. EPA. 2008; WHO. 2006). with an
increased alveolar surface area per kg body weight for the first 2 years (NRC. 1993). Absorption
into the gut from oral ingestion may be altered by gastric pH levels, which are higher in infants
than in adults (WHO. 2006). Absorption during dermal exposure may be affected by the ratio of
surface area, which is higher in infants than in adults (U.S. EPA. 2008; WHO. 2006).
Distribution. The distribution of tetrachloroethylene to specific organs will depend on
organ blood flow and the lipid and water content of the organ, which may vary between life-
stages (NRC. 1993; WHO. 2006). Due to its high lipophilicity, tetrachloroethylene has been
found to distribute widely to all tissues in the body as observed in early lifestages of humans
(Gaillard et al.. 1995; Gamier et al.. 1996; Koppel et al.. 1985) and early lifestages of animals
(Dallas et al.. 1994b; Ghantous et al.. 1986; Savolainen et al.. 1977a; Schumann et al.. 1980;
Szakmary et al.. 1997); however, this is true for adults as well, and it is not clear whether
distribution may vary differentially with life-stage. It should be noted that the total body burden
of tetrachloroethylene increases with age (Clewell et al.. 2004). as would be expected, given that
adult body weight is generally positively correlated with age.
Tetrachloroethylene can cross the placental barrier during prenatal development. Rodent
studies demonstrate that tetrachloroethylene crosses the placental barrier when pregnant dams
are exposed (Ghantous et al.. 1986; Szakmary et al.. 1997). and in humans, it has been shown
that during lactation, tetrachloroethylene distributes to breast milk (NYSDOH. 2005c; Schreiber
et al.. 1993; Sheldon et al.. 1985). However, a noticeable difference exists between the
milk:blood partition coefficients for rats (12) and for humans (2.8; Byczkowski and Fisher.
1994). reflecting the higher fat content of rat milk.
Tetrachloroethylene or its metabolites have been measured in blood of children
(NYSDOH. 2005a. 2010; Popp et al.. 1992; Storm et al.. In Press). A longitudinal study of blood
concentrations of 11 volatile organic chemicals (VOCs) measured in more than 150 poor,
minority children in Minneapolis, MN, found the mean blood tetrachloroethylene levels to be
0.06 ng/mL (Sexton et al.. 2005). When compared to adult data from NHANES III, the blood
level in children was lower (Sexton et al.. 2005). However, these results do not necessarily
represent TK differences between lifestages because the study did not control for exposure
differences between these two cohorts. Lower estimated blood concentrations of
tetrachloroethylene in children compared to adults have also been described in Clewell et al.
(2004). although the variability of the parameters used as well as the results have not been
validated.
Tetrachloroethylene can also cross the blood:brain barrier during both prenatal and
postnatal development; this may occur, to a greater extent, in younger children. Based on the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-407 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
modeled dose of tetrachloroethylene to the brain after a showering/bathing scenario, a study by
Rao and Brown (1993) showed that for a given set of exposures, the younger a person is, the
greater the estimated concentration of tetrachloroethylene in the brain. Modeling showed that
after a 30-minute bathing scenario, a 3-year-old child accumulated higher brain tissue
concentrations of tetrachloroethylene as compared with a 10-year-old and an adult. An autopsy
conducted on the previously mentioned 2-year-old boy found dead after exposure to dry-cleaned
curtains revealed the highest levels of tetrachloroethylene in the brain, 77 mg/kg. Levels in his
blood, heart, and lungs were 66 mg/L, 31 mg/kg, and 46 mg/kg, respectively (Gaillard et al..
1995; Gamier et al.. 1996).
Metabolism. Section 3.3.3 describes the enzymes involved in the metabolism of
tetrachloroethylene. In general, expression of CYP enzymes changes during various stages of
fetal development (Hakkola et al.. 1996a; Hakkola et al.. 1998; Hakkola et al.. 1996b) and during
postnatal development (Clewell et al.. 2004; George et al.. 1995; Hakkola et al.. 1996a; Hakkola
et al.. 1998; Hakkola et al.. 1996b; Tateishi etal.. 1997)(Hakkola et al., 1998b; Shao et al.,
2007). In addition, production of GST enzymes varies significantly during early postnatal
lifestages (McCarver and Hines. 2002; Nakasa et al.. 1997; Raiimakers et al.. 2001)(Dorne et al.,
2001; Mera et al., 1994; Shao et al., 2007).
After maternal oral exposure to tetrachloroethylene it was observed that fetus and infant
blood levels were higher for TCA than for tetrachloroethylene (Gentry et al.. 2003).
demonstrating that metabolism of tetrachloroethylene does occur during these lifestages. In
addition, there is in vitro evidence of an age-related increase in metabolism of
tetrachloroethylene as estimated in the blood (Clewell et al.. 2004; Sarangapani et al.. 2003).
associated with age-related activation of oxidative metabolism pathways, suggesting a decreased
ability to metabolize tetrachloroethylene during early lifestages compared to during adulthood.
One study modeled the role of the age-dependent development of CYP2E1 in oxidative
metabolism (TCA) in the mother and lactating infant (Vieira et al.. 1996). A number of other
human studies suggest that CYP2B6 may also play a role in the metabolism of
tetrachloroethylene (White et al.. 2001). although this enzyme was not detected in placental or
fetal liver samples (Hakkola et al.. 1996a; Hakkola et al.. 1996b). and differences between a
group of 10 prenatal and infant patients showed significantly lower CYP2B6 protein expression
in placental hepatic microsomes as compared with an adult group (Tateishi et al.. 1997). These
findings need to be validated in studies of target tissues in addition to blood to better evaluate
any role of variation and heterogeneity.
Excretion. The major processes of excretion of tetrachloroethylene and its metabolites
are discussed in Sections 3.3 and 3.4, respectively. Excretion profile differences in exhaled
breath and urinary excretion are likely between children and adults. This is due to differences in
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-408 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
ventilation rate, activity level, and the solubility of the compound in blood and tissue, as well as
differences in amounts of water ingested per body weight (NRC. 1993; U.S. EPA. 2008).
Tetrachloroethylene or its metabolites have been measured in exhaled breath (Delfino et
al.. 2003b; Monster and Smolders. 1984b; NYSDOH. 2005a. 2010; Schreiber et al.. 2002; Storm
et al.. In Press), and urine (NYSDOH. 2005c; Popp et al.. 1992; Schreiber et al.. 2002) of
children. However, these studies do not provide clear information whether excretion levels in
children differ from those of adults for a similar exposure concentration.
PBPKModels. A number of PBPK models present toxicokinetic variation between early
lifestages and adulthood for tetrachloroethylene and its metabolites for both humans and animals.
Early lifestage-specific exposure scenarios considered in these models include fetal exposure
(Gentry et al.. 2003) and breast milk exposure (Byczkowski and Fisher. 1995; Byczkowski et al..
1994; Fisher et al.. 1997; Gentry et al.. 2003; Schreiber et al.. 1993)Other PBPK models have
addressed comparisons of early lifestage toxicokinetics with those in adulthood for inhalation
(Mahle et al.. 2007; Pelekis et al.. 2001; Sarangapani et al.. 2003)(Rodriguez et al. 2007),
drinking water (Clewell et al.. 2004). and bathing and showering (Rao and Brown. 1993). When
considering inhalation exposure, Mahle et al. (2007) found no difference in the blood:air
partition coefficient for tetrachloroethylene for children aged 3-10 years compared to adults
(420 years old). This same study reported that rats at PND 10 and at 2 months (adult) have an
age-dependent difference in fat:air, muscle:air, and brain:air partition coefficients, but not for
blood:air, liver:air, or kidney:air (Mahle et al.. 2007). Another study of rats found higher peak
concentrations of tetrachloroethylene in the blood at PND 10 compared to 2 months (adult) after
inhalation exposure, likely due to the lower metabolic capacity of the young rats as observed in
the liver (Rodriguez et al., 2007). Pelekis et al. (2001) found little difference in the suggested
intraspecies uncertainty factor when including lifestage-specific pharmacokinetics. Sarangapani
et al. (2003) also found no age-related difference in tetrachloroethylene blood concentration;
however, this study found that metabolite concentrations were lowest in infancy and increased
with age. For drinking water exposure, Clewell et al. (2004) found an age-related trend in the
average daily dose and cumulative lifetime dose of tetrachloroethylene and its metabolites, with
lower levels of metabolites observed in children compared to higher levels of metabolites
observed in adulthood. In a showering/bathing scenario, Rao and Brown (1993) found that
tetrachloroethylene accumulates in the brain at higher levels in younger versus older children.
Validation and further refinement of the parameters in these PBPK models are necessary, in
particular, modeling of fetal and breast milk exposure, and child-adult differences in partition
coefficients after inhalation, drinking water, and bathing scenarios.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-409 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
4.9.1.1.3. Early life-stage-specific effects
Although limited data exist on tetrachloroethylene toxicity as it relates to early life-
stages, there is enough information to discuss the qualitative differences. In addition to the
evidence described below, Section 4.7 contains information on both human and animal evidence
for reproductive and developmental outcomes such as spontaneous abortion/fetal loss, low birth
weight, IUGR, SGA, congenital abnormalities, sperm quality, developmental delays, and
behavioral changes. Together, Section 4.4 on liver toxicity, Section 4.5 on kidney toxicity,
Section 4.6 on neurotoxicity, and Section 4.8 on toxic effects in other organ systems characterize
a wide array of postnatal developmental effects.
4.9.1.1.3.1. Preconception
Exposures occurring prior to conception may result in adverse reproductive outcomes.
For tetrachloroethylene exposure, adverse outcomes assessed prior to conception include reduced
fertility, altered sperm, and altered reproductive hormones.
Fertility. In humans, limited evidence exists on impacts to fertility. A study of couples
seeking treatment for infertility found that employment in dry cleaning was significantly
associated with infertility among women but not among men, although exposure to
tetrachloroethylene was inferred but not documented (Rachootin and 01 sen. 1983). Another
study observed no impacts on the number of pregnancies or fertility ratio among wives of men
employed as dry cleaners compared to wives employed as laundry workers, although wives of
dry cleaners took longer to become pregnant compared to wives of laundry workers (Eskenazi et
al.. 1991a). Other epidemiological studies have not shown any association between reduced
fertility and working in dry cleaning or exposed to tetrachloroethylene, although these results
were imprecise because the prevalence of exposure was low (Sallmen et al.. 1998; Sallmen et al..
1995). A review of the data by the National Research Council regarding exposures to
tetrachloroethylene, trichloroethylene, or solvent mixtures in drinking water at Camp Lejeune,
NC, found limited/suggestive evidence of an association for female infertility with concurrent
exposure to solvent mixtures, but inadequate/insufficient evidence to determine whether an
association exists for female infertility after exposure cessation, and inadequate/insufficient
evidence to determine whether an association exists for male infertility (NRC. 2009).
In experimental animals, a study found that the percentage of fertilized oocytes in vitro
was reduced in tetrachloroethylene-treated female rats as compared with controls, although this
study found no effect from exposure in drinking water (Berger and Horner. 2003). Other studies
in rats also found no change in fertility (Carney et al.. 2006; Tinston. 1994). and one earlier study
reported an increase in fertility of female rats exposed to tetrachloroethylene (Carpenter. 1937).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-410 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Sperm. Few studies in either humans or animals have examined altered sperm quality,
generally with no observed adverse or consistent effects. Eskenazi and colleagues found that
tetrachloroethylene can have subtle effects on sperm quality (Eskenazi etal.. 1991b); however,
they also reported that altered sperm parameters did not appear to affect reproduction because
wives did not have fewer pregnancies as compared with a national standard (Eskenazi et al..
1991a). A study of couples treated for infertility also examined sperm abnormalities among dry
cleaners but did not see an elevated prevalence of sperm alterations, suggesting that the observed
reduced fertility rate among these couples was related to other reasons (Rachootin and 01 sen.
1983). One rodent study demonstrated inconsistent effects (abnormal sperm at 4 weeks but not 1
or 10 weeks after exposure) in mice, but no adverse effect was observed in rats (Beliles. 2002).
Additionally, reduced testes weight was seen in the offspring of rats after inhalation exposure,
although these were not significant after adjusting for body weight (Tinston. 1994).
Reproductive Hormones. Few studies in either humans or animals have examined altered
hormones related to reproduction, generally with no observed adverse or consistent effects. The
study discussed above of couples seeking treatment for infertility examined employment in dry
cleaning and found inconsistent results for —a ferale diagnosis indicating hormonal
disturbances" among three analyses (Rachootin and 01 sen. 1983). An exploratory study of
menstrual disorders among dry-cleaning workers found associations with unusual cycle length,
menorrhagia, dysmenorrhea, and premenstrual syndrome, but not with oligomenorrhea,
polymenorrhea, irregular cycle, and intermenstrual blood loss (Zielhuis et al.. 1989).
A study of rats exposed to 1,700-ppm tetrachloroethylene did not affect progesterone
levels (Berger and Horner. 2003). The few studies on altered reproductive hormones suggest this
as an area for further research, both in females and males.
4.9.1.1.3.2. Prenatal and birth outcomes
Prenatal and birth outcomes resulting from exposure occurring prior to conception or
during fetal development include fetal death (i.e., spontaneous abortion, perinatal death), birth
defects, and decreased birth weight. It is important to note that maternal toxicity (e.g., reduced
maternal body-weight gain) may influence adverse outcomes in the offspring and was assessed
in a number of experimental animal studies of tetrachloroethylene exposure (Hardin et al.. 1981;
Narotsky andKavlock. 1995; Schwetz et al.. 1975; Szakmary et al.. 1997; Tinston. 1994).
Pregnancy Loss. Human and animal studies examining pregnancy loss are discussed in
detail in Section 4.7. For humans, both occupational and drinking water studies have examined
fetal loss, an outcome for which there is good retrospective recall, and any bias would result in
an underestimation of the true risk (Wilcox and Hornev. 1984). However, the available studies
may be limited by selection bias and small sample sizes.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-411 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
A number of occupational studies have shown spontaneous abortion or perinatal loss
among women employed as dry cleaners (Bosco et al.. 1987; Doyle et al.. 1997; Kyyronen et al..
1989; 01 sen et al.. 1990). or otherwise exposed occupationally (Lindbohm et al.. 1991; Windham
et al.. 1991). An increased risk of spontaneous abortion was not observed in other studies of
women who were dry cleaners or wives of dry cleaners (Ahlborg. 1990b; Eskenazi etal.. 1991a;
Lindbohm et al.. 1991; McDonald et al.. 1986; McDonald et al.. 1987; Taskinen et al.. 1989).
A few residential studies have examined spontaneous abortion or perinatal loss among
women drinking contaminated water (Aschengrau et al.. 2009a; ATSDR. 1998b; Bove. 1996;
Bove et al.. 1995; Lagakos et al.. 1986) or inhaling VOCs (ATSDR. 2008). with no conclusive
results. Lagakos et al. (1986) found no association with drinking contaminated water and risk of
spontaneous abortion and no association for risk of perinatal death prior to 1970; however, a
positive association was observed for perinatal death since 1970. No association was observed
in Aschengrau et al. (2009a). but the authors note that the differences between occupational and
residential studies may be due to the exposure levels. The National Research Council
determined that there is limited/suggestive evidence of an association for miscarriage with
tetrachloroethylene-contaminated drinking water exposure at Camp Lejeune during pregnancy
(NRC. 2009). No increased risk was observed among women living in a community concerned
about vapor intrusion from VOCs including tetrachloroethylene (ATSDR. 2008).
Fetal loss in experimental animals correlates with the observation of spontaneous
abortions in humans, with varying tendencies for fetal loss depending on species (rodents have a
very low propensity to abort, while rabbits and primates have higher rates). There is evidence of
increased preimplantation loss in rats (Szakmary et al.. 1997). increased resorption of pups after
maternal inhalation in rats and rabbits (Schwetz et al.. 1975; Szakmary et al.. 1997). reduction in
litter size and pup survival in rats and guinea pigs (Kyrklund and Haglid. 1991; Narotsky and
Kavlock. 1995; Szakmary et al.. 1997; Tinston. 1994). spontaneous abortion in rabbits
(Szakmary et al.. 1997). and litters with dead pups (Tinston. 1994). However, fetal loss was not
seen in other in vivo studies (Carney et al.. 2006; Hardin et al.. 1981). In vitro studies of
exposure to tetrachloroethylene show decreased fertilized oocytes (Berger and Horner. 2003).
and increased mortality, malformations, and delayed growth and differentiation of embryos
(Saillenfait et al.. 1995).
Birth Defects. After residential exposure to contaminated drinking water, birth defects
related to in utero exposure in humans include eye/ear anomalies and CNS/chromosomal/oral
cleft anomalies (Lagakos et al.. 1986). A study of residents living in a community with vapor
intrusion including tetrachloroethylene examined birth outcomes and observed a significantly
higher prevalence of total and major cardiac defects (ATSDR. 2006); a follow-up study of this
cohort noted that conotruncal heart malformations were particularly elevated (ATSDR. 2008). A
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-412 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
recent study in Massachusetts of maternal exposure to drinking water contaminated with
tetrachloroethylene reported a 20% increased risk (95% CI: 0.8-1.7) between any maternal
exposure at the time of conception and congenital anomalies (oral cleft anomalies, neural tube
defects, and gastrointestinal and genitourinary malformations) in the offspring after adjustment
for maternal and paternal ages (Aschengrau et al.. 2009); however, this study is inconclusive due
to limited adjustment for potential confounding factors and low statistical power. A hypothesis-
generating ecological study found a 3.5-fold increased risk of oral cleft defects in New Jersey
towns with 410-ppb tetrachloroethylene in drinking water (Bove. 1996; Bove et al.. 1995).
although a case-control study of oral cleft defects from a larger area in New Jersey designed to
test this hypothesis did not confirm the earlier observation (Bove. 1996). Three overlapping
studies similarly did not observe any association with birth defects among women who were dry
cleaners or laundry workers, although the number of exposed cases was very small (Kyyronen et
al.. 1989; Olsen et al.. 1990; Taskinen et al.. 1989). While the NAS has determined that there is
inadequate/insufficient evidence to determine whether an association exists between drinking
water at Camp Lejeune, NC, and congenital malformations (NRC. 2009). a follow-up study is
currently underway to examine the incidences of neural tube defects and oral cleft anomalies
(ATSDR. 2003; NRC. 2009).
In experimental animals, an increase in microphthalmia or anophthalmia in rat offspring
was seen after maternal gavage exposure, but no other evaluation of birth defects was undertaken
in this study (Narotsky andKavlock. 1995). Delayed ossification was observed in mice but not
in rats exposed prenatally (Schwetz et al.. 1975); for skeletal retardation, no significant
differences were observed for exposed mice in another study (Szakmary et al.. 1997). Skeletal
malformations were increased in mice pups after maternal inhalation exposure, but no additional
details were given regarding type of malformation (Szakmary et al.. 1997). and no significant
differences were observed in other studies (Carney et al.. 2006; Schwetz et al.. 1975). Internal
organ malformations were significantly increased in mice exposed in utero (Schwetz et al.. 1975;
Szakmary et al.. 1997). and an in vitro study of rat embryos exposed to tetrachloroethylene
showed increased malformations (Saillenfait et al.. 1995). No birth defects were seen in other
studies of rats (Hardin et al.. 1981; Nelson et al.. 1980; Schwetz et al.. 1975) or rabbits (Hardin et
al.. 1981).
Conclusions about the association of birth defects with exposure to tetrachloroethylene
cannot be drawn from the available epidemiological studies, which contain a number of
deficiencies and uncertainties that may introduce a positive or negative bias on observations. A
clear need exists for better studies of tetrachloroethylene exposure and birth defects. In
particular, given the evidence for heart defects reported in animal studies with exposure to TCE
and its metabolites, TCA (Johnson etal.. 1998; Smith et al.. 1989) and DCA (Epstein et al..
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-413 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
1992; see Sections 4.6.2. 4.7.2. and 4.8.2). there is a need for additional studies of heart defects
after exposure to tetrachloroethylene.
Birth Weight. The epidemiological studies reported equivocal findings on birth weight.
At the military base of Camp Lejeune, NC, babies born to women living in housing that received
drinking water containing VOCs including tetrachloroethylene had a slight decrease in mean
birth weight (-26 g, 90% CI: -43, -9) and an increase in small for gestational age (SGA, 22
weeks gestation) (OR: 1.2, 90% CI: 1.0-1.3), most notably among women who had two or more
prior fetal losses (OR: 2.5, 90% CI: 1.5-4.3), compared to unexposed women; no increase in
preterm births was observed (OR: 1.0, 90% CI: 0.9-1.1) (ATSDR. 1998b: Sonnenfeld et al..
2001). The NAS determined that there is inadequate/insufficient evidence to determine whether
an association exists between contaminated drinking water and decreased birth weight at Camp
Lejeune, NC (NRC. 2009).
Risk of intrauterine growth restriction (IUGR) was seen in an occupational study
(OR: 12.5, no CI given) based on one case exposed to tetra- and trichloroethylene (Windham et
al.. 1991). A second residential study of a community with VOC exposure from vapor intrusion
reported that 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) (ATSDR. 2006). However, the analysis did not adjust for
smoking and sociodemographic factors, which are known to also cause birth weight reductions.
Other residential drinking water (Aschengrau et al.. 2008: Lagakos et al.. 1986) and occupational
(01 sen et al.. 1990) studies showed no association between exposure to tetrachloroethylene and
low birth weight.
In experimental animals, exposure to tetrachloroethylene caused decreased birth weight
(Tinston. 1994) and decreased fetal body weight in some studies of rats (Carney et al.. 2006:
Szakmary et al.. 1997) and mice (Schwetz et al.. 1975). However, no effect on birth weight was
found in other studies of mice (Szakmary et al.. 1997). rats (Hardin et al.. 1981: Schwetz et al..
1975). and rabbits (Hardin et al.. 1981: Szakmary et al.. 1997). Experimental animal studies also
observed decreased weight gain after either pre- or postnatal tetrachloroethylene exposure. A
study in rats demonstrated a reduction in overall pup body weight after preconception, prenatal,
and postnatal inhalation exposure (0-1,000 ppm) through 29 days of age (Tinston. 1994).
Another study found that the offspring of rats exposed to tetrachloroethylene (0-900 ppm)
during late pregnancy (GDs 14-20) had reduced weight gain at postnatal Weeks 3-5, but the
same effect was not observed in those exposed earlier in pregnancy (GDs 7-13) (Nelson et al..
1980).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-414 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
4.9.1.1.3.3. Developmental neurotoxicity
Neurotoxicological effects have been reported after low exposure levels to
tetrachloroethylene in children (see Section 4.6 and Table 4-4) and in animals after prenatal
exposure (see Sections 4.6.2 and 4.7.2). Both human and animal evidence supports an
association between neurodevelopmental effects and tetrachloroethylene exposure. While other
neurotoxic effects are seen in adults (see Table 4-5), decreased VCS has been the main
observation in children.
Visual deficits. Recent studies have examined the visual system as a target of
tetrachloroethylene toxicity in both children and adults. Subjects were New York City apartment
residents (NYSDOH. 2005a. 2010; Schreiber et al.. 2002; Storm et al.. In Press) and employees
and children at a daycare center (NYSDOH. 2005a. b, c; Schreiber et al.. 2002) exposed to
tetrachloroethylene by proximity to dry cleaners. Exposure was measured in indoor air, exhaled
air, and blood levels, and the visual system was assessed by visual contrast sensitivity (VCS) and
color confusion index (CCI).
In the day-care studies, visual tests were not conducted on children at the time of
exposure due to their young age (NYSDOH. 2005c; Schreiber et al.. 2002). and a follow-up
evaluation 4 to 5 years after the colocated dry cleaner closed showed no residual changes in VCS
or CCI (NYSDOH. 2005a. b). There is a possibility that the results of these test results for
children could be due to a learning disability or a developmental delay (Storm and Mazor. 2004).
although these data were not available for the control children (Hudnell and Schreiber. 2004).
The residential studies were designed to assess vision in children and adults living in the
same household colocated near dry cleaners (NYSDOH. 2005a. 2010; Schreiber et al.. 2002;
Storm et al.. In Press). Investigators found that children generally performed better than adults
for both VCS and CCI. Children exposed to tetrachloroethylene performed worse than adults for
VCS for the highest category of exposure compared to both child and adult reference subjects
(NYSDOH. 2005a. 2010). indicating there may be increased susceptibility for children. Poorer
CCI scores were associated with levels of tetrachloroethylene-in exhaled breath in children but
not in adults (NYSDOH. 2005a). but a later study found that CCI was not associated with levels
of tetrachloroethylene exposure in either children or adults (NYSDOH. 2010). The investigators
noted that exposure to tetrachloroethylene was highly correlated with race and income, but small
sample sizes made it difficult to fully examine this correlation (NYSDOH. 2005a. 2010).
Additionally, a case study reported reduced VCS in a 2.5-year-old boy after prenatal
exposure to tetrachloroethylene (Till et al.. 2003). as do reports from Till et al.(2001a; 2005;
2001b) and Laslo-Baker et al. (2004) showing visual system functioning deficits in young
children of mothers exposed to multiple solvents during pregnancy, although exposure to
tetrachloroethylene was not uniquely identified. An important factor to consider in the testing of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-415 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
visual function in children is the requirement for sustained attention and cognition (Tschopp et
al., 1998)(Scharre etal.. 1990). For this reason, visual testing of young children, particularly,
contrast sensitivity in children younger than 6 years of age, is difficult, and responses of young
children are more variable than those of adults (Scharre et al.. 1990). A need exists for
developing methods to better evaluate contrast sensitivity effects in the very young-aged child.
Acute Neurotoxicity. Acute neurotoxicity has been observed in children exposed to
tetrachloroethylene. A case study by Koppel et al. (1985) reported that a 6-year-old boy who
directly ingested 12-16 g of tetrachloroethylene suffered from drowsiness, vertigo, agitation, and
hallucinations before lapsing into a coma. One hour after ingestion, his blood
tetrachloroethylene concentration was 21.5 mg/L. He recovered, but because follow-up testing
was not conducted, any potential long-term effects of the exposure are unknown (Koppel et al..
1985). Gamier et al. (1996) reported mild CNS depression (dizziness and drowsiness were the
most common symptoms, along with nausea, vomiting, headache, tinnitus, unconsciousness)
after exposure to coin-operated dry-cleaned items in 5 cases of children and 24 cases of adults
but did not separate the analysis by age group. Gamier et al. also described two additional
reports (published in Danish) of unconsciousness in a 9-year-old boy who died after using his
dry-cleaned sleeping bag (Korn. 1977). and in a 7-year-old girl who was left in a car with dry-
cleaned clothing (Larsen et al.. 1977).
Brain neurochemistry. There are no studies in humans measuring brain neurochemistry
after exposure to tetrachloroethylene, in either children or adults. In experimental animals,
altered brain biochemistry (fatty acid composition) was seen in the offspring after gestational
exposure to rats and guinea pigs (Kyrklund and Haglid. 1991; Nelson et al.. 1980). These studies
do not necessarily indicate effects on brain neurochemistry after gestational exposure compared
to adult exposure.
Neurobehavior. Two cohorts examined behavior in children after exposure to
tetrachloroethylene, with neither finding any association. In the daycare study described above,
18 children were examined for neurobehavioral deficits using a battery of tests for both
neurological and behavioral function. Tests were conducted approximately 5 weeks after
exposure ceased (at ages 4-5 years old) (NYSDOH. 2005c). and again in 13 children at a follow-
up evaluation 4-5 years later (NYSDOH. 2005a) and reported no functional change at either
examination. A large retrospective cohort study in Cape Cod, MA, examined prenatal and
postnatal exposure to drinking water contaminated by tetrachloroethylene leaching into water
distribution pipes (Janulewicz et al.. 2008). Children born in 1969-1983 were included in the
analysis (n = 2,086), and followed during 2002-2003. Data were collected from birth
certificates and self-administered questionnaires including information on medical history for the
mother and child, potential solvent exposure, and water use. Cumulative exposure during the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-416 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
prenatal period was estimated to be 4 x 10 5 to 1,328 g, and exposure during the postnatal period
was estimated to be 2.9 x 10 4 to 3,310 g. No statistically significant association was observed
with attention, learning, or behavioral functions.
Rats exposure to tetrachloroethylene during pregnancy resulted in developmental delay as
measured by the ascent test and rotorod test (Nelson et al.. 1980). although another study found
no adverse effects for running wheel activity, avoidance behaviors, or operant conditioning
(Nelson et al.. 1980). Other effects observed include altered motor activity (Szakmary et al..
1997; Tinston. 1994). decreased muscular strength (Szakmary et al.. 1997). and short-term
reduced response to sound in pups (Tinston. 1994).
Young animals have also been directly exposed postnatally to tetrachloroethylene. Daily
exposure of rats to 1,000-ppm tetrachloroethylene on PNDs 6-29 resulted in sedation and
hypothermia, but the effect ceased 2 hours or less after exposure ended (Tinston. 1994). One
gavage study on young 45-50-gram rats showed behavioral and locomotor effects (Chen et al..
2002a). One study of mice showed no neurobehavioral effects immediately after exposure
ceased at PND 17, but the mice exhibited increased locomotion and total activity and decreased
rearing at PND 60 (Fredriksson et al.. 1993). Following i.p. dosing, 8-week-old male mice
showed effects on the righting reflex and balancing (Umezu et al.. 1997). and 6-week-old rats
showed effects on locomotor activity (Motohashi et al.. 1993).
Autism spectrum disorder: One case-control study examined the relationship between
autism spectrum disorder (ASD) for births in 1994 in the San Francisco Bay Area and estimates
of 19 hazardous air pollutant concentrations for the census tract of the birth residence (Windham
et al.. 2006). Risk estimates for the upper 3rd quartile and upper 4th quartile of
tetrachloroethylene exposure were OR: 1.31 (95% CI: 0.93-1.84) and OR: 1.11 (95%
CI: 0.78-1.59), respectively, with no suggestion of a linear concentration-response pattern. The
low level of exposure detail for individual subjects in the study does not provide sufficient
information either for or against an association between tetrachloroethylene and ASD. The
causes of autism are unknown, but environmental factors have been hypothesized (Grandiean
and Landrigan. 2006). Epidemiologic studies of analytical designs and with more sensitive
exposure-assessment approaches are needed to more clearly define any role of
tetrachloroethylene and other air pollutants.
4.9.1.1.3.4. Developmental immunotoxicity
Section 4.8.1.1.1 and Table 4-38 describe studies relating tetrachloroethylene to immune
response in children. The developing immune system is an area of potential susceptibility
(Dietert. 2008). although there are few published studies relating to immune response after
tetrachloroethylene exposure to either children or adults. The childhood studies examined a
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-417 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
relationship with tetrachloroethylene exposure and allergy, asthma, and infection—immunotoxic
outcomes not reported in any of the studies of adults. In addition, family members of children
diagnosed with leukemia from Woburn, MA, exhibited altered lymphocyte (CD3, CD4, CD8)
and CD4/CD8 ratios (Byers et al.. 1988). though this was a mixed exposure to other
contaminants in addition to tetrachloroethylene. Other immunological conditions have been
observed in adults, but these are distinct from those observed in children discussed below. This
is an area for future research.
Allergy. Lehmann et al. (2002) examined cord blood samples from healthy, full-term
neonates for T-cell populations and associated them with indoor exposure to VOCs measured 4
weeks after birth (likely to reflect late-prenatal exposures) and observed a significant association
of tetrachloroethylene exposure with a reduction of interferon-g-producing Type 1 T-cells.
However, another study examining indoor exposure to VOCs and allergic sensitization and
cytokine secretion in 3-year-old children at high risk for development of allergic disease (low
birth weight, high cord blood IgE, family history of atopy) found no significant association
between tetrachloroethylene exposure and allergic sensitization to egg white and milk (Lehmann
et al.. 2001). No studies of allergy after exposure to tetrachloroethylene were reported in adults.
However, tetrachloroethylene has been demonstrated to adversely affect IL-4 and TNF-a in
rodent mast cells (Seo et al., 2008a) and passive cutaneous anaphylaxis in rats exposed i.p.
(Seo et al., 2008a) and in drinking water (Seo et al., 2008b).
Asthma. In a study of inhalation exposure, Delfino et al. (2003a; 2003b)measured the
concentration of ambient air pollutants, including tetrachloroethylene, and correlated it with
subsequent symptoms of asthma in children in the Los Angeles, CA area. These results
suggested an increased risk with exposure to tetrachloroethylene (Delfino et al.. 2003a).
However, another analysis of the data examined the amount of tetrachloroethylene and other
volatile organic compounds in exhaled breath of asthmatic children (Delfino et al.. 2003b).
Although there was a significant correlation between ambient and exhaled concentrations, the
investigators did not find any association with exhalation concentrations and asthma symptoms
or ambient air concentrations and asthma symptoms, although the OR for exhaled breath was
larger than for ambient air exposure (OR: 1.94. 95% CI: 0.8-4.7; Delfino et al.. 2003b). An
18-year-old without personal or family history of bronchial asthma developed respiratory
symptoms (cough, dyspnea, altered forced expiratory volume) after maintaining dry-cleaning
machines (Boulet. 1988).
Susceptibility to Infection. Only one report on tetrachloroethylene exposure and
childhood infection was found in the published literature. Higher prevalences of kidney and
urinary tract disorders (primarily infection) and lung and respiratory disorders (asthma, chronic
bronchitis, or pneumonia) in children were reported by mothers living in a community with a
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-418 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
past history of VOC-contaminated drinking water compared to prevalences reported by mothers
living in uncontaminated areas (Lagakos et al.. 1986).
4.9.1.1.3.5. Hepatotoxicity
Bagnell and Ellenberger (1977) reported that a child suffered from obstructive jaundice
and hepatomegaly after consuming tetrachloroethylene-contaminated breast milk (10 mg/L),
with conditions improving when breastfeeding was discontinued.
4.9.1.1.3.6. Fatality
A case report found that vapors off-gassing from dry-cleaned fabrics were implicated in
causing the death of a 2-year-old boy who had slept in a room with multiple curtains that had
been incorrectly dry cleaned (Gaillard et al.. 1995) and retained 6 kg of tetrachloroethylene as
estimated by a later experiment repeating the conditions (Gamier et al.. 1996). Another case
reported a death in a 17-year-old employed at a plastics manufacturing plant and using
tetrachloroethylene to clean the inside of a metal mold (NIOSH. 1994).
In the one case of a child's direct ingestion of tetrachloroethylene, a 6-year-old boy who
swallowed 12-16 g tetrachloroethylene lost consciousness and lapsed into a coma (Koppel et al..
1985). This 6-year-old also experienced drowsiness, vertigo, agitation, and hallucinations, but
he later recovered. Follow-up testing on the boy was not reported; therefore, any potential long-
term effects of the exposure are unknown (see Section 2.2.5). Due to the rarity of these cases,
there are little data to support any hypothesis regarding increased susceptibility for acute
mortality in childhood compared to adulthood.
4.9.1.1.3.7. Childhood cancer
The epidemiologic and experimental animal evidence is limited regarding susceptibility
to cancer from exposure to tetrachloroethylene during early life-stages. Generally speaking,
there may be developmental susceptibility for early lifestage exposure to chemicals and cancer
(Andersen et al., 2000)(Qlshan et al.. 2000). The human epidemiological evidence is
summarized above for cancer in the liver (see Section 4.4.1.2), kidney (see Section 4.5.1.2), and
other organ systems (see Section 4.8.1.2). The experimental animal research is summarized
above for cancer in the liver (see Section 4.4.2.2), kidney (see Section 4.5.2.2), and other organ
systems (see Section 4.8.2). Few studies have examined cancer in children after exposure to
tetrachloroethylene; those few have examined total childhood cancer, leukemia, and brain
tumors. A recent review of the data related to exposure to tetrachloroethylene, trichloroethylene,
or solvent mixtures found inadequate/insufficient evidence to determine whether an association
exists for childhood leukemia, neuroblastoma, or brain cancer (NRC. 2009).
Total Childhood Cancer. One study examined childhood cancers in an area in Endicott,
NY, for which vapor intrusion into homes was of concern. Many VOCs were identified in
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-419 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
samples and included trichloroethylene and tetrachloroethylene (ATSDR. 2006). This study
found fewer than six cases of cancer over a 20-year period, in children up to 19 years of age,
which did not exceed expected cases or types.
Childhood Leukemia. Leukemia has been observed in a few studies after exposure to
tetrachloroethylene in adults and children. However, the studies are limited by small sample
sizes, lack of exposure measurements, exposure to multiple contaminants, and possible
participation bias.
A small case-control study of children residing in Woburn, MA, found a strong but
imprecise association between maternal exposure during pregnancy and drinking water
contaminated with multiple solvents including tetrachloroethylene and childhood leukemia, with
a positive dose-response trend, when compared with exposure prior to pregnancy or postnatal
exposure to the infant via lactation (Costas et al.. 2002; MDPH. 1997; see Section 4.9.1.2.4).
However, it is difficult to uniquely identify tetrachloroethylene as the causative agent given the
higher concentrations of trichloroethylene reported. Other population case-control studies of
childhood leukemia have not shown an increased risk from paternal (Lowengart et al.. 1987; Shu
et al.. 1999) or maternal (Infante-Rivard et al.. 2005; Shu et al.. 1999) occupational exposure to
tetrachloroethylene, possibly due to the relatively small sample size. Another study population is
currently being further examined to determine any association between maternal ingestion of
contaminated water and the incidence of childhood cancers (ATSDR. 2003). One in vitro study
of human mononuclear cord blood cells exposed to tetrachloroethylene found that pathways
involved in cancer induction were affected through altered gene expression of inflammatory
responses, tumor and metastatic progression, and the apoptotic process (Diodovich et al.. 2005).
In addition, a follow-up study of children from Camp Lejeune, NC, is currently being conducted
to determine any association between maternal ingestion of contaminated water and the
incidence of childhood leukemia and non-Hodgkin lymphoma (ATSDR. 2003; NRC. 2009). No
data are available on cancer risk in animals from early lifestage tetrachloroethylene exposure.
Childhood Brain Cancer. Very few studies of tetrachloroethylene exposure have
reported brain tumors, and these are generally quite limited. One study of parental occupational
exposure to tetrachloroethylene (8 cases, 11 controls) found no risk of neuroblastoma in the
offspring (OR: 0.5, 95% CI: 0.2-1.4) (De Roos et al.. 2001). This study, like those on childhood
leukemia, is quite limited for examining parental exposure to tetrachloroethylene and childhood
cancer.
4.9.1.1.3.8. Age-dependent adjustment factors (ADAFs)
According to 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
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-420 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
exposures for carcinogens with a mutagenic MOA. Although the contribution of genotoxicity to
tetrachloroethylene carcinogenesis cannot be ruled out for one or more target organs,
uncertainties with regard to the characterization of tetrachloroethylene genotoxicity remain. This
is primarily because in vivo tests of tetrachloroethylene have been equivocal, with at most,
modest evidence of genotoxic effects in rodent tumor tissues examined (including mouse liver
and rat kidney) following exposure at tumorigenic doses. Additionally, no evidence is available
regarding the potential contribution of tetrachloroethylene genotoxicity to other rodent tumor
types (particularly, MCL, testes, and brain) or to human cancers. Ames assays of
tetrachloroethylene have yielded largely negative results. Certain tetrachloroethylene
metabolites (TCVG, TCVC, NAcTCVC, tetrachloroethylene oxide, and DCA) exhibit
genotoxicity, the database of available studies is limited, and not all metabolites have been
sufficiently tested to support clear conclusions about their genotoxic potential. Additionally, the
specific active moiety(ies) that contribute to tetrachloroethylene carcinogenesis are not known.
Thus, because the specific active moiety(ies), mechanisms, or modes of action by which
tetrachloroethylene induces carcinogenesis are not known, early-life susceptibility is not
assumed, and the application of ADAFs is not recommended.
4.9.1.1.3.9. Early lifestage exposure and outcomes in adulthood
Many additional studies have described adverse outcomes in adults only, mainly based on
the assumption of exposure occurring in adulthood; whether or not early lifestage exposures
might have occurred is often not considered. Only one identified study reports an examination of
early lifestage exposure to tetrachloroethylene and latent outcomes in adults. A large prospective
study of the offspring of dry cleaners found a significant increased risk for schizophrenia at
21-33 years of age (Perrin et al.. 2007). This is a preliminary report that did not adjust for
family history of mental disease, a risk factor for schizophrenia.
4.9.1.2. Later Life-Stages
Due to changes in physiology, in the elderly, exposure levels may be distinct from those
observed in younger adults. The elderly have increased ventilation rates per kg body weight
compared to adults (U.S. EPA. 2006a) and spend the majority of their time indoors, where
increased concentrations of tetrachloroethylene have been found compared to those measured
outdoors (U.S. EPA. 2001b). The elderly also experience changes in skin permeability (U.S.
EPA. 2006a). which may lead to increased exposure while showering, bathing, or swimming in
contaminated water (Rao and Brown. 1993; U.S. EPA. 2001b). While dermal exposure is
generally not considered a major route of exposure, this route of exposure is not well
characterized for later life-stages.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-421 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicokinetics in later lifestages can be distinct in younger adults (Ginsberg et al..
2005){U.S. EPA, 2006, 1945671 (Benedetti et al.. 20071 although there is only limited evidence
showing a possible age-related difference in CYP expression {Parkinson, 2004, 729573 }(Dome
and Renwick. 2005; George et al.. 1995). 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).
Few studies examined the exposure to tetrachloroethylene in elderly adults (>65 years
old). One study found elevated blood tetrachloroethylene levels (310-1,770 |ig/L) and urine
trichloroacetic acid levels (22-1,650 (J,g/L) in an elderly couple living above a dry-cleaning
facility (Popp et al.. 1992).
Similarly, few studies examine the effects of tetrachloroethylene exposure in elderly
adults. Another residential study examined two individuals over the age of 60 years and found
that the mean scores of VCS were lower than the 12th percentile of all control subjects (Schreiber
et al.. 2002).
One PBPK modeled tetrachloroethylene in adults aged 65, 75, and 85 years old and
predicted lower concentrations in all compartments for older adults compared to younger adults,
and similar predictions for TCA in older and younger adults (Yoklev and Evans. 2007). The
authors noted that these results indicate that increased susceptibility is likely among older adults
due to metabolic changes associated with aging. Another model predicted a decrease in alveolar
concentration of tetrachloroethylene in 65-year-olds versus 25-year-olds, which the authors
attribute to age-related decreases in cardiac output and ventilation (Guberan and Fernandez.
1974).
These very limited studies suggest that older adults may experience increased exposure to
tetrachloroethylene and resulting increased VCS deficits compared to younger adults. However,
there is no further evidence of effects for older adults exposed to tetrachloroethylene beyond
these studies.
4.9.2. Other Susceptibility Factors
Aside from age, many other factors may affect susceptibility to tetrachloroethylene
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
tetrachloroethylene toxicity.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-422 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
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, ventilation rate, and metabolic enzyme expression, which can
influence the toxicokinetics of chemicals and their metabolites in the body (Parkinson et al.,
2004)(Gandhi et al.. 2004; Gochfeld. 2007). In the case of tetrachloroethylene, there is some
indication that tetrachloroethylene metabolism is different between males and females. One
PBPK model found gender-specific differences that were small (although significant) in
tetrachloroethylene blood concentrations but considerable (twofold at age 40) with regard to
TCA blood concentration levels (Clewell et al.. 2004; see Section 3.5.2 and Figure 3 3). Opdam
and Smolders (1986) exposed six human subjects to concentrations ranging from 0.5-9 ppm and
found alveolar concentrations in male subjects to be only slightly less than those in females (see
Figures 3-6a, b). It is not known whether gender variation of P-lyase activity (see
Section 3.3.3.2.3), the most important activator of toxic products in the conjugation pathway,
exists in humans as it does in rats, with metabolism in males being faster than in females (Volkel
et al.. 1998). although there seems to be little gender difference in the concentrations of
metabolites in blood, regardless of age (Sarangapani et al.. 2003).
In humans, there have been a few studies demonstrating sex-specific effects (see
Section 4.7.2.3), but it has not been determined whether there is a gender difference in response
to exposure to tetrachloroethylene. Among former residents of Camp Lejeune, NC, exposed to
contaminated drinking water, there is limited/suggestive evidence of an association between
breast cancer and tetrachloroethylene, and inadequate/insufficient evidence to determine whether
an association exists for cervical, ovarian/uterine, or prostate cancer (NRC. 2009). Male breast
cancer has also been reported by former residents of Camp Lejeune exposed to contaminated
drinking water; however, this association has not been investigated sufficiently to draw any
conclusions (NRC. 2009).
Ferroni et al. (1992) evaluated neurological effects of tetrachloroethylene exposure
among female dry cleaners and concluded that tetrachloroethylene exposure in dry-cleaning
shops may impair neurobehavioral performance and affect pituitary function. The pituitary is
controlled, in part, by hypothalamic dopamine, which is important to neurotransmission. Study
participants were tested during the proliferation phase of menstruation, which may better capture
changes in prolactin secretion but also may potentially confound findings if there are individual
differences in severity of menstruation and in the timing of a test session relative to the day of
menstruation (U.S. EPA. 2004; see Section 4.6.1.2.5).
Some studies have observed an increased risk for NHL, Hodgkin lymphoma, chronic
lymphocytic leukemia or multiple myeloma in females compared to males (Cohn et al.,
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-423 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
2005)(Andersen et al.. 1999; Blair et al.. 2003; Ji and Hemminki. 2005b. 2006; Miligi et al..
2006; Morton and Marianovie. 1984; see Section 4.6.1.2; Radican et al.. 2008; Spirtas et al..
1991; see Section 4.6.1.2). whereas other studies observed an increase in both males and females
(Travier et al.. 2002) or no increase in either males or females (Boice et al.. 1999; Lynge et al..
2006). Other studies did not examine the outcome in both sexes. Some of these studies are
limited by lack of quantitative exposure information, ecological design, or exposure to mixtures,
differences in exposure potential and level of exposure may explain the difference in risk
between women and men. Differences in physiological parameters may also explain the
observed gender difference in risk.
The studies by Pesch et al.(2000b) and Dosemeci et al. (1999) suggest that there may be
gender differences in risk to renal cell carcinoma with occupational exposure to
tetrachloroethylene; in both studies, the risks were higher in males than in females (see Section
4.5.1.2). In a rat inhalation study, tubule cell hyperplasia was observed in eight males at various
doses but in only one female at the high dose. Also, renal tubule adenomas and
adenocarcinomas were observed only in males; however, chronically induced tetrachloroethylene
neoplastic kidney lesions do not exhibit sex specificity (NTP. 1986b). In a rat gavage study,
there was no gender difference for toxic nephropathy (NCI. 1977). A marked gender difference
was seen between male and female rats in the severity of acute renal toxicity, with male rats
being more affected than female rats (Lash et al.. 2002). but otherwise, no gender variation was
observed for chronic nephrotoxicity not associated with a2[j.-globulin nephropathy (see
Sections 4.5.2.2 and 4.5.4.3.3).
In the liver, male rats showed an increased incidence of spongiosis hepatitis as compared
with females, but there was no gender difference in hepatocellular adenomas and carcinomas;
however, the spleen showed increased effects in males versus females (JISA. 1993; see Sections
4.4.2.1 and 4.4.2.2).
4.9.2.2. Race/Ethnicity
Race/ethnicity can often be seen as an important consideration, and may be due to actual
increased exposure or to variation in expression of metabolic enzymes due to genetic variability
(Garte et al.. 2001). In particular, ethnic variability in expression has been reported for CYP
(Parkinson et al., 2004)(Dorne and Ren wick. 2005; McCarver et al.. 1998; Neafsev et al.. 2009;
Shimada et al.. 1994; Stephens et al.. 1994) and GST (Ginsberg et al.. 2009; Nelson et al.. 1995).
Studies of VCS in residents in apartments colocated with dry cleaners in New York, NY,
found that participants of minority status and low income (<$60,000) were more likely to have
high indoor air levels of tetrachloroethylene (>100 |ig/m3), but analyses of this small sample size
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-424 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
of participants in this exposure category could not definitively separate minority status from
VCS performance (NYSDOH. 2010; Storm et al.. In Press).
Oxidative damage among female dry cleaners appeared to be increased among black
workers compared to female Caucasian workers, although female dry cleaners had decreased
levels of oxidative damage compared to female launderers (Toraason et al.. 2003). In a follow-
up study on the mortality of a cohort of dry cleaners, bladder cancer was elevated among
Caucasian men and women, and kidney cancer was elevated among black men and women;
however, these associations were not strongly related to duration or estimated level of exposure
to tetrachloroethylene (Blair et al.. 2003). One study found that following tetrachloroethylene
exposure, TCA concentration in the urine of six Asian subjects was no different from the levels
found in six Caucasians; however, this study was confounded by significant differences in
alcohol consumption between the Caucasian and Asian populations (Jang and Droz. 1997).
4.9.2.3. Genetics
Human variation in response to tetrachloroethylene exposure may be associated with
genetic variation. For example, in a study of six adults, Monster et al. (1979) found that the
mean coefficient of interindividual variation for tetrachloroethylene uptake was 17%. Human
genetic polymorphisms in metabolizing enzymes involved in biotransformation of
tetrachloroethylene are known to exist (IARC. 1995; Lash et al.. 2001; U.S. EPA. 1991a).
Section 3.3.3.1.5 discusses CYP isoforms and genetic polymorphisms, Section 3.3.3.2.1 covers
GST isoenzymes and polymorphisms, and Section 3.3.4 describes differences in enzymatic
activity.
Reitz et al. (1996) examined tetrachloroethylene metabolism in seven adult human liver
samples and found a fivefold difference in the rate of tetrachloroethylene metabolism between
the 50th and 99th percentiles. Opdam (1989a) found a twofold spread in tetrachloroethylene
blood concentrations in a study population of nine adult human subjects. In this study, the
amount of fat and the blood concentrations seemed to be positively correlated but could not be
confirmed; the author suggested that if the subjects had a wider range of body fat levels (range in
this study was only 7-22 kg), a larger amount of interindividual variation would be expected.
Computer modeling was used to examine the toxicokinetic variability of
tetrachloroethylene (Bois et al.. 1996; Chiu and Bois. 2006). However, whether CYP or GSH
polymorphisms account for interindividual variation in tetrachloroethylene metabolism among
humans, and, thus, differences in susceptibility to tetrachloroethylene-induced toxicities, is not
known.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-425 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
4.9.2.4. Preexisting Disease
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 tetrachloroethylene and its
metabolites. There are limited data indicating that certain diseases may alter susceptibility to
tetrachloroethylene exposure, mainly through altered metabolism. Presence of cancer likely
alters tetrachloroethylene metabolism, because increased CYP2E1 expression has been observed
in these individuals (Neafsev et al.. 2009). Cirrhosis of the liver likely alters tetrachloroethylene
metabolism, because increased CYP2E1 expression has been observed in these individuals
(Neafsev et al.. 2009; also see Section 4.9.2.5.1). Tetrachloroethylene is lipophilic and stored in
adipose tissue (Monster and Houtkooper. 1979); therefore, obese individuals may experience
altered toxicokinetics of tetrachloroethylene compared to nonobese individuals. Obesity also
likely alters tetrachloroethylene metabolism, because increased CYP2E1 expression has been
observed in obese individuals, compared to nonobese individuals (McCarver et al.. 1998;
Neafsev et al.. 2009). For obese individuals, a model predicted a decrease in alveolar
concentration of tetrachloroethylene during exposure and a decrease in elimination, compared to
nonobese individuals (Guberan and Fernandez. 1974).
4.9.2.5. Lifestyle Factors and Nutrition Status
4.9.2.5.1. Alcohol intake
Alcohol is generally regarded as a confounder, although the additive or interactive effects
of these exposures along with tetrachloroethylene are not well characterized. Alcohol intake
likely alters tetrachloroethylene metabolism and causes higher toxicity, because increased
CYP2E1 expression has been observed in individuals who consume alcohol, compared to those
who do not (Parkinson et al., 2004)(Liangpunsakul et al.. 2005; Lieber. 1997; McCarver et al..
1998; Meskar et al.. 2001; Neafsev et al.. 2009; Perrot et al.. 1989). Those exposed to both
tetrachloroethylene and TCE and consumed alcohol demonstrated an elevated color confusion
index (Valic et al.. 1997).
4.9.2.5.2. Tobacco smoking
Smoking, or the number of factors correlated to smoking (e.g., socioeconomic status,
diet, alcohol consumption), is generally regarded as a confounder in epidemiological studies
(Ruder. 2006). although the additive or interactive effects of these exposures along with
tetrachloroethylene are not well characterized. Immunotoxicity and hematotoxicity were
observed in tetrachloroethylene-exposed dry cleaners, particularly for those who were smokers
(Emara et al.. 2010). Sister chromatid exchange in peripheral lymphocytes was observed more
frequently in male smokers exposed to tetrachloroethylene alone or in combination with TCE
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-426 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
(Seiii et al.. 1990). No increase in oxidative damage among tetrachloroethylene-exposed dry
cleaners was observed among smokers compared to nonsmokers (Toraason et al.. 2003).
Regarding esophageal cancer, occupational observations suggest that the magnitude of the risks
for several smoking-related cancers among dry cleaners was greater than could be explained by
smoking alone, suggesting a further contribution from another risk factor, such as occupational
exposure (Blair et al.. 2003; Ruder et al.. 2001; see Section 4.8.1.2.2).
4.9.2.5.3. Nutritional status
Vegetable or vitamin intake may decrease susceptibility to tetrachloroethylene because
CYP2E1 inhibition has been observed in individuals who consume various vegetables, herbs,
and teas, and increased expression in those consuming high-fat diets (Neafsev et al.. 2009).
Coexposure to a-tocopherol (vitamin E) along with tetrachloroethylene resulted in decreased rat
(Costa et al.. 2004) and mouse (Ebrahim et al.. 1996; Ebrahim et al.. 2001) liver cell toxicity. A
similar protective effect was also seen with coexposure to 2-deoxy-D-glucose in mice (Ebrahim
et al.. 1996; Ebrahim et al.. 2001) and taurine in mice (Ebrahim et al.. 2001). An in vitro study
of cultured normal human epidermal keratinocytes demonstrated an increase in lipid
peroxidation in a dose-dependent manner after exposure to tetrachloroethylene, which was then
attenuated by exposure to vitamin E (Ding et al., 2006). However, no associations were found
for blood levels of vitamin E and P-carotene in rats (Toraason et al.. 2003; see Sections 4.3 and
4.4.4.4.3)
4.9.2.5.4. Physical activity
Studies and models have examined the effect of increased workloads on the
toxicokinetics of inhaled tetrachloroethylene alone (Droz et al.. 1989a; Droz et al.. 1989b;
Imbriani et al.. 1988; Jakubowski and Wieczorek. 1988; Pezzagno et al.. 1988) or with TCE
(Opdam. 1989a. b). These studies are equivocal on whether an increase in pulmonary ventilation
increases the amount of tetrachloroethylene taken up during exposure. A model predicted an
increase in alveolar concentration of tetrachloroethylene after exercise, which the authors
attribute to increased cardiac output and ventilation (Guberan and Fernandez. 1974).
4.9.2.6. Socioeconomic Status
Socioeconomic status (SES) can be an indicator for a number of coexposures, 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 tetrachloroethylene exposure.
Children's exposure to tetrachloroethylene 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
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-427 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
examine health effects. Studies of VCS measured in child and adult residents in apartments
colocated with dry cleaners in New York, NY, found that the study participants more likely to be
exposed to high indoor air levels of tetrachloroethylene (>100 |ig/m3) were of minority status,
low income (<$60,000), or, for adults, had significantly lower level of education (NYSDOH.
2005a. 2010; Storm et al.. In Press). However, analyses of the small sample size in this exposure
category could not definitively separate race/ethnicity or SES from VCS performance.
4.9.2.7. Multiple Exposures and Cumulative Risks
When considering health risks, it is important to consider the cumulative impact of
effects that may be due to multiple routes of exposure. EPA published a Framework for
Cumulative Risk Assessment (U.S. EPA. 2003) to address these issues. A human aggregate
exposure model developed by McKone and Daniels (1991) incorporated likely exposures from
air, water, and soil media through inhalation, ingestion, and dermal contact. They asserted that
the aggregate exposure may be age dependent but did not present any data for persons of
differing life-stages.
The limited data summarized by the ATSDR in its draft interaction profile on
tetrachloroethylene, trichloroethylene, 1,1-dichloroethane, and 1,1,1-trichloroethane suggest that
additive joint action is plausible (ATSDR, 2001). Coexposure to other pollutants, including
trichloroethylene and methylchloroform, which produce some of the same metabolites and
similar health effects as tetrachloroethylene, is likely to occur in occupational settings as well as
in nonoccupational sources such as in ground water contamination (e.g., ATSDR. 1998a; Bove et
al.. 2002; Lagakos et al.. 1986; MDPH. 1997; Sonnenfeld et al.. 2001). However, no evidence
from the available studies indicates greater-than-additive effects for liver and kidney toxicity.
Numerous environmental pollutants and therapeutic agents have the potential to induce or
inhibit tetrachloroethylene-metabolizing enzymes. For example, tetrachloroethylene metabolism
is increased by inducers of CYP enzymes such as toluene, phenobarbital, and pregnenolone-
16-a-carbonitrile, whereas CYP inhibitors such as SKF 525A, metyrapone, and carbon monoxide
decrease tetrachloroethylene metabolism (Moslen et al., 1977)(Costa and Ivanetich. 1980; Ikeda
and Imamura. 1973). Likewise, tetrachloroethylene exposure may increase the effects of
exposures to other chemicals or stressors. For instance, adverse effects due to exposure to
chlorinated solvents and alcohol may be increased because tetrachloroethylene may induce
shared metabolic enzymes (see Section 3.3.4).
The acute effects of tetrachloroethylene share much in common functionally with those
of other solvents (e.g., toluene, volatile anesthetics, and alcohols) such as changes in reaction
time, nerve conduction velocity, and sensory deficits. There is emerging evidence that such
agents act on the ligand-gated ion channel superfamily in vitro (Shafer et al.. 2005). particularly
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-428 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
on the inhibitory amino acids NMD A, nicotinic, and GABA receptors in vivo (Bale et al.. 2005).
Other organic solvents induce effects on memory and color vision (Altmann et al.. 1995; Hudnell
et al.. 1996a; Hudnell et al.. 1996b; Mergler et al.. 1991). The consistency of these observations
suggests a common MOA of organic solvents to altered vision pattern. Hence, a concern exists
for neurobehavioral effects from interaction or competitive inhibition between
tetrachloroethylene and exposures with similarly hypothesized MO As.
The interaction between tetrachloroethylene, trichloroethylene, and 1,1,1-trichloroethane
(methylchloroform) was modeled in rats (Dobrev et al.. 2001) and in computer models for
humans (Dobrev et al.. 2002) and was shown to compete for metabolic capacity. The interaction
between tetrachloroethylene and trichloroethylene showed a less-than-additive effect on the liver
and kidney through inhibition of TCA formation (Pohl et al.. 2003). Similarly, when exposed to
tetrachloroethylene, rat liver cells had increased toxicity when coexposed to peroxidation drugs
such as cyclosporine A, valproic acid, and amiodarone (Costa et al.. 2004). and //-hexane and
ethylbenzene inhibited the metabolism of tetrachloroethylene in rats (Skowron et al.. 2001).
4.9.3. Uncertainty of Database and Research Needs for Susceptible Populations
There is some evidence that certain populations may be more susceptible to exposure to
tetrachloroethylene. The factors examined for tetrachloroethylene include age, gender,
race/ethnicity, genetics, preexisting disease, lifestyle factors, nutritional status, socioeconomic
status, and multiple exposures and cumulative risk. Areas where the database is currently
insufficient for characterizing the impact of tetrachloroethylene on susceptible populations are
identified below, along with research needs.
There is limited information on early life exposure to tetrachloroethylene than on other
potentially susceptible populations, there remain a number of uncertainties regarding childhood
susceptibility. Although inhalation is believed to be of most concern for tetrachloroethylene,
pathways of exposure for children are not well characterized. It is not clear to what extent
tetrachloroethylene may pass through the placenta in humans, as shown in rodent studies
(Ghantous et al.. 1986; Szakmary et al.. 1997); for some infants, the primary route of exposure
may be through breast milk ingestion (see Sections 2.2.4 and 3.2), while for other infants, the
dose received through ingestion of breast milk will become insignificant when compared with
the inhalation exposure and subsequent dose (Schreiber. 1997). The amount of
tetrachloroethylene ingested from food is not well described; and it is not known to what extent
tetrachloroethylene is absorbed by a child and to which organs tetrachloroethylene and its
metabolites may be distributed. The neurological effects of tetrachloroethylene constitute the
most sensitive endpoints of concern for noncancer effects, and limited data show that early life-
stages may be more susceptible to visual deficits than are adults (NYSDOH. 2005a. 2010;
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-429 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Schreiber et al.. 2002; Storm et al.. In Press), yet developmental neurotoxic effects, particularly
in the developing fetus, need further evaluation using age-appropriate testing for assessment.
There are a number of adverse health effects observed uniquely in early lifestages, with no
comparable observations in adults to determine relative sensitivity (e.g., birth outcomes, autism,
allergy); conversely, there are some adverse outcomes that have been observed only in adults.
There is suggestive evidence that there may be greater susceptibility for exposures to the
elderly, but the available data are much more limited with related uncertainties. Improved PBPK
modeling that contains physiologic parameter information for infants and children (including, for
example, the effects of maternal inhalation exposure and the resulting concentration in breast
milk) and for older adults, and validation of these models, will aid in determining differences in
life-stage toxicokinetics of tetrachloroethylene. There may be a true difference in outcome after
exposure during one life-stage compared to another, a lack of assessment of these outcomes in all
life-stages, or a lack of assessment of effects of exposures in one life-stage and latent outcomes.
More studies specifically designed to evaluate effects in early and later life-stages are needed in
order to more fully characterize potential life-stage-related tetrachloroethylene toxicity.
For other susceptibility factors, the data are more limited and based mainly on
nonchemical specific data that provide information on variation in physiology, exposure, and
toxicokinetics. Until quantitative conclusions can be made for each susceptibility factor, it will
be very hard to consider the impacts of changes in multiple susceptibility factors. In addition,
further evaluation of the effects of aggregate exposure to tetrachloroethylene from multiple
routes and pathways is needed. Similarly, the effects due to coexposures to other compounds
with similar or different MO As need to be evaluated.
4 10 SUMMARY OF HAZARD IDENTIFICATION
4.10.1. Overview of Noncancer and Cancer Hazard
This section summarizes the noncancer and cancer hazard findings for
tetrachloroethylene. This summary is based on the analyses presented in the preceding sections,
which discussed tetrachloroethylene toxicity on an organ-specific basis, in the following order of
presentation: neurotoxicity (see Section 4.1); kidney and bladder toxicity and cancer (see Section
4.2); liver toxicity and cancer (see Section 4.3); esophageal cancer (see Section 4.4); lung and
respiratory cancer (see Section 4.5); immunotoxicity, hematologic toxicity, and cancers of the
immune system (see Section 4.6); and developmental and reproductive toxicity and reproductive
cancers (see Section 4.7). Section 4.8 discusses genotoxicity, and susceptible populations are
addressed in Section 4.9.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-430 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
The noncancer hazard characterization for tetrachloroethylene is presented in
Section 4.10.2. Findings in humans and in experimental animals within each toxicity domain
(i.e., neurotoxicity [see Section 4.10.2.1], kidney toxicity [see Section 4.10.2.2], liver toxicity
[see Section 4.10.2.3], immunotoxicity and hematologic toxicity [see Section 4.10.2.4], and
reproductive and developmental toxicity [see Section 4.10.2.5]) are first summarized. A tabular
summary of the inhalation (see Table 4-49) and oral (see Table 4-50) studies that are suitable for
dose-response analysis, considering all studies across toxicity domains, is then presented in
Section 4.10.2.6. Neurotoxicity is identified as a sensitive endpoint following either oral or
inhalation exposure to tetrachloroethylene. Section 5 presents dose-response analyses of the
neurotoxicity data set as a basis for derivation of inhalation and oral reference values.
Quantitative dose-response analyses of the findings in other toxicity domains (i.e., kidney, liver,
reproductive and developmental toxicity) are also presented in Section 5.
The cancer hazard characterization for tetrachloroethylene is presented in Section 4.10.3.
Section 4.10.3.1 presents the hazard descriptor, characterizing tetrachloroethylene as —tkely to
be carcinogenic to humans." Section 4.10.3.2 synthesizes the epidemiologic data pertaining to
tetrachloroethylene and several cancer types, including non-Hodgkin lymphoma, multiple
myeloma, bladder, esophageal, kidney, lung, cervical, and breast cancer. Section 4.10.3.3
summarizes the results from three chronic bioassays that identified tetrachloroethylene-induced
rodent cancer, including mononuclear cell leukemia, kidney, and brain tumors in rats and liver
tumors in mice. The available mode-of-action information for the carcinogenicity of
tetrachloroethylene is presented in Section 4.10.3.4. Section 5 presents dose-response analyses
of the rodent bioassay data as a basis for derivation of inhalation and oral cancer slope factors.
4.10.2. Characterization of Noncancer Effects
4.10.2.1. Neurotoxicity
Human and animal studies provide complementary evidence regarding the association of
neurobehavioral deficits and tetrachloroethylene exposure. Tetrachloroethylene exposure in
humans has primarily been shown to affect visual function (including color vision) and
visuospatial memory and other aspects of cognition. Brain-weight changes have been measured
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-431 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-49. Inhalation studies suitable for dose-response analyses
Organ/
system
Study
Species
Duration/dosing
NOAEL/LOAEL'
(ppm)
Effect
CNS
Schreiber et al.
(2002)
Human
4 yr mean duration
0.3 (daycare workers,
mean and median)
Visual contrast
sensitivity
Schreiber et al.
(2002)
Human
5.8 yr (mean),
continuous
0.1 (residents, median
and mean), maybe as
high as 0.4 (mean) and
0.3 (median)
Visual contrast
sensitivity0
NYSDOH (2010)
Human
10 yr mean
duration
0.002, CL05 (children)
0.002, O07 (adults)
Visual contrast
sensitivity
Cavalleri et al.
(1994); Gobbaet
al. (1998)
Human
8.8 yr mean
duration
7. Cavalleri et al
£1994)
Dyschromatopsia
Spinatonda et al.
(1997)
Human
Inhalation (no
information on
duration)
§ (median)
Reaction time
Seeber(1989)
Human
>10 yr mean
duration
12, 53
Visuospatial function,
information processing
speed
Ferroni et al.
(1992)
Human
10.6 yr mean
duration
Reaction time,
continuous performance
Echeverria et al.
(1995)
Human
15 (high-exposure
group) yr mean
duration
11 d (operators)
Visuospatial function
Altmann et al.
(1990)
Human
4-h exposure each
day for 4 d.
10, 5Q
Visual Evoked Potentials
Hake and Stewart
(1977)
Human
7.5-h exposure
each for 5 d.
20, 100, JJQ
EEGs
Kjellstrand et al.
(1985)
Mouse
Acute (1 h)
0, 90, 320, 400, 600,
800, 1,200, 1,800,
3,600
Increased motor activity
Rosengren et al.
(1986)
Gerbil
Subchronic (12 wk,
with 16-wk follow-
up) continuous
0, §g, 300
Brain: protein, DNA
concentration
Mattsson et al.
(1998)
Rat
Subchronic (13 wk)
6 h/d, 5 d/wk
0, 50, 200, 2QQ
Flash-evoked potential
Wang et al. (1993)
Rat
Subchronic (12 wk)
continuous
0, 300. 600
Reduced brain weight,
DNA, protein
Oshiro et al. (2008)
Rat
60 min
500. 1.000. 1.500
Reaction time
500, 1,000, 1,500
False alarms
500. 1.000. 1.500
Trial completions—
Signal Detection Task)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-432 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-49. Inhalation studies suitable for dose-response analyses
(continued)
Organ/
system
Study
Species
Duration/dosing
NOAFI/IOAFI'
(ppm)
Effect
CNS
(continued)
Boves et al. (2009)
Rat
90 minutes
250. 500, 1,000
Impairment in steady
state visual evoked
potential
120 minutes
1,000, 2,000, 3,000,
4,000
Impairment in steady
state visual evoked
potential
Kidney
Mutti et al. (1992)
Human
10 yr duration
15 (median)
Urine and serum markers
of nephrotoxicity
NTP (1986b)
Rat
Chronic bioassay
(104 wk)
0,2QQ, 400
Increased karyomegaly
(74%), megalonuclear-
cytosis
JISA (1993)
Rat
Chronic
(104 wk)
0, 50,2QQ, 600
Increased relative kidney
weight; karyomegaly in
proximal tubules
JISA (1993)
Mouse
Chronic
(104 wk)
0, 10, 50,25Q
Increased relative kidney
weight; karyomegaly in
proximal tubules
Liver
Kjellstrand et al.
(1984)
Mouse
Subchronic (4 wk)
continuous
0, £, 37, 75, 150
Increased liver weight
NTP (1986b)
Mouse
Chronic bioassay
(104 wk)
0, IM, 200
Increased liver
degeneration, necrosis
JISA (1993)
Mouse
Chronic
(104 wk)
0, 10,5Q, 250
Increased angiectasis
Immune and
hematologic
toxicity
Emara et al. (2010)
Human
Mean duration 7 yr
Mean exposure levels
<140 ppm; mean blood
levels 1,685 |ig/L
Reduced RBC count,
reduced hemoglobin,
increased WBC count,
increased lymphocytes,
increased IgE
Reproductive
and develop-
mental
toxicity
Sallmen et al.
(1995); Finland
Human
Exposure during
year before
initiation of
pregnancy,
occupational,
1973-1983
Mean concentration
for dry cleaners in
Nordic countries,
1964-1979 = 24 ppm
(from Lvnse et al..
2006)
Time to pregnancy
Eskenazi et al.
(1991b); United
States
Human
Wives of exposed
men working as
dry cleaners, 1980s
31 ppm average
concentration,
personal samples
(n = 208), any job title,
all sample durations
(Table II. Gold et al..
2008)
Time to conception
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-433 DRAFT-DO NOT CITE OR QUOTE
-------
Organ/
system
Study
Species
Duration/dosing
NOAII/IOAII'
(ppm)
Effect
Reproductive
and develop-
mental
toxicity
Olsenetal.,
(1990); Kwronen
et al. (1989),;
Finland
Human
1st trimester,
occupational,
1973-1983
4.9 ppme
Spontaneous abortion
(continued)
Nelson et al.
(1980)
Rat
7 h/d on GDs 7-13
or 14-20
0, 100. 900
Decreased weight gain in
offspring;
CNS: behavior, brain
acetylcholine
Beliles et al. (1980)
Mouse
5 d exposure;
1, 4, and 10 wk
follow-up
0, 100. 500
Sperm quality
Tinston (1994)
Rat
Developmental—
multigeneration; 6
h/d, 5 d/wk
o, nx), 2m, 1,000
F2A pup deaths by Day
29; F1 andF2
generations: CNS
depression
Carney et al.
(2006)
Rat
Developmental—6
h/d on GDs 6-19
0, 65, 250, 600
Decreased fetal and
placental weight and
incomplete ossification
of thoracic vertebral
centra
"Experimental/observational NOAEL is underlined; LOAEL is double-underlined.
°Schreiber et al. (2002) found mean PCE concentrations of 0.2 ppm (0.09 ppm, median) of four families living in
apartments above active dry cleaning and two families living in an apartment building where dry cleaning had
ceased 1 mo earlier. Ambient monitoring of these six apartments during a period of active dry cleaning indicated
exposure to higher concentrations, mean: 0.4 ppm (median: 0.2 ppm).
d(Echeverria et al.. 1995)—the lowest exposure group is chosen to represent the LOAEL; (3 coefficient for lifetime
or chronic PCE exposure was positive and statistically significant for pattern memory, pattern recognition, and
pattern reproduction.
eLow group (working at dry cleaners but not operator or spot removal >1 h/d); Calculated from mean concentration
for dry cleaners 1964-1979 (24 ppm. Lvnge et al.. 2006) divided by ratio of exposure for operators versus other
work in dry cleaners. Chose a ratio of 5:1 as an intermediate level between 7:1 from Gold et al. (2008) (pg. 816)
that included transfer type machines in the United States and 3.5:1 from Raisanen et al., 2001 which included only
dry to dry primarily nonvented machines in Finland.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-434 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-50. NOAELs and LOAELs in selected studies involving oral
exposure to tetrachloroethylene
Organ/
System
Study
Species
Duration/exposure
Route
Dose/exposure
(NOAEL/LQAELf
(mg/kg-day)
Effect
CNS
Fredriksson et
al. (1993)
Mouse
PND 10-16/oral
gavage
0,5, 320
Day 60: Increased
locomotion,
decreased rearing
Kidney
NCI (1977)
Mouse, Rat
Chronic
(78 wk)/oral gavage
0, 536. 1,072 (male mice);
0, 386. 772 (female mice);
0, 475. 950 (male and female
rats)
Toxic
nephropathy
Liver, kidney
Jonker et al.
(1996)
Rat
4 wk/oral gavage
0, MM, 2,400
Liver weight,
enzyme levels;
kidney weight,
kidney enzyme
levels
Liver
Berman et al.
(1995)
Rat
14 d/oral gavage
0, 50, 150, 500, 1.500, 5,000
Liver weight,
ALT
Liver
Buben and
O'Flaherty
(1985)
Mouse
(40 g)
6 wk/oral gavage
0, 20,100, 200, 500, 1,000,
1,500, 2,000
Liver weight,
triglycerides
Hematologic
toxicity
Marth et al.
(1987; 1985b:
1989)
Mouse
(2 wk old,
20 g)
7 wk/drinking water
0, 0.05. 0.1 mg/kg-bw/day
Reversible
hemolytic anemia,
increased serum
triglycerides,
decreased
cholesterol
Development
Bove et al
(1995); United
States
Human
1st trimester,
drinking water,
1985-1988
<1,3.5, 7.5 and >10 ng/Lb
Oral clefts
"NOAELs are underlined once; LOAELs are double-underlined.
bBove et al. reported risks for categories of drinking water concentration of <1, >1-5, >5-10, and >10 |ig/L.
Exposure levels are the midpoints of these exposure categories. Supported by Aschengrau et al. (2009) who
observed an increased risk of oral clefts associated with any exposure to PCE versus no exposure (1-5, 197 |ig/L).
1
2 in animal studies. A more in-depth discussion of the human neurotoxicological studies can be
3 found in Section 4.1.1.3. The animal inhalation and oral or i.p. exposure studies are discussed in
4 Sections 4.1.2.1 and 4.1.2.2, respectively.
5 Visual contrast sensitivity deficits as well as color discrimination deficits are commonly
6 present prior to detectable pathology in the retina or optic nerve head. These deficits are, thus,
7 among the earliest signs of disease and potentially more sensitive measures than evoked
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-43 5 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
potentials from visual stimuli (Regan. 1989). Several independent lines of evidence can be
found in the occupational and residential exposure studies to support an inference of visual
deficits following chronic tetrachloroethylene exposure. The studies that observed effects on
color vision using the Lanthony D-15 color vision test include cross-sectional and longitudinal
designs in dry cleaning (Cavalleri et al.. 1994; Gobba et al.. 1998) and residential (Schreiber et
al.. 2002) settings. Decrements in color confusion were reported among 22 dry-cleaning workers
exposed to a mean TWA of 7 ppm for an average of 8.8 years (Cavalleri et al.. 1994). A
significant dose-response relationship between CCI value and tetrachloroethylene concentration
(p < 0.01) was also seen in Cavalleri et al. (1994). As noted previously, the color vision testing
in this study was blinded to exposure level of the study participants, and the study participants
were well matched in terms of age, smoking, and alcohol use. A follow-up of these workers
2 years later (Gobba et al.. 1998) showed greater loss in color discrimination in those who were
subsequently exposed to a higher concentration [increase in geometric mean from 1.7 to 4.3
ppm], with no change in those exposed to lower concentrations [decrease in geometric mean
from 2.9 to 0.7 ppm]). Although Gobba et al. (1998) demonstrated persistent color confusion
effects in this follow up evaluation, the study exposures are not clearly characterized over the
course of the 2-year duration. Nakatsuka et al. (1992) did not observe an association with color
vision among dry cleaners in China (n = 64, geometric mean: TWA 11 and 15 ppm in females
and males, respectively), but the relative insensitivity of the specific type of color vision test
used in this study (Lanthony. 1978) is a likely explanation for these results. Effects on color
vision were also seen among 14 dry cleaners in the small study in Malaysia by Sharanjeet-Kaur
et al. (2004). but this study provides little weight to the strength of the evidence because of the
lack of exposure information (other than job title), and differences between dry cleaners and
controls regarding test conditions and smoking habits. Two other small studies also reported
lower scores on the Lanthony D-15 color vision test in much lower exposure settings, but the
differences were not statistically significant. A study of residents living above dry cleaners
(mean tetrachloroethylene exposure during active dry cleaning = 0.4 ppm), reported mean CCI
scores of 1.33 and 1.20 for 17 exposed and 17 controls, respectively (p = 0.26). A study of
workers in a daycare center located in a building with a dry-cleaning business (mean
tetrachloroethylene exposure: 0.32 ppm) reported mean CCI scores of 1.22 and 1.18 in the
exposed daycare workers and controls, respectively (p = 0.39) (Schreiber et al.. 2002). Another
residential exposure study observed decrements in color vision in children but not in adults
(NYSDOH. 2005a). Overall, the evidence reveals a high degree of consistency in this aspect of
visually mediated function.
Visual contrast sensitivity changes were reported in two NYSDOH residential studies. In
a small pilot study (4 children and 13 adults), mean scores for visual contrast sensitivity (using a
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-436 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
near vision visual contrast sensitivity test) across spatial frequencies were statistically
significantly lower in exposed residents than in controls, indicating poorer visual function in the
exposed groups (Schreiber et al.. 2002). Controls were age- and sex-matched to the exposed
group, and both groups were English speaking and of predominately Caucasian ethnicity;
however, they were drawn from different geographic areas. In addition, two of the four exposed
children had diagnoses of learning disabilities or developmental delays, which could affect
performance on this type of test. In the larger study (NYSDOH. 2005a. b, 2010). the test
(Functional Acuity Contrast Test, FACT) assessed far vision visual contrast sensitivity, and the
test had a low rate of detecting visual contrast changes. For both contrast vision and color
vision, a number of analyses in NYSDOH (2005a. 2010; Storm et al.. In Press) suggest a
vulnerability among children. However, exposure to >0.015 ppm (>100 |ig/m3)
tetrachloroethylene was highly correlated with race and children's age. Additionally, the sample
sizes in the highest exposure group, especially in higher income, nonminority groups, makes it
difficult to fully examine possible effects of income, race, and age on vision. Therefore, while
both studies report visual contrast sensitivity changes, with exposed children being more
sensitive, there are concerns with the methodological and analytic approaches in these studies.
Acute human exposure studies reported increased latencies of up to 3.0 ms in visual
evoked potentials (Altmann et al.. 1990) and changes in EEGs (magnitude of effect was not
specified) (Hake and Stewart. 1977; Stewart et al.. 1970) at higher exposures ranging from 340
to 680 mg/m3.
In rats, acute inhalation exposure to tetrachloroethylene results in significant changes to
the flash-evoked potential at 800 ppm (Mattsson et al.. 1998) and a decrease in F2 amplitudes of
the steady state visual-evoked potential at 250 ppm (Boves et al.. 2009). In a subchronic
exposure study (13 weeks, up to 800-ppm tetrachloroethylene), changes in flash-evoked potential
responses were not observed at tetrachloroethylene exposures up to 200 ppm. In the 800-ppm
group, there was a significant increase in the amplitude and a significant increase in latency
(-3.0 ms) of the mid-flash-evoked potential waveform (N3), but histopathological lesions were
not observed in the examination of the visual system brain structures (e.g.. visual cortex; optic
nerve; Mattsson et al.. 1998).
Effects on visuospatial memory in humans were also reported in each of the studies that
examined this measure (Altmann et al.. 1995; Echeverria et al.. 1994; Echeverria et al.. 1995;
Seeber. 1989). These effects (increased response times) were seen in occupational and
residential studies, and the occupational studies were quite large, involving 101, 65, and 173 dry-
cleaning workers in Seeber (1989). Echeverria et al. (1995). and Echeverria et al. (1994).
respectively. Several different types of tests were used including digit reproduction (Seeber.
1989). switching, pattern memory, and pattern recognition (Echeverria et al.. 1994; Echeverria et
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-437 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
al.. 1995). and the Benton test (Altmann et al.. 1995). Exposure ranges for the increased reaction
time observations (LOAELs) ranged from 4.99 to 102 mg/m3 (Altmann et al.. 1995; Echeverria
et al.. 1995; Ferroni et al.. 1992). The changes in the cognitive tasks were observed at exposures
(LOAELs) ranging from 53.9 to 364.22 mg/m3 (Echeverria et al.. 1995; Seeber. 1989;
Spinatonda et al.. 1997). All of these studies except Altmann et al. (1995) indicate that the
neurobehavioral assessment was blinded to knowledge of the exposure level of the subject, and
all of the studies adjusted for potentially confounding factors. It should be noted, however, that
residual confounding from education-level differences between exposed and referent subjects
may still be present in Altmann et al. (1995).
Increased reaction time, increased number of false alarms, and decreased trial
completions in a signal detection task (measures of decreased attention) were reported in an
acute (60 minutes) exposure (6,782 mg/m3 or higher) study in rats (Oshiro et al.. 2008).
Additionally, operant tasks that test cognitive performance have demonstrated performance
deficits in rats and mice following acute tetrachloroethylene oral (Warren et al.. 1996) and i.p.
(Umezu et al.. 1997) exposures. These findings are consistent with observed effects on cognition
and memory in humans. However, no studies, to date, have evaluated the persistent effects of
tetrachloroethylene exposure on cognitive performance deficits in animal models.
An occupational exposure study (n=60) (Ferroni et al.. 1992) and a residential exposure
study (n = 14) (Altmann et al.. 1995). with mean exposure levels of 15 and 0.7 ppm,
respectively, reported significant increases in simple reaction time of 24 ms (11% increase)
(Ferroni et al.. 1992) and 40 and 51.1 ms (15 and 20% increases', respectively', for two separate
measurements) (Altmann etal.. 1995) for the exposed subjects. A third study, Lauwerys et al.
(1983). reported better performance on simple reaction time in 21 exposed workers (mean TWA:
21 ppm) compared with controls when measured before a work shift but not when measured
after work.
The changes in brain weight, DNA/RNA, and neurotransmitter levels that were observed
in the animal studies are highly supportive of the neurobehavioral changes observed with
tetrachloroethylene exposure. Changes in brain DNA, RNA, or protein levels and lipid
composition were altered following inhalation, with changes observed in cerebellum,
hippocampus, and frontal cortex (Rosengren et al.. 1986; Savolainen et al.. 1977a; Savolainen et
al.. 1977b; Wang et al.. 1993). The replication of these changes in biochemical parameters and
effects in brain weight in both rats and gerbils is pathognomonic. Changes in neurotransmitters
systems (Briving et al.. 1986; Honma et al.. 1980a; Honmaetal.. 1980b) and circadian rhythm
(Motohashi et al.. 1993) in animal studies are consistent with neuroendocrine alterations
observed in humans (Ferroni et al.. 1992).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-43 8 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
In conclusion, the weight of evidence across the available studies of humans and animals
exposed to tetrachloroethylene indicates that chronic exposure to tetrachloroethylene can result
in decrements in color vision, visuospatial memory, and possibly other aspects of cognition and
neuropsychological function, including reaction time.
4.10.2.2. Kidney Toxicity
The epidemiologic studies support association between inhalation tetrachloroethylene
exposure and chronic kidney disease, as measured by urinary excretion of renal proteins and
ESRD. The elevated urinary RBP levels seen in two studies (Mutti et al.. 1992; Verplanke et al..
1999) and lysozyme or P-glucuronidase in Franchini et al. (1983) provide some evidence for
effects to the proximal tubules from tetrachloroethylene exposure. Exposures in the studies that
observed renal toxicity were time-weighted averages of 8, 10, and 15 ppm. None of the
reviewed studies reported exposure-response relationships, and this is an important limitation of
the available data. Calvert et al. (2010) supports an association between inhalation
tetrachloroethylene exposure and ESRD, particularly hypertensive ESRD, and observed a
twofold elevated incidence among subjects who worked only in a shop where tetrachloroethylene
was the primary cleaning solvent compared to that expected based on U.S. population rates. An
exposure-response pattern was further suggested because hypertensive ESRD risk was highest
among those with longest employment durations. No human studies investigating drinking water
or other oral tetrachloroethylene exposures on kidney toxicity have been published.
Adverse effects on the kidney have been observed in studies of rodents exposed to high
concentrations of tetrachloroethylene by inhalation (JISA. 1993; NTP. 1986b) , oral gavage
(Ebrahim et al.. 1996; Goldsworthy et al.. 1988; Green et al.. 1990; Jonker et al.. 1996; NCI.
1977)(Ebrahim et al., 2002), and i.p. injection of tetrachloroethylene metabolites (Elfarra and
Krause. 2007). The nephrotoxic effects include increased kidney-to-body weight ratios, hyaline
droplet formation, glomerular —nephrosis' karyomegaly (enlarged nuclei), cast formation, and
other lesions or indicators of renal toxicity. The male rat has been shown to be more sensitive to
nephrotoxicity following exposure to tetrachloroethylene. These findings support a LOAEL of
200 ppm and a NOAEL of 50 ppm. Overall, multiple lines of evidence support the conclusion
that tetrachloroethylene causes nephrotoxicity in the form of tubular toxicity, mediated
potentially through the tetrachloroethylene GSH conjugation products TCVC and TCVCS.
4.10.2.3. Liver Toxicity
Two of four studies of occupationally exposed dry cleaners showed early indications of
liver toxicity, namely sonographic changes of the liver and altered serum concentrations of one
enzyme indicative of liver injury (Brodkin et al.. 1995; Gennari et al.. 1992). Frank liver disease
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-439 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
was not seen among these workers nor were changes in other biomarkers indicative of liver
toxicity (e.g., serum transaminases), which was not unexpected, given subjects with signs of liver
disease were excluded in both studies. LOAELs in these human studies were between 12 and
16 ppm (TWA).
Liver toxicity has been reported in multiple animal species by inhalation and oral
exposures to tetrachloroethylene. The effects are characterized by increased liver weight, fatty
changes, necrosis, inflammatory cell infiltration, triglyceride increases, and proliferation. The
mouse has been shown to be more sensitive to hepatic toxicity than the rat in multiple subchronic
and chronic studies (e.g.. JISA. 1993; NCI. 1977; NTP. 1986b; Schumann et al.. 1980). After
subchronic or chronic inhalation exposures in mice, liver toxicity is manifested by increased liver
weight (Kiellstrand et al.. 1984). liver enlargement (Kiellstrand etal.. 1984; Odum et al.. 1988a).
cytoplasmic vacuolation (fatty changes) (Kiellstrand et al.. 1984; NTP. 1986b; Odum et al..
1988b). centrilobular hepatocellular necrosis (JISA. 1993; NTP. 1986b). and inflammatory cell
infiltrates, pigment in cells, oval cell hyperplasia, and regenerative foci (NTP. 1986b). The
LOAEL for the inhalation studies, 9 ppm, is from a 30-day-exposure mouse study reporting
increased liver weight and morphological changes, and is supported by a finding of irreversible
macromolecular binding in mouse liver following a single, 6-hour exposure at 10 ppm. The
JISA (1993) chronic mouse inhalation bioassay reported liver necrotic foci at 50 ppm and higher.
With oral administration in mice, liver toxicity (increased liver weight, hepatocellular
swelling, necrosis, lipid accumulation, and increased DNA synthesis) has been observed at
100 mg/kg-day (Buben and O'Flahertv. 1985; Schumann et al.. 1980) and above (Berman et al..
1995; Ebrahim et al.. 1996; Goldsworthy and Popp. 1987; Jonker et al.. 1996). At
150 mg/kg-day administered for 30 days (Philip et al.. 2007). tetrachloroethylene increased ALT
levels transiently and stimulated fatty degeneration and necrosis, with ensuing regenerative
repair. These findings support a LOAEL of 100 mg/kg-day and a NOAEL of 20 mg/kg-day.
4.10.2.4. Immunotoxicity and hematologic toxicity
The strongest human study examining immunologic and hematologic effects of
tetrachloroethylene exposure in terms of sample size and use of an appropriately matched control
group is of 40 male dry-cleaning workers (mean exposure levels: <140 ppm; mean duration:
7 years; mean blood tetrachloroethylene levels: 1,685 |ig/L) by Emara et al. (2010). Statistically
significant decreases in red blood cell count and hemoglobin levels and increases in total white
cell counts and lymphocyte counts were seen in the exposed workers compared to age- and
smoking-matched controls. Similar effects were seen in mice (Ebrahim et al.. 2001). In
addition, increases in several other immunological parameters, including T-lymphocyte and
natural killer cell subpopulations, IgE, and interleukin-4 levels were observed in
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-440 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
tetrachloroethylene-exposed dry-cleaning workers (Emara et al.. 2010). These immunologic
effects suggest an augmentation of Th2 responsiveness. However, the limited available data
from studies in children (Delfino et al.. 2003a; Lehmann et al.. 2001; Lehmann et al.. 2002) do
not provide substantial evidence of an effect of tetrachloroethylene exposure during childhood on
allergic sensitization or exacerbation of asthma symptomology. The observation of the
association between increased tetrachloroethylene exposure and reduced interferon-y in cord
blood samples may reflect a sensitive period of development and points to the current lack of
understanding of the potential immunotoxic effects of prenatal exposures. The available data
pertaining to risk of autoimmune disease in relation to tetrachloroethylene exposure are limited
for ascertainment of disease incidence and exposure-assessment difficulties in population-based
studies.
The available data from experimental studies assessing immunotoxic responses in
animals are very limited (Aranyi et al.. 1986; Germolec et al.. 1989; Hani oka et al.. 1995a).
Additional data from inhalation, oral, and dermal exposures of different durations are needed to
assess the potential immunotoxicity of tetrachloroethylene along multiple dimensions, including
immunosuppression, autoimmunity, and allergic sensitization. The data from Aranyi et al.
(1986) suggest that short-term exposures may result in decreased immunological competence
(immunosuppression) in CD-I mice. The relative lack of data, taken together with the concern
that other structurally related solvents have been associated with immunotoxicity, particularly
relating to autoimmune disease (Cooper et al.. 2009). contributes to uncertainty in the database
for tetrachloroethylene. The limited laboratory animal studies of hematological toxicity
demonstrated an effect of tetrachloroethylene exposure on red blood cells (decreased RBCs)
(Ebrahim et al.. 2001). or decreased erythrocyte colony-forming units (Seidel et al.. 1992). with
reversible hemolytic anemia observed in female mice exposed to low drinking water levels
(0.05 mg/kg-bw day) of tetrachloroethylene beginning at 2 weeks of age in one series of studies
(Marth. 1987; Marth et al.. 1985a; Marth et al.. 1989). Ebrahim et al. (2001) also observed
decreased hemoglobin, platelet counts, and packed cell volume, and increased WBC counts.
Although limited experimental animal studies examining the immunotoxicity and hematologic
toxicity of tetrachloroethylene are available in the peer-reviewed published literature, the results
of these studies support the human epidemiology studies described above.
4.10.2.5. Reproductive and Developmental Toxicity
4.10.2.5.1. Reproductive toxicity
The literature contains few studies of effects on reproduction among subjects with
exposure to tetrachloroethylene. One study of primarily unionized workers in the dry-cleaning
and laundry industries in California observed subtle deficits in sperm quality in relation to
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-441 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
increasing levels of three measures of exposure, including tetrachloroethylene in exhaled breath
(Eskenazi etal.. 1991b). However, three clinically recognized measures of sperm quality were
not associated with exposure in the study population. The results of (Eskenazi et al.. 1991b) are
compelling, but more studies are needed to understand the spectrum of effects on sperm and their
impact on fecundity. Some studies that relied on detailed work histories and monitoring data to
classify exposure suggested that maternal or paternal exposure to tetrachloroethylene or work in
dry cleaning reduces fertility or delays conceptionfEskenazi et al.. 1991a; Sallmen et al.. 1998;
Sallmen et al.. 1995). However, the risk estimates were imprecise because the number of
participants reporting exposure to tetrachloroethylene was small. As a consequence, the existing
literature is limited and inconclusive concerning effects of tetrachloroethylene on reproduction
and fertility.
Results of several studies of maternal occupational exposure to tetrachloroethylene
suggest an increased risk of spontaneous abortion, particularly at higher levels of exposure
(Doyle et al.. 1997; Kyyronen et al.. 1989; Lindbohm et al.. 1990; Olsen et al.. 1990; Windham
et al.. 1991). Most of the studies evaluated exposure during the first trimester of pregnancy.
Some of the studies observed an increased odds ratio ranging from 1.4 to 4.7, but risk estimates
were statistically imprecise, and some studies were limited in their ability to evaluate potential
confounding (Bosco et al.. 1987; Lindbohm et al.. 1990; Olsen et al.. 1990; Windham et al..
1991). In general, the studies that used a more precise definition of exposure, or categorized
exposure into levels of increasing dose or intensity, observed higher risk estimates (Doyle et al..
1997; Kyyronen et al.. 1989; Lindbohm et al.. 1990; Olsen et al.. 1990). The Finnish studies
controlled for reported exposure to other substances in the workplace as well as for several
potential confounders. Increased risks were not found among dry cleaners in Sweden using a
similar study design (Ahlborg. 1990b; Olsen et al.. 1990). Although there is no evidence of an
increased risk associated with paternal exposure, the studies were not of sufficient size or detail
in exposure estimates to draw conclusions (Eskenazi etal.. 1991b; Lindbohm et al.. 1991;
Taskinen et al.. 1989). No associations with incidence of spontaneous abortion were observed
between two populations exposed to tetrachlorethylene in drinking water (Aschengrau et al..
2008; Lagakos et al.. 1986). The populations were likely exposed to lower levels compared to
the occupational populations. In addition, the window of exposure used to assess risk in both
studies may not have been precise enough to detect a small elevation in risk for spontaneous
abortion.
The database of experimental animal studies for tetrachloroethylene includes evaluations
of reproductive and fertility outcomes in rats and mice following inhalation exposures.
Additionally, an in vitro assay of oocyte fertilizability is available. An assessment of fertility
and reproductive function in rats exposed to tetrachloroethylene via inhalation over the course of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-442 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
two generations was conducted by Tinston (1994). Effects on offspring included decreased pup
weights and postnatal survival in both generations, as well as behavioral alterations in the F1
pups. Decreased mean testes weight was observed in Fla males; however, no effects on male or
female fertility or other evidence of alterations in reproductive function were observed. For
males, this finding is supported by the results of a study by Beliles et al. (1980). who found no
sperm abnormalities in rats following up to 10 weeks of tetrachloroethylene inhalation
exposures. While Beliles et al. (1980) identified an increase in abnormal sperm heads in mice
after 4 weeks of exposure, no other reproductive toxicity data in mice were available to aid in the
interpretation of this finding. A limitation of the (Tinston. 1994) study included a concern about
a short peri-parturition exposure gap. Additionally, the study was conducted in accordance with
standard EPA and OECD toxicological study guidelines in place at the time but did not assess
endpoints that are included in the guidelines that were revised and harmonized in 1998.
4.10.2.5.2. Developmental toxicity
A few epidemiologic studies evaluated developmental toxicity endpoints such as
decreased birth weight, intrauterine growth restriction (IUGR; also known as small for gestation
age [SGA]), and congenital anomalies. Overall, no associations were noted in several studies
that assessed maternal or paternal occupational exposure to tetrachloroethylene and increased
incidence of stillbirths, congenital anomalies, or decreased birth weight (Bosco et al.. 1987;
Kvvronen etal.. 1989; Lindbohm. 1995; Olsen et al.. 1990; Taskinen et al.. 1989; Windham et
al.. 1991). However, congenital anomalies were analyzed as a combined group, and the number
of exposed cases for specific types of anomalies was not sufficient to evaluate risk with
statistical precision. Some studies of tetrachloroethylene in drinking water reported that
exposure during pregnancy is associated with low birth weight (Bove et al.. 1995; Lagakos et al..
1986). eye/ear anomalies (Lagakos et al.. 1986). and oral clefts (Aschengrau et al.. 2009b; Bove
et al.. 1995; Lagakos etal.. 1986). No associations with tetrachloroethylene exposure were
reported for small for gestational age (Bove et al.. 1995) or other classifications of congenital
anomalies [e.g., musculoskeletal, cardiovascular (Lagakos et al.. 1986)1). Although a small
increase in risk of small for gestational age was reported for infants exposed prenatally to
tetrachloroethylene at the Camp Lejeune military base, the finding remains inconclusive until
ATSDR completes its reanalysis. Aschengrau et al. (2008) did not observe associations with
birth weight or gestational age in a Cape Cod population living in communities receiving
drinking water containing a wide range of tetrachloroethylene concentrations. Participants in
some of the studies of drinking water contamination were exposed to multiple pollutants (Bove
et al.. 1995; Lagakos etal.. 1986). and it was not possible to disentangle substance-specific risks.
Diagnoses of attention deficit or educational histories reported by the mothers were not increased
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-443 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
in relation to the amount of tetrachloroethylene delivered to the homes during pregnancy or
childhood (Janulewicz et al.. 2008). Finally, a more than threefold risk of schizophrenia was
associated with dry cleaning as a surrogate for prenatal tetrachloroethylene exposure (Perrin et
aL^.2007). The longitudinal design and use of a national registry to identify psychiatric
diagnoses were strengths of the study, but tetrachloroethylene exposure was not directly
analyzed.
The animal inhalation study database includes assessments of developmental toxicity in
rats, mice, and rabbits following exposures during gestation and assessments of developmental
neurotoxicity in rats following pre- and/or postnatal exposures of the offspring. Additional
supportive studies include in vitro assays of embryo development and oocyte fertilizability, a
developmental assay in Japanese medaka, and two oral gavage studies that assessed
developmental toxicity in rats and developmental neurotoxicity in mice. The tetrachloroethylene
database included assessments of the various potential manifestations of developmental toxicity,
i.e., alterations in survival, growth, morphology, and functional development. Indications of
effects on prenatal survival following in utero exposure included increased pre- and/or
postimplantation loss in rats, mice, and rabbits (Schwetz et al.. 1975; Szakmary et al.. 1997).
These findings were supported by evidence of embryo mortality in a rat whole embryo culture
(WEC) assay (Saillenfait et al.. 1995) and decreased viability in a Japanese medaka assay
(Spencer et al., 2001). Decreased prenatal growth was observed in mice (Schwetz et al.. 1975)
and rats (Szakmary et al.. 1997). Morphological alterations associated with prenatal exposures to
tetrachloroethylene included delays in skeletal ossification in mice (Schwetz et al.. 1975) and
rats (Carney et al.. 2006; Szakmary et al.. 1997)(. which were often associated with fetal weight
decrements, and increased incidences of malformations in mice, rats, and rabbits (Szakmary et
al.. 1997). Evidence of tetrachloroethylene exposure-related malformations was also observed in
the rat WEC and medaka assays (Saillenfait et al., 1975; Spencer et al., 2001) and in a gavage
prenatal developmental toxicity screening study in rats (Narotsky and Kavlock. 1995).
Alterations in neurological function following pre- and/or postnatal inhalation exposures to
tetrachloroethylene were observed in rats by Szakmary et al. (1997). Nelson et al. (1980). and
Tinston (1994). These findings were supported by a study that found altered spontaneous motor
activity in young adult rats that had been exposed orally to tetrachloroethylene postnatally during
a critical period of nervous system development (Fredriksson et al.. 1993). Additionally,
reductions in brain acetylcholine and dopamine were observed in rat offspring following
gestational tetrachloroethylene exposures (Nelson et al.. 1980). Limitations of the inhalation
developmental toxicity studies include the lack of dose-response information due to the use of a
single treatment level in the prenatal developmental toxicity assessment by Schwetz et al. (1975);
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-444 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
the lack of either maternal or developmental toxicity in Hardin et al. (1981); and absence of
methodological details in study reporting (Szakmary et al.. 1997).
4.10.2.5.3. Synthesis of human and animal reproductive and developmental toxicity
The finding of spontaneous abortions in several human studies of dry cleaners is
supported by the occurrence of reduced birth weight and mortality in several animal studies.
Although not a consistent finding in epidemiology studies, the finding of low birth weight in a
study of contaminants in drinking water (Bove et al.. 1995) is supported by reduced birth weight
in five animal studies (Carney et al.. 2006; Nelson et al.. 1980; Schwetz et al.. 1975; Szakmary et
al.. 1997) and in the F1 generation but not the F2 generation of Tinston (1994). There are no
human observations of behavioral changes to compare with the animal evidence of CNS effects.
The subtle effects on sperm seen in humans (Eskenazi etal.. 1991b) correspond to one report of
abnormal sperm in mice. Overall, the developmental and reproductive toxicity database for
tetrachloroethylene was judged to include a range of data from appropriate well-conducted
studies in several laboratory animal species plus limited human data and was considered
sufficient for hazard characterization and dose-response assessment, based upon EPA risk
assessment guidelines (U.S. EPA. 1991a. 2006b). Based upon a consideration of the available
database of animal developmental and reproductive toxicity studies for tetrachloroethylene, the
overall inhalation NOAEL is 100 ppm, based on Tinston (1994). The overall inhalation LOAEL
is 300 ppm, based on Tinston (1994) and Schwetz et al. (1975). in which increased mortality and
decreased body weight of the offspring were observed.
4.10.2.6. Summary of Noncancer Toxicities and Identification of Studies for Dose-
Response Analyses
Noncancer effects of tetrachloroethylene identified in exposed humans and animals
include toxicity to the central nervous, renal, hepatic, immune, and hematologic systems, and on
development and reproduction. Neurotoxic effects have been characterized in human
occupational and residential studies, as well as in experimental animal studies, providing
evidence of an association between tetrachloroethylene exposure and neurobehavioral deficits.
Tetrachloroethylene exposure primarily results in visual changes, increased reaction time, and
decrements in cognition in humans; in animal studies, effects on vision, visual-spatial function,
and reaction time, as well as brain-weight changes were also seen. Adverse effects on the kidney
in the form of tubular toxicity, potentially mediated through the tetrachloroethylene GSH
conjugation products TCVC and TCVCS, have been reported in numerous well-conducted
animal studies. Although epidemiological studies have not systematically investigated
nephrotoxicity, an association between inhalation tetrachloroethylene exposure and chronic
kidney disease, as measured by urinary excretion of renal proteins and ESRD, is supported. The
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-445 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
developmental and reproductive toxicity database for tetrachloroethylene includes a range of
data from appropriate well-conducted studies in several laboratory animal species plus limited
human data. Evidence of liver toxicity is primarily from several well-conducted rodent studies,
including chronic bioassays.
Other toxicity endpoints are less well characterized. The few published reports of experi
ental studies examining immune or hematologic system toxicity are consistent with the
limited findings in the human occupational exposure studies. These include, as noted by NRC
(2010). a series of reports by Marth (1987; 1985a; 1989) providing evidence of hemolytic
anemia in young (2-week-old) female mice exposed at low levels of tetrachloroethylene in
drinking water (0.05 or 0.1 mg/kg-day for 7 weeks). The relative lack of additional data,
including confirmatory reports of immunotoxic or hematologic toxicity with low continuous
exposures beginning in early lifestages, taken together with evidence of immunotoxicity from
structurally related solvents (Cooper et al.. 2009). contributes to uncertainty in the database for
tetrachloroethylene. No epidemiological studies identified potential noncancer respiratory
toxicities, and no lung effects in rodents were reported in chronic bioassays (NCI. 1977; NTP.
1986b) or other published reports.
The tables below present the inhalation (see Table 4-49) and oral (see Table 4-50)
findings of tetrachloroethylene toxicity, arranged by organ toxicity domains, which are suitable
for dose-response analyses. The NOAELs and LOAELs from candidate dose-response studies
are identified. In examining the studies judged to be suitable for dose-response analyses, it is
evident that the neurotoxicological findings consistently occur at the lowest exposure levels.
Additionally, the database for neurotoxicity comprises a number of both occupational and
residential human studies as well as animal studies that are suitable for dose-response analyses.
Residential inhalation exposures to tetrachloroethylene resulted in visual contrast sensitivity
changes and cognitive and motor changes at exposures approximately 5- to 10-fold lower than
the lowest sensitive exposure in other organ toxicity domains. Similarly, with oral doses,
developmental neurotoxicity effects were observed at levels at least fivefold lower (Fredriksson
et al.. 1993). Therefore, the CNS effects are identified as a sensitive endpoint following either
oral or inhalation exposure to tetrachloroethylene. Section 5 presents dose-response analyses of
the neurotoxicity data set as a basis for derivation of inhalation and oral reference values.
Quantitative dose-response analyses of the findings in other toxicity domains (i.e., kidney, liver,
reproductive and developmental toxicity) are also presented in Section 5. In addition to
providing information regarding the relative sensitivity of different organs/systems to
tetrachloroethylene, such quantitative analyses may be useful for cumulative risk assessment in
which multiple chemicals have a common target organ/system other than the CNS.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-446 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
4.10.3. Characterization of Cancer Hazard
Following EPA (2005a) Guidelines for Carcinogen Risk Assessment, tetrachloroethylene
is 4ikely to be carcinogenic in humans by all routes of exposure." This characterization is based
on credible evidence of carcinogenicity in epidemiologic studies and conclusive evidence that
the administration of tetrachloroethylene, either by ingestion or by inhalation to sexually mature
rats and mice, increases tumor incidence (JISA. 1993; NCI. 1977; NTP. 1986b). Several rodent
tumor types were significantly increased with tetrachloroethylene administration in at least two
studies. Mouse liver tumors (hepatocellular adenomas and carcinomas) and rat mononuclear cell
leukemia were reported in both sexes in two lifetime inhalation bioassays employing different
rodent strains, and mouse liver tumors were also reported in both sexes in an oral bioassay (NCI.
1977). Tumors reported in single inhalation bioassays include kidney and testicular interstitial
cell tumors in male F344 rats (NTP. 1986b). brain gliomas in male and female F344 rats (NTP.
1986b). and hemangiomas or hemangiosarcomas in male Crj :BDF1 mice (JISA. 1993). Several
metabolites of tetrachloroethylene also are considered rodent carcinogens. TCA and DCA
produce liver tumors in mice, and DCA also induces liver tumors in rats (reviewed in EPA's
TCA Toxicological Review). Other tetrachloroethylene metabolites have not been tested in a
rodent bioassay.
The specific active moiety(ies) and mode(s) of action involved in the carcinogenicity of
tetrachloroethylene and its metabolites are not known. For rat kidney tumors, it is generally
believed that metabolites resulting from GSH conjugation of tetrachloroethylene are involved.
The hypothesized modes of action for this endpoint include mutagenicity, peroxisome
proliferation, a2[j,-globulin nephropathy, and cytotoxicity not associated with a2[j,-globulin
accumulation. For mouse liver tumors, it is generally believed that metabolites resulting from
P450-mediated oxidation of tetrachloroethylene are involved. The mode of action (MO A)
hypotheses for this endpoint concern mutagenicity, epigenetic effects (especially DNA
hypomethylation), oxidative stress, and receptor activation (focusing on a hypothesized PPARa-
activation MO A). However, the available evidence is insufficient to support the conclusion that
either rat kidney or mouse liver tumors are mediated solely by one of these hypothesized modes
of action. In addition, no data are available concerning the metabolites or the mechanisms that
may contribute to the induction of other rodent tumors (including mononuclear cell leukemia,
brain gliomas, or testicular interstitial cell tumors in exposed rats and hemangiosarcomas in
exposed mice). Furthermore, no mechanistic hypotheses have been advanced for the human
cancers suggested to be increased with tetrachloroethylene exposure in epidemiologic studies,
including bladder cancer, non-Hodgkin lymphoma and multiple myeloma. Although
tetrachloroethylene is largely negative in genotoxicity assays including in the Ames mutagenicity
test, tetrachloroethylene has been shown to induce modest genotoxic effects (micronuclei
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-447 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
induction following in vitro or in vivo exposure, and DNA binding and single-strand breaks in
tumor tissue) and mutagenic effects under certain metabolic activation conditions. In addition,
some tetrachloroethylene metabolites have been shown to be mutagenic. Thus, the hypothesis
that mutagenicity contributes to the tetrachloroethylene carcinogenesis cannot be ruled out for
one or more target organs, although the specific metabolic species or mechanistic effects are not
known.
4.10.4. Synthesis of Epidemiologic Studies
The available epidemiologic studies provide a pattern of evidence associating
tetrachloroethylene exposure and several types of cancer, specifically bladder cancer,
non-Hodgkin lymphoma, and multiple myeloma. Associations and exposure-response
relationships for these cancers were reported in studies using higher quality (more precise)
exposure assessment methodologies for tetrachloroethylene. Confounding by common lifestyle
factors such as smoking are unlikely explanations for the observed results. For other sites,
including esophageal, kidney, lung, liver, cervical, and breast cancer, more limited data
supporting a suggestive effect are available.
With respect to bladder cancer, the pattern of results from this collection of studies is
consistent with an elevated risk for tetrachloroethylene of a relatively modest magnitude (i.e., a
10-40% increased risk). The results from five of the six studies with relatively high quality
exposure-assessment methodologies provide additional evidence of an association with effect
estimates ranging from 1.44 to 4.03 (Aschengrau et al.. 1993; Blair et al.. 2003; Calvert et al.. In
Press; Lynge et al.. 2006; Pesch et al.. 2000a; Siemiatvcki. 1991). The Lynge et al. (2006) risk
estimates were slightly higher among the subgroup from Denmark and Norway, in which the
number of subjects with unclassifiable data was negligible (relative risk: 1.69, 95% CI: 1.18,
2.43). An exposure-response gradient was seen in a large case-control study by Pesch et al.
(2000a). using a semiquantitative cumulative exposure assessment, with an adjusted odds ratio of
0.8 (95% CI: 0.6, 1.2), 1.3 (95% CI: 0.9, 1.7), and 1.8 (95% CI: 1.2, 2.7) for medium, high, and
substantial exposure, respectively, compared to low exposure, based on the JTEM approach. An
exposure-response gradient was not seen in the study by Lynge et al. (2006) using employment
duration without consideration of exposure concentration. In addition, relative risk estimates
between bladder cancer risk and ever having a job title of dry-cleaner or laundry worker in four
large cohort studies ranged from 1.01 to 1.44 (Ji and Hemminki. 2005a; Pukkala et al.. 2009;
Travier et al.. 2002; Wilson et al.. 2008). As expected, the results from the smaller studies are
more variable and less precise, reflecting their reduced statistical power. Confounding by
smoking is an unlikely explanation for the findings, given the adjustment for smoking by Pesch
et al. (2000a) and in other case-control studies.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-448 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
The results from the collection of studies pertaining to non-Hodgkin lymphoma also
indicate an elevated risk for tetrachloroethylene. There is little evidence of an association in the
large cohort studies examining risk in relation to a broad occupational category of work in
laundry or dry cleaning (i.e., relative risk estimates ranging from 0.95 to 1.05 in females in
(Andersen et al.. 1999). females and males in Ji and Hemminki (2006). The results from five
cohort studies that used a relatively high quality exposure-assessment methodology generally
reported relative risks between 1.7 and 3.8 (Anttila et al.. 1995; Boice et al.. 1999; Calvert et al..
In Press; Radican et al.. 2008; Selden and Ahlborg. 2011). There is also some evidence of
exposure-response gradients in studies with tetrachloroethylene-specific exposure measures
based on intensity, duration, or cumulative exposure (Boice et al.. 1999; Miligi et al.. 2006;
Seidler et al.. 2007). Higher non-Hodgkin lymphoma risks were seen in these studies in the
highest exposure categories, with the strongest evidence from the large case-control study in
Germany in which a relative risk of 3.4 (95% CI: 0.7, 17.3) was seen in the highest cumulative
exposure category (trend />value = 0.12) (Seidler et al.. 2007). Confounding by life-style factors
are unlikely explanations for the observed results because common behaviors, such as smoking
and alcohol use, are not strong risk factors for non-Hodgkin lymphoma (Besson et al.. 2006;
Morton et al.. 2005).
The larger cohort studies that use a relatively nonspecific exposure measure (broad
occupational title of launderers and dry cleaners, based on census data) do not report an
increased risk of multiple myeloma, with effect estimates ranging from 0.99 to 1.07 (Andersen et
al.. 1999; Ji and Hemminki. 2006; Pukkala et al.. 2009). Some uncertainty in these estimates
arises from these studies' broader exposure-assessment methodology. Results from the cohort
and case-control studies with a higher quality exposure-assessment methodology, with an
exposure measure developed specifically for tetrachloroethylene, do provide evidence of an
association, however, with relative risks of 7.84 (95% CI: 1.43, 43.1) in women and 1.71 (95%
CI: 0.42, 6.91) in men in the cohort of aircraft maintenance workers (Radican et al.. 2008) and
1.5 (95% CI: 0.8, 2.9) in a case-control study in Washington (Gold et al.. 2010b);
tetrachloroethylene exposure). Gold et al. also reported increasing risks with increasing
exposure duration (based on job titles) (Gold et al.. 2010a) and based on a cumulative
tetrachloroethylene exposure metric (Gold et al.. 2010b). Two smaller studies with
tetrachloroethylene-specific exposure measures based on intensity, duration, or cumulative
exposure did not observe an exposure-response trend: a study by Seidler et al. (2007) observed
no cases among the highest exposure groups, and a study by Boice et al. (1999) of aerospace
workers observed one death among routinely exposed subjects and six deaths among subjects
with a broader definition of routine or intermittent exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-449 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Suggestive but limited evidence was also seen in the collection of epidemiologic studies
pertaining to tetrachloroethylene exposure and esophageal, kidney, lung, liver, cervical, and
breast cancer. One difference between these sets of data and the data for bladder cancer,
non-Hodgkin lymphoma, and multiple myeloma is a more mixed pattern of observed risk
estimates and an absence of exposure-response data from the studies using a quantitative
tetrachloroethylene-specific cumulative exposure measure.
For esophageal cancer, the SIR in the only large cohort study (n = 95 cases), a study
using broad exposure categories, was 1.18 (95% CI: 0.96, 1.46) (Pukkala et al.. 2009). The point
estimates of the association in seven of eight smaller studies, four studies with specific exposure
assessments, and four other studies with less precise assessments were between 1.16 and 2.44
(Blair et al.. 2003; Boice et al.. 1999; Calvert et al.. In Press; Lynge and Thygesen. 1990;
Pukkala et al.. 2009; Selden and Ahlborg. 2011; Sung et al.. 2007; Travier et al.. 2002). Two
small case-control studies with relatively high quality exposure-assessment approaches, Lynge et
al. (2006) and Vaughan et al. (Vaughan et al.. 1997) reported an odds ratio of 0.76 (95%
CI: 0.34, 1.69) and of 6.4 (95% CI: 0.6, 68.9), respectively. Some uncertainties in these estimate
arises from the lack of job title information for 25% of the cases and 19% of the controls, and the
variability in the results from the sensitivity analysis using different assumptions regarding the
correct classification of individuals in this group or the small number of exposed cases. One of
the two larger studies examining exposure-response suggested a positive relationship (Calvert et
al.. In Press). Based on smoking rates in blue-collar workers, the twofold risk estimate reported
in (Calvert et al.. In Press) and Blair et al. (2003) was higher than that which could reasonably be
attributable to smoking.
One primary study that supports an association between tetrachloroethylene exposure and
kidney cancer, the largest international case-control study (245 exposed cases from Australia,
Denmark, Germany, Sweden, and the United States), reported a relative risk of 1.4 (95% CI: 1.1,
1.7) for any exposure to dry-cleaning solvents (Mandel etal.. 1995). This study was able to
adjust for smoking history, BMI, and other risk factors for kidney cancer. Results from the large
cohort studies, using a more general exposure classification based on national census occupation
data, present more variable results, with relative risks of 0.94, 1.11, and 1.15 in Pukkula et al.
(2009). Travier et al. (2002). and Ji et al., respectively. The results from the smaller studies
using a relatively specific exposure-assessment approach to refine classification of potential
tetrachloroethylene exposure in dry-cleaning settings are mixed, with some studies reporting
little or no evidence of an association (Aschengrau et al.. 1993; Boice et al.. 1999; Dosemeci et
al.. 1999; Lynge et al.. 2006; Pesch et al.. 2000a). and other studies reported elevated risks
(Anttila et al.. 1995; Blair et al.. 2003; Calvert et al.. In Press; Schlehofer et al.. 1995). An
increasing trend in relative risk with increasing exposure surrogate was not seen in any of the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-450 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
larger occupational exposure studies with three or more exposure categories (Lynge et al.. 2006;
Mandel et al.. 19951 but some indication of higher risk with higher exposure (or duration) was
seen in other studies (Blair et al.. 2003).
For lung cancer risk, the results from seven large cohort studies of dry cleaners are
consistent with an elevated lung cancer risk of 10-40% (Blair et al.. 2003; Calvert et al.. In
Press; Ji et al.. 2005b; Lynge and Thygesen. 1990; Pukkala et al.. 2009; Schlehofer et al.. 1995;
Selden and Ahlborg. 2011; Travier et al.. 2002). Similar results were seen in four of the five
occupational studies that were identified as having a relatively strong exposure-assessment
methodology (Anttila et al., 1997)(Blair et al.. 2003)(Boice et al.. 1999; Calvert et al.. In Press).
However, Selden and Ahlborg (2011) observed similar, but slightly higher, relative risks for
laundry workers compared with dry-cleaning workers in their study. These studies were unable
to control for potential confounding from cigarette smoking; however, and the magnitude of the
association in these studies is consistent with that expected assuming the prevalence of smoking
among dry-cleaners and laundry workers was slightly higher (e.g., 10% higher) than among the
general population. Features of the selection of study participants and study analysis in the
available case-control studies reduce the potential for confounding by smoking, however. Two
case-control studies were limited to either nonsmokers or ex-smokers and both of these studies
indicate an approximate twofold increased risk with a history of work in the dry-cleaning
industry (OR: 1.8, 95% CI: 1.1, 3.0) in Brownson et al., (1993). and OR: 1.83, 95% CI: 0.98,
3.40 among women in Pohlabeln et al. (2000). The other case-control studies adjusted for
smoking history, and the results for these (somewhat smaller studies) are similar to the
previously cited estimates. Among the studies that evaluated exposure-response gradients, the
evidence for a trend in risk estimates was mixed (Blair et al.. 2003; Boice et al.. 1999; Brownson
et al.. 1993; Calvert et al.. In Press; Paulu et al.. 1999; Travier et al.. 2002).
For liver cancer, studies carrying greater weight in the analysis based on the large number
of observed events or exposed cases, or based on a strong exposure-assessment approach show a
mixed pattern of results. The one case-control study with a large number of exposed liver cancer
cases and a relatively high quality exposure-assessment methodology reported an odds ratio
estimate of 0.76 (95% CI: 0.38, 1.72) for liver cancer and dry cleaning (Lynge et al.. 2006). A
recent multiple Nordic country cohort study and two cohort studies of Swedish subjects with
broad exposure-assessment approaches, and whose subjects overlapped with Lynge et al. (2006).
reported SIRs of 1.02 (95% CI: 0.84, 1.24), 1.22 (95% CI: 1.03, 1.45), and 1.23 (95% CI: 1.02,
1.49) for liver and biliary tract cancer and work as a dry-cleaner or laundry worker in Travier et
al. (2002). Ji and Hemminki (2005c). and Pukkala et al. (2009). respectively. Three other studies
with strong exposure-assessment approaches specific to tetrachloroethylene, but whose risk
estimates are based on fewer observed liver cancer cases or deaths, reported risk estimates of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-451 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
1.21 to 2.05 for the association between liver cancer and tetrachloroethylene (Blair etal.. 1979;
Boice et al.. 1999; Bond et al.. 1990; Selden and Ahlborg. 2011). However, dry cleaning or
workers employed after 1960 when tetrachloroethylene use was more prevalent did not have a
higher liver cancer risk estimate than laundry workers (Lynge et al.. 2006; Selden and Ahlborg.
2011). Exposure response was not observed, and the SIR for tetrachloroethylene-exposed
subjects with the longest employment duration in Selden and Ahlborg (2011) was lower than that
for subjects with shorter employment duration. Potential confounding may be an alternative
explanation as no study adjusted for known and suspected risk factors for liver cancer (Boice et
al.. 1999; Bond et al.. 1990; Ji and Hemminki. 2005c; Lynge et al.. 2006; Pukkala et al.. 2009;
Selden and Ahlborg. 2011; Travier et al.. 2002). Nine other cohort and case-control studies with
fewer observed events and/or a broad exposure-assessment methodology carried less weight in
the analysis and reported a mixed pattern of results (Blair et al.. 2003; Calvert et al.. In Press;
Lindbohm et al.. 2009; Lynge et al.. 1995; Stemhagen et al.. 1983; Suarez et al.. 1989; Sung et
al.. 2007; Vartiainen et al.. 1993) Lee et al., 2006;. Lee et al. (2006) reported a risk estimate of
2.57 (95% CI: 1.21, 5.46) for the association between liver cancer and residence in a village with
ground water contamination, but subjects were from a region with a high prevalence of HCV
infection, and HCV status may confound the observed association.
For cervical cancer, the results from the two large cohort studies with a broad exposure
assessment are consistent with an elevated cervical cancer risk of 20-30% (Pukkala et al.. 2009;
Travier et al.. 2002). Results from four smaller cohort and case-control studies with a relatively
high quality exposure-assessment methodology presented a pattern of more variable results, with
relative risks of 0.98 (95% CI: 0.65, 1.47), 1.19 (95% CI: 0.64, 1.93), 2.10 (95% CI: 0.68, 4.90),
and 3.20 (95% CI: 0.39, 11.6) in Lynge et al. (2006). Selden and Ahlborg (Calvert et al.. In
Press; 2011). and Anttila et al. (1995). respectively. A fourth study with higher quality
exposure-assessment specific to tetrachloroethylene did not observe any cervical cancer deaths
among women, but less than one death was expected (Boice et al.. 1999). Calvert et al. was the
only study to report an exposure response gradient with employment duration. Dry cleaning or
workers employed after 1960 when tetrachloroethylene use was more prevalent did not have
higher cervical cancer risks compared with laundry workers (Lynge et al.. 2006)(Selden and
Alhborg, 2011). Lack of data on socioeconomic status—a proxy for exposure to the human
papilloma virus, a known risk factor for cervical cancer—indicates great uncertainty for asserting
this association with tetrachloroethylene exposure. Potential confounding by socioeconomic
status is an alternative explanation with some support provided by Lynge et al. (2006). a case-
control study with controls of similar socioeconomic status as cases and that did not observe an
association between cervical cancer and dry cleaning.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-452 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
The results from the large studies of breast cancer risk in women in relation to
tetrachloroethylene exposure are mixed. The largest, based on 1,757 breast cancer cases in
female dry-cleaners and laundry workers, reported a statistically significant deficit in the risk of
breast cancer incidence compared to the populations of Nordic countries (Pukkala et al.. 2009).
Findings in the other four studies were based on fewer events or exposed cases; two of four
studies with a nonspecific exposure-assessment methodology provided evidence for association
between breast cancer in females and tetrachloroethylene exposure (Anderson et al.. 1999;
Aschengrau et al.. 2003; Chang. 2005; Lynge and Thygesen. 1990; Sung et al.. 2007). but no
association was seen in two other large cohort studies with a relatively high quality exposure-
assessment methodology to tetrachloroethylene (Blair et al.. 2003; Selden and Ahlborg. 2011).
Small studies also observed mixed findings (Aschengrau and Seage. 2003; Boice et al.. 1999;
Calvert et al.. In Press; Peplonska et al.. 2007; Radican et al.. 2008; Sung et al.. 2007).
Although cohort studies were unable to control for potential confounding from reproductive
history or menopausal status, observations in case-control studies controlled for these potential
confounders in statistical analyses and provided support of an association between female breast
cancer and tetrachloroethylene compared to controls (Aschengrau and Seage. 2003; Band et al..
2000; Peplonska et al.. 2007). Three studies examined exposure response, and two of these
studies with semiquantitative or quantitative exposure-assessment approaches reported risk
estimates in females monotonically increased in higher exposure groups (Aschengrau et al..
2003; Blair et al.. 2003). A third study examining exposure duration observed an inverse relation
(Peplonska et al.. 2007). Exposure duration is more uncertain than use of a semiquantitative
surrogate given increased potential for bias associated with exposure misclassification. Because
of the limitation in statistical power, none of the five studies reporting on male breast cancer is
adequate to examine tetrachloroethylene exposure (Anderson et al.. 1990; Chang et al.. 2005;
Lynge and Thygesen. 1990; Pukkala et al.. 2010; Selden and Ahlborg. 2011).
4.10.5. Synthesis of Rodent Cancer Bioassay Findings
One oral gavage (NCI. 1977) and two inhalation (JISA. 1993; NTP. 1986b) cancer
bioassays provide evidence of tetrachloroethylene carcinogenicity in rats and mice. In male and
female rats, inhalation exposure to tetrachloroethylene significantly increased the incidence of
mononuclear cell leukemia (MCL) in independent bioassays of the F344/N (JISA. 1993; NTP.
1986b) or F344/DuCrj (JISA. 1993) strain. Tetrachloroethylene reduced MCL latency in
females in both studies. In addition, the NTP bioassay reported dose-related increases in the
severity of MCL in males and females. Additional tumor findings in rats included significant
increases in the NTP bioassay of two rare tumor types, kidney tumors in males, and brain
gliomas in males and females. Additionally, the NTP (1986b) bioassay reported increases in the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-453 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
rate of testicular interstitial cell tumors, a tumor type of high incidence in unexposed male F344
rats. Other evidence, including that brain gliomas occurred earlier with tetrachloroethylene
exposure than in control animals, and that the related compound trichloroethylene is a kidney
carcinogen in rats and humans and a testicular carcinogen in rats, support the significance of
these findings. A third rat bioassay, of oral gavage exposure in Osborne-Mendel rats, was
inconclusive with respect to carcinogenicity due to a high incidence of respiratory disease in all
animals and shortened survival in tetrachloroethylene-exposed animals (NCI. 1977).
In male and female mice, tetrachloroethylene exposure via inhalation (JISA. 1993; NTP.
1986b) or oral gavage (NCI. 1977) significantly increased the incidence of hepatocellular
adenomas and carcinomas. The NCI and NTP studies employed the B6C3Fi strain, while the
JISA study examined the Crj :BDF1 strain. The JISA study reported increases in hemangiomas
or hemangiosarcomas of the liver, spleen, fat, and subcutaneous skin in exposed male CrJ:BDFl
mice.
In summary, tetrachloroethylene increased the incidence of liver tumors (hepatocellular
adenomas and carcinomas) in male and female mice and of MCL in both sexes of rats. These
findings were reproducible in multiple lifetime bioassays employing different rodent strains and,
in the case of mouse liver tumors, by inhalation and oral exposure routes. Additional tumor
findings in rats included significant increases in the NTP bioassay of testicular interstitial cell
tumors and kidney tumors in males, and brain gliomas in males and females. In mice,
hemangiosarcomas in liver, spleen, fat, and subcutaneous skin were reported in males in the
JISA study. The rat and mouse findings are summarized in Tables 4-51 and 4-52, respectively,
and in the sections below.
4.10.5.1. Carcinogenicity Findings in Rats
The NCI oral gavage study in Osborne-Mendel rats was considered to be inconclusive
because of the high incidence of respiratory disease, and high mortality with tetrachloroethylene
exposure. Lesions indicative of pneumonia were observed in almost all rats at necropsy. A high
incidence of toxic nephropathy was evident in tetrachloroethylene-exposed male and female rats.
Early mortality was also seen in tetrachloroethylene-exposed animals; 50% of the high dose
males and females had died by Weeks 44 and 66, respectively. Therefore, this bioassay is not
considered further in the below evaluation of the carcinogenicity of tetrachloroethylene in rats.
The NTP (1986b) and JISA (1993) inhalation bioassays reported increases in the
incidence of mononuclear cell leukemia (MCL) in male and female F344/N or F344/DuCij rats.
Supplemental analyses by NTP indicated that tetrachloroethylene produced a dose-related
increase in the severity of MCL in both males and females. Additionally, NTP found that
tetrachloroethylene exposure significantly shortened the time to onset of MCL in females.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-454 DRAFT-DO NOT CITE OR QUOTE
-------
1 Although survival was unaffected, the incidence of advanced MCL increased in female rats that
2 died before the scheduled study termination. MCL incidences were higher in the concurrent than
3 in the historical chamber control groups at the performing laboratory (males: 28/50 [56%] vs.
4 117/250 [47%]; females: 18/50 [36%] vs. 73/249 [29%]). The concurrent control rates were also
5 higher than the NTP program historical rate for untreated control groups (males: 583/1,977
6 [29%]; females: 375/2,021 [18%]).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-455 DRAFT-DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-51. Tumor incidence in rats exposed to tetrachloroethylene
Bioassay
Doses/exposures
Sex
Reported cumulative tumor incidencea(%)
Admin.
Continuous
equivalent
Hepatocellular
adenomas or
carcinomas
Hemangioma
or hemangio-
sarcomasb
Renal
adenomas or
carcinomas
Mononuclear
cell leukemia0
Testicular
interstitial
cell tumors
Brain
gliomas
NCI (1977)d
Osborne-Mendel
rats
Gavage:
5 d/wk,
78 wk
Vehicle
500 mg/kg-day
1,000 mg/kg-day
0e
471 mg/kg-day
941 mg/kg-day
Male
None reported3
1/20
1/49
0/50
2f/20 (5)
lf/49 (2)
0/50 (0)
None reported
None
reported
None
reported
Vehicle
500 mg/kg-dayf
1,000 mg/kg-day
0f
474 mg/kg-day
974 mg/kg-day
Female
None reported
None reported
0/20 (0)
0/50 (0)
0/50 (2)
None reported
N/A
None
reported
NTP (1986b)
F344/N rats
Inhalation:
6 h/d,
5 d/wk,
104 wk
0
200 ppm
400 ppm
0
36 ppm
72 ppm
Male
0/50 (0)
1/50 (2)
1/50 (2)
0/50
0/50
0/50
1/49 (2)
3/49 (6)
4/50 (8)
28/50 (56)
37/50 (77)
37/50 (74)
36/50 (76)
39/49 (80)
41/50 (82)
1/50 (2)
0/50 (0)
4/50 (8)
0
200 ppm
400 ppm
0
36 ppm
72 ppm
Female
0/50
0/50
0/50
0/50
0/50
0/50
0/47
0/44
0/46
18/50 (36)
30/50 (60)
29/50 (58)
N/A
1/50 (2)
0/50 (0)
2/50 (4)
JISA (1993)
F344/DuCij rats
Inhalation:
6 h/d,
5 d/wk,
104 wk
0
50 ppm
200 ppm
600 ppm
0
9 ppm
36 ppm
108 ppm
Male
0/50
0/50
0/50
0/50
0/50
0/50
0/50
1/50 (2)
2/50 (4)
1/50 (2)
2/50 (4)
11/50 (22)
14/50 (28)
22/50 (44)
27/50 (54)
47/50 (94)
46/50 (92)
45/50 (90)
48/50 (96)
2/50 (4)
0/50 (0)
0/50 (0)
0/50 (0)
0
50 ppm
200 ppm
600 ppm
0
9 ppm
36 ppm
108 ppm
Female
1/50 (2)
0/50 (0)
1/50 (2)
0/50 (0)
1/50
0/50
0/50
0/50
0/50 (2)
0/50 (0)
0/50 (0)
1/50 (2)
10/50 (20)
17/50 (34)
16/50 (32)
19/50 (38)
N/A
0/50
0/50
1/50
0/50
a -r
>1
to 0\
2 ^
3
¦t*
Ltl
On
~n
H
O
O
aNone reported: Individual animal data were not available, and summary data did not include a line item for this tumor type.
bThese tumors were reported as hemangioendotheliomas in the JISA (19931 report. The term has been updated to hemangioma (benign) or hemangiosarcoma
(malignant). Note that these incidences do not match those tabulated in Table 12 of the JISA report summary. The incidences reported here represent a
tabulation of hemangioendotheliomas from the individual animal data provided in the JISA report.
°Reflects the number of animals with MCL reported under multiple organs," spleen, or liver.
dThis study was inconclusive with respect to carcinogenicity due to a high incidence of respiratory disease in all animals and shortened survival in PCE-exposed
animals.
eGavage doses listed were adjusted several times during the course of the study. Male rats received the listed TWA daily doses through Week 78, and surviving
animals were observed up to study termination in Week 110.
f —Wked tumor, malignant"(NCI. 1977).
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-52. Tumor incidence in mice exposed to tetrachloroethylene
Bioassay
Doses/exposures
Sex
Reported cumulative tumor incidence (%)
Administered
exposure
Continuous
equivalent
exposures
Hepatocellular
adenomas or
carcinomas
Hemangioma
or hemangio-
sarcoma3
Renal
adenomas or
carcinomas
Malignant
lymphoma
Testicular
interstitial cell
tumors
Brain
gliomas
NCI (1977)b
B6C3Fi mice
Gavage:
5 d/wk,
78 wk
Vehicle
450 mg/kg-day
900 mg/kg-day
0
536 mg/kg-day
1,072 mg/kg-
day
Male
2/20 (10)
32/48 (67)
27/45 (60)
None reported0
0/20 (0)
1/49 (2)
0/48 (0)
None
None reported
None reported
Vehicle
300 mg/kg-dayd
600 mg/kg-day
0
386 mg/kg-day
772 mg/kg-day
Female
0/20 (0)
19/48 (40)
19/48(40)
None reported
None
reported
None
N/A
None reported
NTP0986tD
B6C3Fi mice
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
100 ppm
200 ppm
0
18 ppm
36 ppm
Male
17/49 (35)
31/49 (70)
41/50 (82)
1/49 (2)
0/49 (0)
0/50 (0)
0/49 (0)
1/49 (2)
0/50 (0)
None
1/49 (2)
0/48 (0)
0/49 (0)
None
0 ppm
100 ppm
200 ppm
0
18 ppm
36 ppm
Female
4/48 (8)
17/50(38)
38/50 (76)
0/48 (0)
3/50 (6)
0/50 (0)
None
None
N/A
1/48 (2)
0/49 (0)
0/50 (0)
JISA (1993)
Crj:BDFl mice
Inhalation:
6 h/d,
5 d/wk,
104 wk
0 ppm
10 ppm
50 ppm
250 ppm
0
1.8 ppm
9.0 ppm
45 ppm
Male
13/50 (28)
21/50 (43)
19/50 (40)
40/50 (82)
4/50 (4)
2/50 (2)
7/50 (13)
9/50 (18)
0/50
1/50
1/50
0/50
9/50
7/50
7/50
9/50
3750
0/50
0/50
1750
0/50
0/50
0/50
0/50
0 ppm
10 ppm
50 ppm
250 ppm
0
1.8 ppm
9.0 ppm
45 ppm
Female
3/50 (6)
3/47 (6)
7/49 (15)
33/49 (67)
1/50
0/47
2/49
3/49
0/50
0/47
0/49
0/49
14/50
10/47
16/49
10/49
N/A
0/50
0/47
0/49
0/49
a -r
>1
to 0\
2 ^
3
¦t*
Ltl
¦o
~n
H
O
O
a Administered gavage doses listed were increased after 11 wk by 100 mg/kg-day in each low-dose group or by 200 mg/kg-day in each high-dose group. Animals
received the listed TWA daily doses through Week 78, and surviving animals were observed up to study termination in Week 90.
bThese tumors were reported as hemangioendotheliomas in the JISA (19931 report. The term has been updated to hemangioma (benign) or hemangiosarcoma
(malignant). Note that these incidences do not match those tabulated in Table 12 of the JISA report summary. The incidences reported here represent a
tabulation of hemangioendotheliomas from the individual animal data provided in the JISA report.
°None reported: Individual animal data were not available, and summary data did not include a line item for this tumor type,
histiocytic sarcomas, epididymides, or seminal vesicles.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
The Japanese bioassay (JISA. 1993) also reported a significant dose-dependent increase
in MCL in male and female F344/DuCrj rats exposed for 104 weeks to 50-, 200-, and 600-ppm
tetrachloroethylene. MCL latency was decreased in female rats, with the first appearance in
Week 100 in controls and Weeks 66-70 in exposed rats. As in the NTP study, there was a
higher control incidence of MCL (22% in males and 20% in females) than the reported historical
rate of MCL for the Japanese laboratory of 147/1,149 [13%] in males and 147/1,048 [14.0%] in
females.
Additional tumor findings in rats included a significant increase in the NTP bioassay of
two rare tumor types, kidney tumors in males and brain gliomas in both sexes of exposed F344/N
rats. Kidney tumors rarely occur in unexposed F344/N male rats, with historical incidences
reported to be 0.2% in 1968 controls. The reported incidences with 0, 200, or 400 ppm
tetrachloroethylene exposure were 1/49, 3/47, and 4/50, respectively. Additional support for the
significance of the kidney tumors comes from evidence that the related chemical
trichloroethylene induces this tumor type in humans and in male rats (U.S. EPA. 2009b). For
brain gliomas, the laboratory and overall program historical control incidences were 2/247
(0.8%)) and 4/1971 (0.2%), respectively. Reported incidence with 0, 200, or 400 ppm
tetrachloroethylene exposure was 2/50, 0/48, and 4/50 in males and 1/50, 0/50, and 2/50 in
females, respectively. The significance of the brain tumor findings is supported by the earlier
occurrence with tetrachloroethylene exposure, suggesting an effect on latency. In males,
tetrachloroethylene-induced brain tumors were seen beginning at Week 88 compared with Week
99 in controls. Female brain tumors were first seen at 75 weeks in tetrachloroethylene-exposed
animals compared with 104 weeks in control group females.
The NTP (1986b) study also reported an increase in the rate of testicular interstitial cell
tumors, a tumor type of high incidence in unexposed F344 rats. The reported incidences of
testicular interstitial cell tumors in male rates exposed to 0-, 200-, or 400-ppm
tetrachloroethylene were 36/50, 39/49, and 41/50, respectively. A higher incidence (47/50, or
92%) was seen in control rats in the JISA (1993) study than in the NTP (1986b) study. In the
JISA study, exposure to 0-, 50-, 200- or 600-ppm tetrachloroethylene resulted in incidences of
47/50, 46/50, 45/50, and 48/50, respectively. Support for the significance of the testicular
interstitial cell tumors comes from evidence that the related chemical trichloroethylene induces
this tumor type in rats. Trichloroethylene did not induce increases in testicular interstitial cell
tumors in the F344 rat in a bioassay with a reported incidence of 47/48 (98%) in the vehicle
control. However, increases were seen in male Marshall rats, in which the incidences were
16/46, 17/46, 21/33, and 32/39 in the vehicle control and 500, or 1,000 mg/kg-day
trichloroethylene exposure groups, respectively.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-458 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
In conclusion, evidence for the carcinogenicity of tetrachloroethylene in rats was
provided by increases in MCL incidence in both sexes in two inhalation bioassays. Rare kidney
tumors in males and rare brain gliomas in males and females were increased in a single bioassay
(NTP. 1986b). Additionally, the NTP (1986b) bioassay reported increases in the rate of
testicular interstitial cell tumors, a tumor type of high incidence in unexposed male F344 rats.
The available oral gavage cancer bioassay was inconclusive due to respiratory infection in all
groups and high mortality in tetrachloroethylene-exposed animals.
4.10.5.2. Carcinogenicity Findings in Mice
In both sexes of mice, tetrachloroethylene increased the incidence of liver tumors in
multiple bioassays. In male and female B6C3Fi mice exposed for 2 years by oral gavage,
significant increases were noted in hepatocellular carcinomas and adenomas (NCI. 1977). The
reported incidence with 0-, 500-, and 1,000-mg/kg-day tetrachloroethylene were 2/20, 32/48, and
27/45 in males and 0/20, 19/48, and 19/49 in females, respectively. Tumor latency was
significantly decreased with tetrachloroethylene exposure. A significant association between
increased mortality and dose of tetrachloroethylene was seen, with liver tumors found in many of
the mice that died early. In lifetime inhalation studies of B6C3Fi (NTP. 1986b) and Crj:BDFl
mice, tetrachloroethylene similarly increased liver tumors. Statistically significant, dose-related
increases in the incidence of hepatocellular carcinoma and in combined hepatocellular adenoma
and carcinoma were seen in both sexes. The reported incidence of liver carcinomas and
adenomas with 0-, 100-, and 200-ppm tetrachloroethylene in the NTP inhalation bioassay were
17/49, 31/49, and 41/50 in males and 4/45, 17/42, and 38/48 in females, respectively. In male
mice, hepatocellular carcinomas metastasized to the lungs in 2/49, 7/49, and 1/50 animals.
Metastatic hepatocellular carcinomas were found in the lungs of 0/48, 2/50, and 7/50 female
mice. In the JISA study, the reported incidence of liver carcinomas and adenomas with 0-, 10-,
50-, and 250-ppm tetrachloroethylene were 13/50, 21/50, 19/50, and 40/50 in males and 3/50,
3/47, 7/49, and 33/49 in females, respectively.
Additional evidence of carcinogenicity from the lifetime bioassays in mice included a
significant increase in the incidence of hemangiosarcomas (reported as malignant
hemangioendotheliomas) or hemangiomas (reported as benign hemangioendotheliomas) of the
liver, spleen, fat, and subcutaneous skin in males. This tumor type was not reported in the NCI
oral gavage bioassay, and no increase was reported in the NTP inhalation bioassay. Other
findings in the JISA study were Harderian gland adenomas and enlargement of the nucleus in the
kidney proximal tubular cells in male mice at the highest exposure.
Other supporting evidence for carcinogenicity is the known hepatocarcinogenicity of
tetrachloroethylene metabolites. The major urinary metabolite of tetrachloroethylene in humans
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-459 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
and rodents, TCA, is hepatocarcinogenic in mice. TCA significantly increased the incidence of
liver tumors in male and female B6C3Fi mice exposed via drinking water for 52-104 weeks
(Bull et al.. 2002; Bull et al.. 1990; Bull et al.. 2004; DeAngelo et al.. 2008; Herren-Freund et al..
1987; Pereira. 1996; Pereira and Phelps. 1996). Incidence of tumors increased with increasing
TCA concentrations (Bull et al.. 2002; Bull et al.. 1990; DeAngelo et al.. 2008; Pereira. 1996).
The development of tumors in animals exposed to TCA progressed rapidly, as evidenced by
significant numbers of tumors in less-than-lifetime studies of 82 weeks or less. The
tetrachloroethylene metabolite DCA also causes liver cancer in mice (Bull et al.. 1990; Bull et
al.. 2004; Daniel et al.. 1992; DeAngelo etal.. 1999; Herren-Freund et al.. 1987). Additionally,
DCA and TCA are hepatocarcinogenic in mice when coadministered in the drinking water for 52
weeks (Bull et al.. 2004). Treatment-related liver tumors were observed in male F344/N rats
exposed via drinking water to DCA (DeAngelo et al.. 1996) but not TCA (DeAngelo et al.. 1997)
for 60 or 104 weeks. However, the extent to which DCA is available to the liver is unclear,
because it is thought to be formed in the kidney following P-lyase processing of TCVC and may
be largely excreted in urine without circulating systemically. The carcinogenicity of TCA and
DCA has not been evaluated in female rats or in other species of experimental animals.
In conclusion, evidence for the carcinogenicity of tetrachloroethylene in mice is provided
by increases in hepatocellular carcinomas and adenomas in both sexes of mice in a gavage
bioassay (B6C3Fi mice) and in two inhalation bioassays (one of the B6C3Fi strain and the other
of the Crj :BDF1 strain). In male Crj :BDF1 mice, hemangiosarcomas or hemangiomas of the
liver, spleen, fat, and subcutaneous skin were increased (JISA. 1993). Supporting evidence
includes the hepatocarcinogenicity of tetrachloroethylene metabolites TCA and DCA, alone and
in combination.
4.10.5.3. Carcinogenic Mode of Action Hypotheses
This section summarizes the supporting evidence for the modes of action posited for the
rat and mouse tumors presented in Table 4-51 and 4-52. The discussion focuses on
tetrachloroethylene-specific studies, for which the database is especially limited. Evidence from
studies of metabolites of tetrachloroethylene is also summarized. A tabular summary of the
hypothesized MOA and key events, and the supporting evidence from studies of
tetrachloroethylene and its metabolites, are provided in Table 4-56. Overall, these findings
support the conclusion that the mechanisms by which tetrachloroethylene induces rodent
carcinogenesis are not yet fully characterized, completely tested, or understood.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-460 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
4.10.5.3.1. Hypothesized modes of action for rat tumors
4.10.5.3.1.1. Testicular interstitial cell tumors
No data are available concerning either the metabolites or the mechanisms that may
contribute to the induction of testicular interstitial cell tumors occurring in exposed rats.
Evidence for the related compound trichloroethylene, while suggestive of a MOA involving
hormonal disruption, is inadequate to specify and test a hypothesized sequence of key events. It
is concluded that the specific active moiety(ies), mechanisms, or modes of action by which
tetrachloroethylene induces this type of tumor is not known.
4.10.5.3.1.2. Brain gliomas
No data are available concerning either the metabolites or the mechanisms that may
contribute to the induction of rare brain gliomas occurring in exposed rats. It is concluded that
the specific active moiety(ies), mechanisms, or modes of action by which tetrachloroethylene
induces this type of tumor is not known.
4.10.5.3.1.3. Mononuclear cell leukemia
Regarding the metabolites that potentially contribute to MCL development, a role for
GSH-derived intermediates was posited based on findings for the related compound
trichloroethylene. However, TCVC, a GSH-derived metabolite of tetrachloroethylene, induced
no kidney or bone marrow effects when administered to two calves as a single dose (Lock et al..
1996). Aside from this evaluation of bone marrow toxicity of TCVC in the juvenile cow, a
species of unknown sensitivity to tetrachloroethylene-induced leukemia, other studies aimed at
elucidating the active metabolites contributing to leukemic effects have not been reported. In
particular, no such studies are available in the F344 rat, the species and strain in which leukemic
effects have been consistently observed in both sexes. Additionally, no data are available
concerning the contributing mechanisms. It is, thus, concluded that the specific active
moiety(ies), mechanisms, or modes of action by which tetrachloroethylene induces this type of
tumor are not known.
4.10.5.3.1.4. Renal tumors
It is likely that several mechanisms contribute to tetrachloroethylene-induced kidney
cancer. Mutagenicity, peroxisome proliferation, a2[j,-globulin nephropathy, and cytotoxicity not
associated with a2[j,-globulin accumulation are MO As that have been investigated. Except for
a2[j,-globulin accumulation, which is more likely due to tetrachloroethylene itself (Lash and
Parker. 2001). other mechanisms hypothesized to contribute to tetrachloroethylene-induced renal
carcinogenicity are thought to be mediated by tetrachloroethylene metabolites rather than with
the parent compound. Metabolites from the GSH conjugation pathway are posited to induce
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-461 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
renal tumorigenicity, as opposed to, or to a greater extent, than the metabolites resulting from
oxidative CYP processing. The glutathione conjugation of tetrachloroethylene in the kidney,
discussed in Section 3, leads sequentially to TCVG and TCVC. TCVC can be further processed
by P-lyase to yield an unstable thiol, 1,2,2-trichlorovinylthiol, that may give rise to a highly
reactive thioketene, a chemical species that can form covalent adducts with cellular nucleophiles
including DNA. TCVC can also undergo FM03 or P450 oxidation to reactive intermediates;
additionally, sulfoxidation of both TCVC and its A-acetyl ated product occurs, resulting in
reactive metabolites (Ripp et al.. 1999; Ripp et al.. 1997; Werner et al.. 1996). TCVG, TCVC,
and NAcTCVC are mutagenic in Salmonella tests, as is tetrachloroethylene in the few studies of
conditions that could generate GSH-derived metabolites (Dekant et al., 1986)(Dreessen et al..
2003; Vamvakas et al.. 1987; Vamvakas et al.. 1989b; Vamvakas et al.. 1989c). Evidence of in
vivo genotoxicity in the kidney is limited to reports of modest effects following i.p. exposures,
including low level binding to rat kidney DNA (Mazzullo et al.. 1987) and DNA single-strand
breaks in mouse kidney (Walles. 1986). Given the known mutagenicity of the GSH-derived
tetrachloroethylene metabolites that are formed in the kidney, and the observed in vitro
mutagenicity of tetrachloroethylene under conditions that would generate these metabolites, a
mutagenic MOA contributing to the development of the kidney tumors cannot be ruled out.
It has been suggested that the low-level renal tumor production observed in exposed rats
is secondary to sustained cytotoxicity and necrosis leading to activation of repair processes and
cellular regeneration. However, nephrotoxicity occurs in both sexes of rats and mice, whereas
cell replication and tumorigenesis occurs only in male rats. In addition, tetrachloroethylene
induces kidney tumors at lower doses than those required to cause a2[j,-globulin accumulation,
raising serious doubt that a2[j,-globulin plays a key role—especially any major role—in rat
kidney tumor formation. Rodent studies of tetrachloroethylene addressing renal a2[j,-globulin
accumulation are summarized in Table 4-53.
Because tetrachloroethylene has been shown to induce peroxisome proliferation, an
indicator of PPARa-activation, the possibility exists that certain responses resulting from
activation of this receptor might be involved in cancer-causing activity leading to
tetrachloroethylene-induced renal tumors. However, as summarized in Table 4-54, chemical-
specific studies are limited and show only modest effects at exposures exceeding those required
for renal carcinogenesis. There is no evidence causally linking PPARa-activation to kidney
tumorigenesis for tetrachloroethylene or other compounds.
In summary, the complete mechanisms of tetrachloroethylene-induced renal
carcinogenesis are not yet understood. Given the known mutagenicity of the GSH-derived
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-462 DRAFT-DO NOT CITE OR QUOTE
-------
Table 4-53. Renal a2ji-globulin accumulation in tetrachloroethylene-exposed
rodents
Species/strain/
sex/number
Exposure level/du ration
Effects
Reference
Mouse, B6C3Fi, both
sexes (49 or 50 mice per
sex per dose group)
0, 100, 200 ppmfor 104 wk,
inhalation
Karyomegaly and cytomegaly of the
proximal tubules in all exposed mice;
nephrosis in exposed females, casts
increased in all exposed males and in high-
dose females.
NTP (1986b)
Rat, F344, both sexes (50
mice per sex per dose
group)
0, 200, 400 ppmfor 104 wk,
inhalation
Karyomegaly and cytomegaly of the
proximal tubules in all exposed rats.
NTP (1986b)
Rat, F344 (both sexes, 5
per group)
0 or 1,000 mg/kg-day for
10 d, corn oil gavage
Increases in a2(i-hyaline droplets in
exposed males but not females. Correlated
to increased cell proliferation and protein
droplet nephropathy.
Goldsworthy
etal. (1988)
Rat, F344 (both sexes, 12
per group)
0, 500 mg/kg-day daily for 4
wk, corn oil gavage
Increases in a2(i-hyaline accumulation in
proximal tubule cells.
Bergamaschi
etal. (1992)
Rat, F344 (both sexes)
and B6C3Fi mice (both
sexes); 10 per group for
oral studies, 5 per group
for inhalation studies
0, 1,000 or 1,500 mg/kg-day
daily by corn oil gavage for
42 d; 0 or 1,000 ppm for 10 d
Accumulation of a2(i-globulin in proximal
tubules of male rats; nephrotoxicity in male
rats (formation of granular tubular casts and
evidence of tubular cell regeneration).
Inhalation exposure demonstrated
formation of hyaline droplets in kidneys of
male rats.
Green et al.
(1990)
Table 4-54. Renal peroxisome proliferation in tetrachloroethylene-exposed
rodents
Species/strain/sex/number
Effect
Dose
Time
Rat, F344; and mouse,
B6C3Fi; both sexes
(5/group)
Odumetal. (1988b)
Mice of both sexes: Analysis in mice was limited to
pooled tissue, but showed slight increases in
(3-oxidation in mouse kidney
200, and 400
ppm, inhalation
14, 21,28 d
Rats: Modest increases in PCO in male rat kidneys
at 200 ppm for 28 d only, but elevated in female rat
kidney at all doses and times.
200, and 400
ppm, inhalation
14, 21,28 d
Rat, F344 (male only,
5/group) and mouse,
B6C3Fi (male only,
5/group)
Goldsworthy and Popp
(1987)
Mice: Increased PCO activity
1,000
mg/kg-day for
10 d, corn oil
gavage
10 d
Rats: Increased kidney weight
1,000
mg/kg-day for
10 d, corn oil
gavage
10 d
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-463 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
tetrachloroethylene metabolites that are formed in the kidney, and the observed in vitro
mutagenicity of tetrachloroethylene under conditions that would generate these metabolites, a
mutagenic MOA contributing to the development of the kidney tumors cannot be ruled out.
4.10.5.3.2. Hypothesized modes of action for mouse tumors
4.10.5.3.2.1. Hemangiosarcomas
No data are available concerning either the metabolites or the mechanisms that may
contribute to the induction of hemangiosarcomas or hemangiomas observed in the liver, spleen,
fat, and subcutaneous skin in male mice. It is concluded that the mechanisms or modes of action
by which tetrachloroethylene induces this type of tumor are not known.
4.10.5.3.2.2. Hepatocellular tumors
As noted by NRC (2010). it is likely that key events from several pathways, comprising
several simultaneous mechanisms, operate in tetrachloroethylene-induced liver cancer. MOA
hypotheses for mouse liver tumors concern genotoxicity, epigenetic effects (especially DNA
hypomethylation), oxidative stress, and receptor activation (i.e., a hypothesized PPARa-
activation MOA). Because it has been suggested that hepatocarcinogenesis caused through a
PPARa-activation MOA is not relevant to humans (e.g.. Klaunig et al.. 2003). and such a
conclusion would have significant implications for hazard conclusions and dose-response
analyses, this hypothesized MOA is discussed in relatively more detail than other topics.
The limited tetrachloroethylene-specific data for PPARa-activation support the view that
this is not the primary MOA for hepatocarcinogenesis (see Table 4-55). Philip et al. (2007)
reported significantly increased expression of CYP4A, a marker of PPARa-activation, in SW
mice at only the highest dose (1,000 mg/kg-day) and at the earliest time point (7 days), in
contrast to the robust dose-dependent proliferative response of a more prolonged nature (lasting
for 14-30 days post exposure) observed at the same and lower (150, 500, and 1,000 mg/kg-day)
levels of tetrachloroethylene. The authors suggested that these data are not supportive of a close
mechanistic relationship of carcinogenicity and PPARa-activation for tetrachloroethylene-
derived TCA. Limitations of this interpretation include the possible lack of sensitivity of
CYP4A protein expression as a marker of peroxisome proliferation, and the unknown sensitivity
of the SW mouse to tetrachloroethylene hepatocarcinogenicity. Other investigators (e.g..
Schumann etal.. 1980)) have reported liver toxicity and repair at 100 mg/kg-day in the B6C3Fi
strain, whereas repeated exposures to 1,000 mg/kg-day were reported by Philip et al. (2007) and
Odum et al. (1988b) to only modestly increased peroxisomal markers in SW and B6C3Fi mice,
respectively. Odum et al. (1988b) also observed moderate increases in peroxisome proliferation
in rats, a species insensitive to tetrachloroethylene hepatocarcinogenicity. In all, these findings
indicate that the modest peroxisome proliferation observed in response to tetrachloroethylene
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-464 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
may lack specificity with respect to species, tissue, and dose. Studies of the temporal sequence
of events are limited. Given the limitations in the database of tetrachloroethylene-specific
studies, it can be concluded that the few studies demonstrating peroxisome proliferation by
tetrachloroethylene are insufficient to demonstrate a causative role of this effect in the induction
of other key events posited for the PPARa-activation MOA hypothesis, and for
hepatocarcinogenesis by tetrachloroethylene.
Table 4-55. Rodent studies of induction of hepatic peroxisome proliferation
or its markers by tetrachloroethylene
Species/strain/sex/number
Effect
Dose
Time
Rat, F344; and mouse, B6C3F,;
both sexes (5/group)
Odumetal. CI988b)
Mice of both sexes: increased relative
liver weight, centrilobular lipid
accumulation and peroxisome
proliferation; increased PCO
(up to 3.7-fold)
200, and 400 ppm,
inhalation
14, 21,28 d
Male mice: mitochondrial proliferation
400 ppm,
inhalation
28 d
Rats of both sexes: increased PCO (up to
1.3-fold)
200, and 400 ppm,
inhalation
14, 21,28 d
Rat, F344 (male only, 5/group)
and mouse, B6C3F, (male only,
5/group)
Goldsworthv and Podd (1987)
Mice: Increased relative liver weight;
4.3-fold PCO increase
1,000 mg/kg-day
for 10 d, corn oil
gavage
10 d
Rats: Increased relative liver weight;
modest but not significant (1.4-fold) PCO
increase
1,000 mg/kg-day
for 10 d, corn oil
gavage
10 d
Mouse, Swiss-Webster, male (4
mice/group)
Philip et al. (2007)
Increased plasma ALT
150, 500, and 1,000
mg/kg-day,
aqueous gavage
24 hours to 14 d
after initial
exposure
Mild to moderate fatty degeneration and
necrosis, with focal inflammatory cell
infiltration
150, 500, and 1,000
mg/kg-day,
aqueous gavage
24 hours to 30 d
after initial
exposure
Increased mitotic figures and DNA
synthesis
150, 500, and 1,000
mg/kg-day,
aqueous gavage
Peaked on 7 d,
sustained at 14-
30 d
CYP4A increased at 7 but not 14 d, only
at 1,000 mg/kg-day
1,000 mg/kg-day,
aqueous gavage
7 but not 14 d
Studies of other PPARa agonists, and of transgenic models of PPARa-activation, more
generally support the view that the hypothesized PPARa-activation MOA may not be a limiting
factor in rodent hepatocarcinogenesis (see Section 4.3.5.5). PPARa-activation may play a
significant role in mouse liver tumor induction by some compounds, such as Wy-14,643.
However, recent studies suggest that DEHP can induce tumors in a PPARa independent manner
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-465 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
without any loss of potency (Ito et al.. 2007). and that PPARa-activation in hepatocytes is itself
insufficient to cause tumorigenesis (Yang et al.. 2007a). Additional analyses, presented in
Section 4.3.5.3.2, demonstrate that peroxisome proliferation and associated markers are poor
quantitative predictors of hepatocarcinogenesis in rats or mice. These findings raise serious
concerns about human health risk assessment MO A conclusions based exclusively on evidence
of PPARa-agonism and other key events in the hypothesized PPARa-activation MO A, given that
other modes, mechanisms, toxicity pathways, and molecular targets may contribute to or be
required for the observed adverse effects. Indeed, for tetrachloroethylene and most other PPARa
agonists, chemical-specific data to define the range of effects that may contribute to human
carcinogenesis are insufficient. Similarly, the epidemiologic data are inadequate to inform
conclusions of human relevance (Guvton et al.. 2009).
A recent review (Rusvn et al.. 2006) addressed other mechanistic effects of the PPARa
agonist DEHP and proposed that tumors arise from a combination of molecular signals and
pathways, rather than from a single event such as PPARa-activation. As reviewed in
Section 4.3.5.1, the metabolites of tetrachloroethylene have been shown to induce a number of
effects that may contribute to carcinogenicity, including mutagenicity, alterations in DNA
methylation, and oxidative stress. Given the demonstrated mutagenicity of several
tetrachloroethylene metabolites, the hypothesis that mutagenicity contributes to the MOA for
tetrachloroethylene hepatocarcinogenesis cannot be ruled out, although the specific metabolic
species or mechanistic effects are not known. Epigenetic effects and oxidative stress, including
those produced secondary to cytotoxicity, may also contribute. Currently, the available database
of tetrachloroethylene-specific studies addressing these mechanisms is very limited.
4.10.5.3.3. Mode-of-action summary
Table 4-56 reviews the hypothesized modes of action for tetrachloroethylene-induced
cancer in rodents, which are not intended to be interpreted as being mutually exclusive. The
evidence summarized in this table supports the view that there are significant gaps in the
scientific knowledge of mechanisms contributing to tetrachloroethylene-induced cancer.
Multiple metabolites formed from tetrachloroethylene are toxic and carcinogenic in rodents.
Given this knowledge, and the known complexity and heterogeneity in cancer development, in
general, the available evidence supports a hypothesis of multiple, contributing mechanistic
effects that may, in turn, be affected by multiple modifying factors.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/2011 4-466 DRAFT-DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-56. Summary of hypothesized modes of action for tetrachloroethylene-induced cancer in rodents
(continued)
Tumor type, sex, strain,
species
Hypothesized MOA and key events
Evidence that PCE or PCE metabolites
induces key events
Necessity of MOAs
key events for
carcinogenesis
Sufficiency of MOA
for carcinogenesis
Kidney adenocarcinoma in
male F344/N rats (continued)
Tubular cell necrosis and
nephrotoxicity followed by
hyperplasia
Nephrotoxicity of PCE reported in multiple
studies in both sexes of rats and mice at
carcinoeenic doses (e.e.. NTP (1986b))
No PCE-specific
studies3
No PCE-specific
studies
a2(i-globulin accumulation:
• Excessive accumulation of hyaline
droplets containing «2|i-globulin
in renal proximal tubules
• Subsequent cytotoxicity and
necrosis
• Sustained regenerative tubule cell
proliferation
• Development of intralumenal
granular casts from sloughed
cellular debris associated with
tubule dilatation and papillary
mineralization
• Foci of tubule hyperplasia in the
convoluted proximal tubules
• Renal tubule tumors
In F344 rats, PCE induced hyaline droplets at
500 me/ke-dav for 4 wk (Bereamaschi et al..
1992). or >1.000 me/ke-dav for 10
(Goldsworthv et al.. 1988) or 42 d (Green et
al.. 1990s)
No evidence of mineralization in PCE
bioassavs (JISA. 1993; NTP. 1986b) or of
hyaline droplets with <400 ppm for 28 d
(Green et al.. 1990s) in F344 rats
No PCE-specific
studies3
No PCE-specific
studies
PPARa-activation:
• Metabolites (e.g., TCA) activate
PPARa
• Alterations in cell proliferation
and apoptosis
• Clonal expansion of initiated cells
In F344 rat kidney, PCE increased PCO in
males only at 200 ppm for 28 d (PCO
increased in females at 200 and 400 ppm, at
14. 21 and 28 d) (Odum et al.. 1988a): in
B6C3Fi male mouse kidney, PCE increased
PCO with 1,000 mg/kg-day p.o. for 10 d
(Goldsworthv and Pood. 1987)
No PCE-specific
studies
No data from other
chemicals on PPARa
involvement in
kidney tumors.
No PCE-specific
studies
Hemangiosarcomas in male
Crj:BDFi mice
None hypothesized
N/A
N/A
N/A
a -r
>1
to 0\
2 ^
3
¦t*
On
^1
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-56. Summary of hypothesized modes of action for tetrachloroethylene-induced cancer in rodents
(continued)
Tumor type, sex, strain,
species
Hypothesized MOA and key events
Evidence that PCE or PCE metabolites
induces key events
Necessity of MOAs
key events for
carcinogenesis
Sufficiency of MOA
for carcinogenesis
a -r
>1
N> 0\
2 ^
0
1
On
00
Liver hepatocellular carcinoma
in male and female B6C3Fi
and Crj:BDFi mice
Mutagenicity induced by one or more
metabolites advances acquisition of
multiple critical traits contributing to
carcinogenesis
PCE lacks mutagenicity in Salmonella
(Ames), other genotoxicity tests (Bartsch et
al.. 1979: Connor etal.. 1985: DeMarini et
al.. 1994: Emmert et al.. 2006: Greim et al..
Hardin etal.. 1981: Haworth et al..
Kringstad et al.. 1981: Milman et al..
NTP. 1986b: Roldan-Ariona et al..
Shimada et al.. 1985: Warner et al..
Watanabe et al.. 1998[see Table 4-4011
1975
1983
1988
1991
1988
Limited PCE genotoxicity studies in mouse
liver: Positive/equivocal Comet assay in CD1
mice (Cederberg et al.. 20101. positive
micronucleus assay in ddY mice post (but
not pre) partial hepatectomy (Murakami and
Horikawa. 1995) at 1,000 mg/kg-day; DNA
binding in male Balb/c mice at 1.4 mg/kg i.p.
(Mazzullo et al.. 1987): DNA single-strand
breaks in NMRI mice with 660 mg/kg i.p.
rWalles. 19861
Certain metabolites of PCE (e.g., DCA) are
mutagenic in vitro and in vivo (see Tables
4-41 and 4-42)
No PCE-specific
studies3
No PCE-specific
studies;
Mutagenicity is
assumed to cause
cancer, as a
sufficient cause
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 4-56. Summary of hypothesized modes of action for tetrachloroethylene-induced cancer in rodents
(continued)
Tumor type, sex, strain,
species
Hypothesized MOA and key events
Evidence that PCE or PCE metabolites
induces key events
Necessity of MOAs
key events for
carcinogenesis
Sufficiency of MOA
for carcinogenesis
Epigenetic changes, particularly
DNA methylation, induced by one or
more metabolites (TCA, DCA, and
other reactive species) advance
acquisition of multiple critical traits
contributing to carcinogenesis
No PCE-specific studies
In mouse liver, TCA and DCA decrease
global DNA methylation and promoter
hvDomcthvlation (e.e.. of c-mvc) (Ge et al..
2001b: Taoetal.. 1998)
No PCE-specific
studies3
No PCE-specific
studies; dys-
regulation of
methylation
represents a common
early molecular
event in most tumors
and is hypothesized
to cause cancer
Liver hepatocellular carcinoma
in male and female B6C3Fi
and Cij:BDFi mice (continued)
Cytotoxicity and secondary oxidative
stress:
• One or more reactive
intermediates induce
hepatotoxicity
• Oxidative stress results (from
hepatocyte injury, from infiltrating
inflammatory cells and/or as part
of the intra- and/or intercellular
repair processes)
• Oxidative stress advances
acquisition of multiple critical
traits contributing to
carcinogenesis
PCE induces hepatotoxicity characterized by
increased liver weight, fatty changes,
necrosis, inflammatory cell infiltration, and
proliferation (e.e.. NTP. 1986b)
No PCE-specific
studies3
No PCE-specific
studies
a -r
>1
to 0\
2 ^
3
¦t*
On
VO
~n
H
O
O
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
s
>!
§•
rs
st
5
>!
Si
Table 4-56. Summary of hypothesized modes of action for tetrachloroethylene-induced cancer in rodents
(continued)
Tumor type, sex, strain,
species
Hypothesized MOA and key events
Evidence that PCE or PCE metabolites
induces key events
Necessity of MOAs
key events for
carcinogenesis
Sufficiency of MOA
for carcinogenesis
a -r
>1
N> 0\
2 ^
0
1
PPARa-activation:
• TCA, after being produced in the
liver, activates PPARa
• Alterations in cell proliferation
and apoptosis
• Clonal expansion of initiated cells
In B6C3Fi mouse liver, PCE increased PCO
(three- to fourfold) with 200 and 400 ppm
(OdumetaL 1988a) or 1,000 mg/kg-day p.o.
(Goldsworthv andPopp. 1987)
In SW mouse liver, PCE increased CYP4A at
7 but not 14 d, at 1,000 mg/kg-day; increased
mitotic figures and DNA synthesis at 7-30 d
with 150, 500, and 1,000 mg/kg-day (Philip
etal.. 20071
TCA activates PPARa, induces peroxisome
proliferation and hepatocyte proliferation in
mice and rats (e.g.. DeAngelo et al.. 2008:
Dees and Travis. 1994: Laughter et al.. 2004:
Pereira and Phelps. 1996: Sanchez and Bull.
1990: Stauber and Bull. 19971
No PCE-specific
studies; liver tumor
response from WY
dramatically
diminished in
PPARa-null mice
(Peters et al.. 19971:
liver tumor response
from DEHP
unchanged in
PPARa-null mice
ato et al.. 20071.
No inference
possible with PCE.
No PCE-specific
studies; PPARa-
activation in a
transgenic mouse
model caused all the
key events in the
MOA, but not
carcinogenesis,
suggesting that the
MOA is not
sufficient for
carcinogenesis
(Yang et al.. 2007a1.
Consistent with
hypothesis that PCE
liver carcinogenesis
involves multiple
mechanisms.
a Associations (e.g., per Hill (Hill. 1965) considerations) noted for some chemicals between hypothesized sequence of key events and carcinogenesis.
o
~n
H
O
O
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
5. DOSE-RESPONSE EVALUATION
5 1 INHALATION REFERENCE CONCENTRATION (RfC)
This section presents quantitative risk estimates for chronic noncancer inhalation
tetrachloroethylene exposure. Although the RfD is commonly presented first in the IRIS
toxicological reviews, the RfC is presented in Section 5.1 and the RfD in Section 5.2 because the
RfD was developed by route-to-route extrapolation of the RfC to the oral route of exposure. The
analysis is based on the noncancer hazard characterization for tetrachloroethylene presented in
Section 4.10.2, which identified neurotoxicity as a sensitive endpoint following either inhalation
or oral exposure to tetrachloroethylene. Neurotoxicity is thus selected as the critical effect for
deriving the noncancer inhalation RfC. All neurotoxicity studies suitable for dose-response
analysis are evaluated in the selection of principal studies.
5.1.1. Choice of Candidate Studies and Critical Effect
5.1.1.1. Choice of Critical Effect
The database of human and animal studies on inhalation toxicity of tetrachloroethylene is
adequate to support derivation of inhalation reference values. As summarized in Section 4.10, a
number of targets of toxicity from chronic exposure to tetrachloroethylene have been identified
in published animal and human studies. These targets include the central nervous system (CNS),
kidney, liver, immune, and hematologic system, and development and reproduction. In general,
neurological effects were judged to be associated with lower tetrachloroethylene concentrations
compared with other noncancer endpoints of toxicity.
5.1.1.2. Overview of Candidate Principal Studies
The evidence for neurotoxicity in humans includes controlled experimental chamber
(Altmann et al.. 1990; Hake and Stewart. 1977) and epidemiologic (Altmann etal.. 1995;
Echeverria et al.. 1995; Ferroni et al.. 1992; Hake and Stewart. 1977; Seeber. 1989; Spinatonda
et al.. 1997) studies that used standardized neurobehavioral batteries or employed assessment of
visual function (Cavalleri et al.. 1994; Gobba et al.. 1998; NYSDOH. 2010; Schreiber et al..
2002; Storm et al.. In Press), a neurological outcome known to be sensitive to volatile organic
compounds. Of the 12 candidate studies in humans, seven epidemiological studies of
tetrachloroethylene examined occupational exposure (Cavalleri etal.. 1994; Echeverria et al..
1995; Ferroni et al.. 1992; Gobba et al.. 1998; Schreiber et al.. 2002; Seeber. 1989; Spinatonda et
al.. 1997) three epidemiological studies examined residential exposure to tetrachloroethylene
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-1 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
(Altmann et al.. 1995; NYSDOH. 2010; Schreiber et al.. 2002; Storm et al.. In Press) and two
were acute experimental chamber studies (Altmann et al.. 1990; Hake and Stewart. 1977).
Together, the epidemiologic evidence supports an inference of a broad range of cognitive, motor,
behavioral, and visual functional deficits following tetrachloroethylene exposure (U.S. EPA.
2004).
The research in animal models comprises acute and subchronic studies of the effects of
tetrachloroethylene on functional neurological endpoints (functional observation battery, motor
activity) (Kiellstrand etal.. 1985; Oshiro et al.. 2008). on sensory system function as assessed by
evoked potential (Boves et al.. 2009; Mattsson et al.. 1998; U.S. EPA. 1998) or pathological
changes in the brain (Wang et al.. 1993). The studies in animal models support the human
studies, with notable effects on motor activity and motor function following exposure to
tetrachloroethylene during either adulthood or the developmental period. Changes in evoked
potentials following acute and subchronic exposures were also seen. In addition, postmortem
effects in animals were observed with pathological alterations in brain DNA, RNA, or protein
levels and brain weight changes.
The studies considered for derivation of the RfC are summarized in the following
sections and in Table 5-1 and Figure 5-1. Table 5-1 identifies the species, exposure duration,
and ambient (experimental) concentrations. For epidemiologic studies, the reported
concentrations, and the observed effect and its magnitude associated with the NOAEL or the
LOAEL are provided. Additionally, human equivalent concentrations (HECs) for LOAELs or
NOAELs are presented to better allow examination of effect levels across studies and species.
HECs are calculated using the RfC methodology for a Category 3 gas, extrathoracic effects, and
adjusted to equivalent continuous exposure (U.S. EPA. 1994).1 The studies in Table 5-1 are
listed in order of increasing HEC, and displayed graphically in Figure 5-1.
5.1.1.3. Selection of Principal Studies
The candidate principal studies of CNS effects listed in Table 5-1 were evaluated
according to study characteristics identified in Table 5-2. Human studies were preferred to
animal studies, as were studies of chronic duration. Certain human studies are considered as
more methodologically sound based on study quality attributes identified in Table 5-2 and are
preferred for supporting an RfC. The sections below summarize the evaluation of study.
1 NOAEL* [hec] = NOAEL* [ADJ] (ppm) x {Hb/^)A/Hb/g)H> where, NOAEL* [Hec] = the NOAEL or analogous effect
level such as the benchmark concentration (BMC), NOAEL*[ADj] = the NOAEL or analogous effect level adjusted
for duration of experimental regimen; experimental exposure times duration (number of hours exposed/24 hours)
times week (number of days of exposure/7 days), and (Hb/g)A/Hb/g)H = the ratio of the blood/gas (air) partition
coefficient of the chemical for the laboratory animal species to the human value. The value of one is used for the
ratio if (Hb/g)A>Hb/g)H
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-2 DRAFT—DO NOT CITE OR QUOTE
-------
_ >g
o
0 ^
1 ^
& Si
^ &
k ^
^ >!
»9 O
^ S
¦1 ?».
o
3:
>!
§•
Cb
ss
S
>!
Table 5-1. Neurotoxicological inhalation studies considered in the development of an RfC
Study
Species
Duration
NOAEL/LQAELa
PPm
Effect
(effect magnitude)
at LOAEL
Human equivalent continuous
concentrationsb
(NOAEL/LOAEL)
NOAEL /
LOAEL
ppm
mg/m3
NYSDOH (1978;
Storm et al.. In Press)
Human
10 yr (mean),
continuous
0 002. 0 05 (children)
0.002. 0.07 (adults)
Visual contrast sensitivity (6%
f in children)
NOAEL
0.002
0.01
Schreiber et al. (1976)
Human
4 yrs (mean),
occupational
0.3 (daycare workers,
mean and median)
Visual contrast sensitivity0
LOAEL
0.1
0.7
Schreiber et al. (1976)
Human
5.8 yr (mean),
continuous
0.1 (residents, median
and mean), maybe as
high as 0.4 (mean) and
0.3 (median)
Visual contrast sensitivity0
LOAEL
0.4d
3d
Altmann et al. (1981)
Human
10.6 yr (median)
continuous
0.7 (mean)
0.2 (median)
Cognitive function (14% f),
reaction time (15%-20 |)
visual memory (15% J,)
LOAEL
0.7
5
Cavalleri et al. (2007);
Gobba et al. (1982)
Human
8.8 yr (mean),
occupational
|L(Cavalleri et al.,1994)
Dyschromatopsia (color
vision) (6% f )d
LOAEL
2
15
Spinatonda et al.
(2005)
Human
Inhalation (no
duration
information),
occupational
£ (median)
Reaction time (15% t)
LOAEL
3
19
a -r
§.<§>
N> G\
2 S3
•5
LtJ
~n
H
O
O
-------
_ >g
J $ ki
o i K'
! §•
i? s o
a- i-S «
£ o s
^ & 2
fc. £ a
<$ 2 5"
3 -3" Q
O Sl
^s a
o ^
Table 5-1. Neurotoxicological inhalation studies considered in the development of an RfC (continued)
Study
Species
Duration
NOAEL/IOAFl;'
PPm
Effect
(effect magnitude)
at LOAEL
Human equivalent continuous
concentrationsb
(NOAEL/LOAEL)
Seeber(1989)
Human
>10 yr (mean),
occupational
12, 53
Visuospatial function and
information processing speed
(5-30% change depending on
subtest)
LOAEL
4
29
Ferroni et al. (1992)
Human
10.6 yr (mean),
occupational
H
Reaction time (10% f),
continuous performance (7-11%
1)
LOAEL
5
36
Echeverria et al.
(1995)
Human
15 yr (high-exposure
group; mean),
occupational
11,21,41
Cognitive and visuospatial
measures (4-14% change
depending on subtest)
LOAEL
8
56
Altmann et al. (1990)
Human
4 hr/d for 4 d
10,5fl
Visual-evoked potentials (2-3
ms t)
NOAEL
4
24
Mattsson et al.
(1998)
Rat
Subchronic (13 wk)
6hrs/d, 5d/wk
0, 50, 200, Ml
Flash-evoked potential
(3 ms |)
NOAEL
36
242
Rosengren et al.
(1986)
Gerbil
Subchronic (12 wk,
with 16-wk follow-
up) continuous
0, M, 300
Brain: protein, DNA
concentration (10-15% change
depending on brain region there
were both f and J,)
LOAEL
60
408
Kjellstrand et al.
(1985)
Mouse
60 min
0, 90, 320, 400, 600,
800, 1,200, 1,800,
3,600
Increased locomotor activity
(20% t)
LOAEL
90e
6,102e
Boves et al. (2009)
Rat
90 min
250. 500, 1,000
Impairment in steady state
visual-evoked potential (10% J,)
LOAEL
250e
l,695e
120 min
1,000, 2,000, 3,000,
4,000
Impairment in steady state
visual-evoked potential (20% J,)
LOAEL
l,000e
6,780e
rs S4 O5
>S
Zri
K> G\
2 S3
•_3 O
N ^
-k
~n
H
O
O
-------
_ >g
o
0 ^
1 ^
& Si
^ &
k ^
^ >!
»9 O
^ S
¦1 ?».
o
3:
>!
§•
Cb
ss
S
>!
Table 5-1. Neurotoxicological inhalation studies considered in the development of an RfC (continued)
Study
Species
Duration
NOAEL/LQAEL'
ppm
Effect
(effect magnitude)
at LOAEL
Human equivalent continuous
concentrationsb
(NOAEL/LOAEL)
Wane et al. (1993)
Rat
Subchronic (12 wk)
continuous
0. 300. 600
Reduced brain weight
(10.10 g),DNA
(J.0.05-0.06 mg), protein
(12.5-3.5 mg)
NOAEL
300e
2,034e
Oshiro et al. (2008)
Rat
60 min
500. 1,000, 1,500
False alarms (10% f)
LOAEL
500e
3,390e
500, 1,000, 1,500
Reaction Time (200 ms f)
NOAEL
500e
3,390e
a -r
N> G\
2 S3
•5
Note: Principal studies shaded in blue. 1 ppm = 6.78 mg/m3.
aExperimental/observational NOAEL is underlined, LOAEL is double-underlined.
Calculated using RfC methodology for a Category 3 gas, extrathoracic effects, and adjusted to equivalent continuous exposure. Occupational exposures were
multiplied by 5/7(d) x 10/20 (m3/d, breatliing rate) and experimental exposure were multiplied by hours exposed/24 (lir) x 5/7(d).
°Effect magnitude could not be determined from information in published paper.
dAtmospheric monitoring indicated slightly higher exposure levels were experienced by subjects. Schreiber et al. (1976) found mean tetrachloroethylene
concentrations of 0.2 ppm (0.09 ppm, median) of four families living in apartments above active dry cleaning and two families living in an apartment building
where dry cleaning had ceased 1 month earlier. Ambient monitoring of these six apartments during a period of active dry cleaning indicated exposure to higher
concentrations, mean = 0.4 ppm (median 0.2 ppm) and is used as the LOAEL for this study.
eHECs are the human equivalents for the same duration as in the experiments, not adjusted to continuous daily exposures.
~n
H
O
O
-------
_ >g
o
0 ^
1 ^
& Si
^ &
k ^
^ >!
»9 O
^ S
¦1 ?».
s
>!
§•
Cb
ss
S
>!
0
1
§.<§>
>!
N> G\
2 ^
On
~n
H
O
O
NYSDOH (2010) —Visual contrast sensitivity; children, residential
0.0]fiig/m3
Schreiber et al (2002)—Visual contrast sensitivity; adult, occupational
o
0.7 mg/m3
Schreiber et a1 (2002) — Visual contrast sensitivity; adult, residential
3 mg/m3
Altmann et al (1995)— Simple reaction time; adults, residential
V o
5 mg/m3
Cavalleri et al (1994^: Gobba et al (1998^ —Color confusion: adult, occupational
F=Oh
15 mg/m3
POD
NOAEL - •
LOAEL-O
Acute Exposure Studies - '
Principal studies are underlined and shaded + 2SE
HUMAN CHRONIC STUDIES
HUMAN ACUTE STUDIES
Spinatonda et al (1997) — Reaction Time; adult, occupational
19^ng/m3
Seeber (1989^—Visuospatial function: adult, occupational
hOH
^ 29 mg/m3
Ferroni et al (1992) —Reaction time, continuous performance; adult, occupational
^ 36 mg/m3
Echeverria et al (1995^— Cognitive and visuospatial measures: adult, occupational
I—O—I
56 mg/m3
Altmann et al (1990^qVEPs; humans, 4 days*
24 mg/m3
Hake and Stewart (1977)-^EEGs, humans, 5 days*#
ANIMAL STUDIES
Mattsson et al (199^- FEP; rat, 12 wks
242 mg/m3
Rosengren et al (1986)— Brain weight; gerbil,3 mths
408 n?«/m3
Kjellstrand et al (1985) — Locomotor activity; mice, 60 min*#
6102^ng/m3
Boyeset al (2009)-J^EP; rats, 90 min*
169ivmg/m3
Boyeset al (2009)-rats, 120min*
6780 mg/m3
Wanget al (1993)—Brain weight; rat, 12 wks
2 03 4 mg/m3
Oshiro et al (2008) — Fa^ alarms; rats, 60 min*
3390 mg/m3
Oshiro et al (2008)—Re^tion time; rats, 60 min*
3390 mg/m3
0.001
0.01
0.1
1 10
[PERC] mg/m3
100
1000
10000
Figure 5-1. Exposure-response array for neurotoxicological inhalation studies considered for RfC development
(listed in Table 5-1). PODs (HEC for LOAELs and NOAELs) are displayed and labeled by study, effect, and
duration. Principal studies selected for RfC derivation are shaded in blue and the POD range (± 2SE) are presented.
-------
Table 5-2. Summary of rationale for identifying studies on
tetrachloroethylene for RfC development
Consideration
Data characteristics
Decision context
Species studied
Animal and human
neurotoxicity studies
Human data are preferred to reduce interspecies extrapolation
uncertainties. Animal data are considered as supporting studies
when adequate human studies are available.
Relevance of
exposure paradigm
scenario
Acute, subchronic
and chronic exposure
durations
Peak and chronic
exposure intensities
Subchronic or chronic studies, if adequate, are preferred over studies
of acute exposure durations.
Studies of residential exposures, if available and of adequate quality
are preferred. In residential settings, exposure is more likely to be
continuous, and of lower concentrations compared with the more
intermittent, higher concentration exposure experienced in work
settings. The potential influence of peak or intensity concentrations
is more common with occupational than residential exposures.
Study quality attributes for human toxicity studies
Study populations
Comparability of
referent and exposed
groups
Referent and exposed groups were evaluated and compared. In
addition to age, potential confounders for neurobehavioral measures
include education, lifestyle factors such as alcohol consumption and
SES are controlled for to limit selection bias and confounding.
Use of a study design (e.g., matching procedures) or analysis
(procedures for statistical adjustment) that adequately addresses the
relevant sources of potential confounding for a given outcome adds
weight to the consideration of the study as principal rather than
supportive.
Measurement of
exposure
Area or individual
measures of exposure
Stronger studies have exposure estimates which are supported by
ambient monitoring and/or biological monitoring. Measurement or
assignment of exposure should not be influenced by knowledge of
results of tests of neurobehavioral function. Higher quality
assessment strategies in occupational studies are based on
assignment of exposure potential to individual subjects considering
individual job titles and tasks with consideration of changes over
time.
Use of higher quality assessment strategies adds weight to the
consideration of the study as principal rather than supportive.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-7 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-2. Summary of rationale for identifying studies on
tetrachloroethylene for RfC development (continued)
Consideration
Data characteristics
Decision context
Measurement of
effect(s)
Standardized
neurological tests:
validity and
reliability
Neurobehavioral function (reaction time measures, cognitive
function, and motor activity) assessed using a standardized test
battery (e.g., Neurobehavioral Evaluation System) is preferred,
because wide administration to occupational populations in different
settings has resulted in a high degree of validity with context of
potential population norms. WHO and ATSDR recommend these
test methods to evaluate nervous system deficits in adults and
children. Other standardized methods were used to evaluate color
vision and visual contrast sensitivity.
Administration or interpretation of the test should not be influenced
by knowledge of exposure status adds weight to the consideration of
the study as principal rather than supportive.
Use of standardized neurological tests and use of sensitive methods
to detect neurological changes adds weight to the consideration of
the study as principal rather than supportive.
Study quality attributes for animal toxicity studies
Study populations
Comparability of
animal models to
effects observed in
humans
Studies in animal models reporting effects concordant to observed
solvent associated effects in humans were considered preferable.
Measurement of
effect(s)
Validity and
comparability of
neurological tests
Neurological tests and methods that have been validated in animal
models were preferred. Endpoints in animals that were concordant
or comparable with evaluated endpoints in humans were the most
preferred.
1 characteristics, presenting human studies by exposure paradigm (residential, occupation, and
2 controlled exposure), followed by animal studies.
5.1.1.3.1. Evaluation of epidemiologic studies of residentially exposed populations
3 Three epidemiological studies of residential exposures were examined as candidate
4 principal studies for deriving a RfC (Altmann et al.. 1995; NYSDOH. 2010; Schreiber et al..
5 2002) and Storm et al. (In Press). As outlined in Table 5-2, residential exposures come closest to
6 the chronic, continuous exposures addressed by reference values. The exposed populations in
7 these studies lived in buildings colocated with dry cleaners. Additional strengths of all of these
8 studies included high quality exposure assessment, matching of controls by age and sex, and use
9 of standardized testing. In addition, statistical analyses adjusted for race/ethnicity, age, and other
10 covariates such as smoking or alcohol use. On the other hand, there were differences in
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-8 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
comparability between referent and exposed groups in each of these studies for which statistical
analyses could not sufficiently adjust, limiting their use as principal studies. The studies were
described in detail in Section 4; study-specific issues relevant to principal study selection are
summarized below.
The NYSDOH pilot study (Schreiber et al.. 2002) reported deficits in visual contrast
sensitivity (VCS) in residents exposed to tetrachloroethylene compared to controls. Schreiber
et al. (2002) evaluated 17 exposed subjects, including four children (in New York City) and
17 control subjects (recruited from among NYSDOH employees living in Albany, NY) and
reported reduced group-mean visual contrast sensitivity scores in residents compared to
unexposed referents at a human equivalent LOAEL (LOAELrec) of 3 mg/m3 (arithmetic mean
concentration). A key limitation of this study, in addition to its small sample size and potential
for selection bias owing to health department employees being referents for exposed residents,
was that vision testing was not blinded to exposure classification.
NYSDOH (2010) and Storm et al. (In Press) is a larger study of 104 exposed adult and
children residents of 24 buildings with colocated dry cleaners using tetrachloroethylene and
101 unexposed adults and children in 36 buildings without colocated dry cleaners. High quality
exposure assessment addressed some of the concerns of selection bias in the previous study of
Schreiber et al. (2002); for example, the study employed a larger number of subjects and
referents from the same geographical area. Additionally, exposure and effects were assessed in
family units, allowing comparison of parents and children in the same household. Storm et al.
(In Press) identified a human equivalent NOAEL (NOAELrec) of 0.01 mg/m3 (median
concentration) in children and a NOAELrec of 0.48 mg/m3 (median concentration) in adults.
However there are other concerns as to the comparability of referent and exposed subjects.
Those living in households with higher levels of tetrachloroethylene were more likely to be of
minority race and of lower income status compared to referent families. Additionally, exposed
subjects were younger (p < 0.05) and of lower educational attainment (p < 0.05) than those in
referent buildings. Another concern is that, although a standardized visual test (Functional
Acuity Contrast Test [FACT]) was used, it was of far distance VCS only. The test was also less
sensitive than that employed in other studies because the response was scored as either maximum
(perfect) or less than maximum, with no gradations of reduced response. Statistical analyses
appropriately examined the association between these exposure metrics and vision and adjusted
for a number of relevant covariates. However, the small number of nonminority and high
income subjects in the highest tetrachloroethylene exposure group, and the lower mean education
level of the high exposure group, limit conclusions that observed effects were completely
independent of education level, race/ethnicity or income. This raises concerns about the
comparability between exposed and referent subjects. Consequently, due to ceiling effect of the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-9 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
testing method and potential confounding of education level, race/ethnicity or income, NYSDOH
(2010) and Storm et al. (In Press) were not selected as principal studies.
Altmann et al. (1995) reported visuospatial and cognitive deficits (from two tests of
simple reaction time, continuous performance and visual memory) among 19 residents compared
to 30 unexposed referents at a LOAELrec of 5 mg/m3 (arithmetic mean concentration).
Statistical analyses appropriately adjusted for covariates and possible confounders of age,
gender, and education in logistic regression models; however, the paper lacked reporting of
logistic regression coefficients and effect magnitudes, limiting a clear assessment of the effects
observed. Furthermore, the referent group in Altmann et al. (1995) had a higher educational
attainment than tetrachloroethylene-exposed subjects. Altmann et al. (1995) adjusted for a
potential effect of education, a surrogate for socioeconomic status, on visuospatial test
performance in multiple regression models. However, the National Research Council (NRC.
2010) noted potential for residual confounding, as education was examined as a categorical, not
continuous, variable using three groups which might affect interpretation of cognitive testing of
continuous performance and visual memory. Nonetheless, effects of tetrachloroethylene
exposure were seen on reaction time, an endpoint that is not influenced by education level.
There was potential bias in subject selection: 19 of 95 potentially eligible subjects participated in
the study and the study did not identify reasons for excluding the remaining 76 subjects.
Altmann et al. (1995) was not selected as a principal study given the limited reporting, concerns
for potential selection bias and concern about residual confounding for some of the adverse
outcomes observed.
In sum, none of these residential studies was selected as a principal study. These studies
nonetheless provide qualitative evidence for hazard identification of neurological deficits in
visual function, reaction time, and cognitive function. The database of residential studies also
adds support for the choice of key endpoints in principal studies, and informs uncertainty factor
(UF) selection, as described in Section 5.1.3.
5.1.1.3.2. Evaluation of epidemiologic studies of occupationally exposed populations
Seven occupational studies assessed visual function or other neurobehavioral effects and
were considered as candidate studies for deriving the RfC (Cavalleri et al.. 1994; Echeverria et
al.. 1995; Ferroni et al.. 1992; Gobba et al.. 1998; Schreiber et al.. 2002; Seeber. 1989;
Spinatonda et al.. 1997). The primary strength of each of these studies is their use of
standardized tests methodology to evaluate neurobehavioral or visual function. Additional
details regarding the evaluation of occupational study characteristics that informed selection of
candidate studies are provided below.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-10 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Ferroni et al. (1992) was a prevalence study of 60 female dry cleaners or other
dry-cleaning workers and 30 sex-, age-, and vocabulary test score-matched controls from an
industrial cleaning plant that did not use organic solvents. Compared to responses in referents,
dry cleaners had a 10% increased simple reaction time and decrements in response on
two subtests of the shape comparison test, one of vigilance (7% decrease) and one of stress
(11% decrease) at the LOAEL of 102 mg/m3 [LOAELrec = 36 mg/m3] (median concentration).
Study details are sparsely reported, and results are not accurately reported in the published paper.
Ferroni et al. (1992) does not clearly identify whether age-matching was for individual subjects,
or for the group's average age. A crude exposure assessment was used based on ambient
monitoring data assigned to the group of dry cleaners and statistical analyses did not control
adequately for confounding characteristics among participants. As compared to the other
occupational studies, this study had poorer quality in terms of comparability of referent and
exposed groups and measurement of exposure and analysis methods, in part because of poor
reporting of study details and results, and therefore was not selected as a principal study.
Spinatonda et al. (1997) was a prevalence study of 35 dry cleaners and 39 age- and
education-matched unexposed subjects that reported a 15% increased latency to a vocal response
time at a LOAEL of 54 mg/m3 [LOAELrec =19 mg/m3] (median concentration). The study
design is sparsely reported and the paper lacks details of subject selection, including the
population from which controls were drawn, and demographic information for evaluation of
comparability of dry cleaners and controls. Exposure was assessed by a -grab sample" that is
inferior to a time weighted average estimate. The study developed an index of cumulative
exposure to tetrachloroethylene for each exposed subject by multiplying the tetrachloroethylene
concentration by the number of years worked. Statistical analyses comprised Wests comparing
average latency in dry cleaner and control groups, and regression models fit to responses of
exposed subjects only, a weaker approach than fitting multiple logistic regression models to data
from all subjects. Additionally, the statistical analyses did not control for alcohol consumption,
which is also associated with response time, indicating a greater potential for confounding. As
compared to the other occupational studies, this study had poorer quality in terms of
comparability of referent and exposed groups and measurement of exposure, in part because of
poor reporting of study details and results, as well as less robust statistical analyses controlling
for alcohol consumption. Therefore, Spinatonda et al. (1997) was not selected as a principal
study.
Schreiber et al. (2002) was a small study examining 9 adult staff at a day-care facility
colocated in the same building as a dry cleaner, comparing group mean visual contrast values to
age- and sex-matched referents values and identifying a LOAEL of 2 mg/m3
[LOAELrec = 0.7 mg/m3] (arithmetic mean concentration). Referents in this study were
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-11 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
acquaintances, local retail shop employees, staff of other local day-care centers, or NYSDOH
employees. Exposed and referent subjects were similar on sex and age; however, the paper lacks
any details of whether referents were of similar education or socioeconomic status. Use of
NYSDOH employees located in Albany, NY, may indicate referents and exposed subjects may
be different on education and other variables. Exposure assignment to subjects was based on
ambient monitoring during time of active dry cleaning; no personal monitoring was conducted.
Schreiber et al. (2002) used a standardized test (FACT) for near vision; however, a shortcoming
is that assessment of vision was 6 weeks after exposure ceased, when measured
tetrachloroethylene concentrations were 100-fold lower than during active dry cleaning. While
Schreiber et al. (2002) adopted a valid and sensitive test to measure vision, it was not selected as
a principal study due to its few subjects, concern that testers were not blinded to exposure
classification, concern about comparability of exposed and referent subjects, and lack of
concurrent exposure and outcome assessment.
Seeber (1989) evaluated the neurobehavioral effects of tetrachloroethylene on
101 dry-cleaning workers (employed in coin-operated or while-you-wait shops), and reported
effects on several measures of cognition at a LOAEL of 83 mg/m3 [LOAELHec = 29 mg/m3]
(time-weighted average mean concentration), compared to referents from several department
stores and receptionists from large hotels. A strength of the study was the relatively large sample
sizes used for all three groups, 57, 44, and 84 subjects in the lowest, highest and referent groups,
respectively. No information was provided on the methods used to identify subjects or their
reasons for participating in the study, although the authors reported that 29 service technicians
were excluded from participation because of either discontinuous exposure conditions with peak
concentrations or long periods of no exposure. The exposure assessment targeted estimates of
long-term exposure from interview data, active sampling of room air, and passive sampling of
personal air, including during entire shifts in summer and in winter. This information was used
in assigning dry cleaners to two exposed groups (83 and 364 mg/m3). The administered tests of
neuropsychological function included standardized tests of symptoms and personality; tests of
sensorimotor function, including finger tapping and aiming; and the Mira and Santa Ana
dexterity tests. Another strength of this study is its use of blinded examiners to test subjects.
Because the dry-cleaner groups and the control group differed in gender ratios, age, and scores
on the intelligence test, stratified regression analysis was used to statistically control for the
influence of these potentially confounding factors on test scores. Additional adjustment for
group differences in alcohol consumption did not alter the results. Seeber (1989) had relatively
good quality in terms of the addressing comparability of referent and exposed groups,
measurement of effect, and measurement of exposure. Therefore, it was selected as a principal
study.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-12 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Cavalleri et al. (1994) and Gobba et al. (1998) are two studies of the same exposed
population. Cavalleri et al. (1994) reported poorer performance (6% decrement on average) on a
test of color vision among 35 dry cleaning and laundry workers compared to 35 controls matched
on age, alcohol consumption, and smoking. The LOAEL for all workers in this study was
42 mg/m3 [LOAELrec =15 mg/m3] (time-weighted average mean concentration). Controls were
not matched on education or intelligence, but these factors have not been shown to be associated
with color vision. Exposure was assessed for individual subjects from personal monitoring over
the full work shift and represented an 8-hour time weighted average. Standard testing methods,
including an established protocol, were used to detect changes in color vision, which was
assessed by the Lanthony D-15 Hue desaturated panel. Statistical analyses included comparison
of group mean Color Confusion Indexes (CCIs) by the arithmetic mean of three exposure
groupings, all workers (42 mg/m3), dry cleaners (49 mg/m3), and ironers (33 mg/m3). Multiple
logistic regression analyses adjusted for effects of age, alcohol consumption, and smoking.
Gobba et al. (1998) examined color vision in 33 of these 35 dry cleaners and laundry
workers after a 2-year period, and reported a further decrement in color vision (9% decrement on
average) among 19 subjects whose geometric mean exposure had increased from 12 mg/m3 to
29 mg/m3 over the 2-year period. No improvement was observed among 14 subjects whose
geometric mean exposure had decreased from 20 mg/m3 to 5 mg/m3. The mean responses of
both subgroups supported a persistence of deficits in visual function, and suggested a worsening
of effects when exposure increased for individuals. A strength of Gobba et al. (1998) is subjects
serving as their self-controls, with scores on the test of color vision compared from the initial and
follow-up study. Given the vision deficits reported by Cavalleri et al. (1994). Gobba et al.
(1998) serves to confirm and extend those findings.
Cavalleri et al. (1994) is preferred to Gobba et al. (1998) as a principal study for
reference value derivation, for several reasons. First, the earlier study more clearly associated a
deficit in color vision with tetrachloroethylene exposure, through comparison to a suitable and
well characterized, unexposed reference group. The Gobba study did not include unexposed
controls, and therefore cannot distinguish the possible impact of age on the CCI scores of
subjects who were two years older at the second evaluation. Second, the Gobba et al. (1998)
study suggests that the earlier exposure was sufficient to cause the CCI deficit in at least those
subjects (n = 14) whose exposure decreased after the earlier evaluation. While the Gobba et al.
(1998) study also demonstrated further deficits in those whose exposure increased after the first
study (n = 19), it is not straightforward to relate the higher measurement to the incremental
deficit, given the lack of improvement in the subset with decreased exposure and the lack of
information concerning the other confounding variables considered in the first evaluation—
absolute age, smoking and alcohol status. In any case, a deficit existed in this subset before the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-13 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
follow-up period, at a lower exposure than that of the second evaluation. Third, the exposures in
Cavalleri et al. (1994) were reported as time-weighted average arithmetic means, which are
expected to represent total risk better than time-weighted average geometric means (as reported
in Gobbaetal.. 1998) when data are grouped (Crump. 1998). The point of departure (POD) was
therefore taken from the Cavalleri et al. (1994) study. The exposure level for the full study
sample is used as the LOAEL, using the following reasoning. Although no apparent CCI deficit
was seen in ironers, their reported exposure range (0.52-11.28 ppm, or 3.5-76 mg/m3) was
completely contained within the range of exposures for dry cleaners (0.38-31.19 ppm, or
2.6-210 mg/m3). Yet elevated CCI scores were observed at exposures lower than the mean
exposure of the ironers (4.8 ppm, or 33 mg/m3), indicating that the mean exposure of the ironers
cannot be considered a NOAEL. For these reasons, Cavalleri et al. (1994) is selected as a
principal study.
Echeverria et al. (1995) examined 65 dry cleaners in Detroit, MI, using a standardized
neurobehavioral battery, and found changes in cognitive and visuospatial function. A LOAEL of
156 mg/m3 [LOAELrec = 56 mg/m3] (time-weighted average mean concentration) was
identified, based on comparison of the two higher exposure categories with an internal referent
group comprising mainly counter clerks, who were matched to exposed dry cleaners on age and
education. The study had a high quality exposure-assessment approach and appropriate
statistical analyses that adjusted for covariates including alcohol. A potential selection bias may
have resulted from the 18% participation rate among dry-cleaning shop owners, if the low
participation could be explained by the health status of employees. The study also lacked an
unexposed referent group; subjects were categorized into three exposure groups. Without an
unexposed control group, however, the exposure level for the lowest exposure group (i.e., the
internal referent group), cannot be classified as a NOAEL or a LOAEL. This study was of
relatively good quality in terms of the comparability of referent and exposed groups,
measurement of effect, and measurement of exposure and, although there are concerns about the
lack of an unexposed referent group, this study was selected as a principal study.
5.1.1.3.3. Evaluation of experimental human exposure studies
The two human controlled exposure studies (Altmann et al.. 1990; Hake and Stewart.
1977) were of fewer subjects, shorter exposure durations and effects were observed at higher
exposure concentrations than chronic studies of residential and occupational exposure. While
subjects in Altmann et al. (1990) could serve as their own controls, there was not an unexposed
group. Therefore, neither study was selected as a principal study given the availability of
suitable human data of chronic duration. These studies do provide qualitative evidence for
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-14 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
hazard identification of neurological deficits in visual function and neurological function and add
support for choice of key endpoints in principal studies.
5.1.1.3.4. Evaluation of animal neurotoxicity studies
The animal neurotoxicity studies mostly consist of acute duration studies (Boves et al..
2009; Ki ell strand et al.. 1985; Oshiro et al.. 2008) and subchronic (repeated dosing) studies
which generally involve lower exposures than the acute animal studies (Mattsson et al.. 1998;
Rosengren et al.. 1986; Wang et al.. 1993). However, these studies covered shorter exposure
duration periods than the available human studies, and require extrapolation of animal
observations to humans. They were not considered principal studies given the availability of
suitable human data from chronic exposures. The findings in the animal studies contribute to the
weight of evidence that tetrachloroethylene exposure results in neurological deficits and is
considered supportive of the human studies in terms of hazard identification.
5.1.1.3.5. Selection of principal studies
To summarize, three studies (Cavalleri etal.. 1994; Echeverria et al.. 1995; Seeber. 1989)
had more of the preferred qualities compared to other epidemiologic studies of occupational and
residential tetrachloroethylene and were considered principal studies for deriving an RfC. None
of these three studies stands out as a clearly superior candidate for identifying the POD.
Endpoints selected for the RfC were reaction time measures (Echeverria et al.. 1995). cognitive
changes (Echeverria et al.. 1995; Seeber. 1989) and visual function changes (Cavalleri et al..
1994).
5.1.2. Additional Analyses: Feasibility of Dose-Response Modeling
The present analysis defines a POD using the traditional NOAEL/LOAEL approach. The
NOAELs/LOAELs were adjusted to an equivalent continuous exposure (U.S. EPA. 1994) and
described in Section 5.1.1) so that comparisons could be made between studies. Ambient
(inhaled) concentration of tetrachloroethylene was used as the dose metric in deriving the RfC.
Because the application of dose-response modeling offers advantages over traditional
LOAEL/NOAEL approaches, the data sets from the endpoints in the three principal studies (see
Table 5-3) (Cavalleri etal.. 1994; Echeverria et al.. 1995; Seeber. 1989) were evaluated to
determine feasibility of dose-response modeling. In all of the studies, it was determined that
PODs could not be derived using dose-response modeling, for varying reasons. First, Seeber
(1989) included a control and two exposure groups in a neurobehavioral analysis of workers
exposed to tetrachloroethylene. Information that could be used to identify benchmark response
levels (BMRs) corresponding to minimally biologically significant response levels for the
administered tests was not located.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-15 DRAFT—DO NOT CITE OR QUOTE
-------
1 In evaluating the CCIs in Cavalleri et al. (1994). normative data for color confusion
2 Qregren et al.. 2002; Lomax et al.. 2004) were considered. However, the normal ranges are
Table 5-3. Application of uncertainty factors to four neurological endpoints
from three studies used to derive the RfC
Neurological endpoint
Human
equivalent
NOAEL/LOAEL
(mg/m3)
Uncertainty factors (UFs)
Candidate
RfC
mg/m3
Reference
Composite
UF
ufa
UFh
UFS
ufd
ufl
Cognitive Domain
Visual reproduction,
pattern memory, pattern
recognition—adult,
occupational
56
(LOAEL)
1,000
1
10
1
10
10
0.056
Echeverria
et al.
(1995)
Digit symbol,
cancellation, digit
reproduction, perceptual
speed—adult,
occupational
29
(LOAEL)
1,000
1
10
1
10
10
0.029
Seeber
(1989)
Reaction Time Domain
Reaction time in pattern
memory— adult,
occupational
56
(LOAEL)
1,000
1
10
1
10
10
0.056
Echeverria
et al.
(1995s)
Visual Function Domain
Color confusion—
adults, occupational
15
(LOAEL)
1,000
1
10
1
10
10
0.015
Cavalleri
et al.
(1994)
3 influenced strongly by age, which was not available for the data set at a similar level of
4 resolution as the normative data. The variability in the available data was not amenable to
5 modeling with available models. Finally, Echeverria et al. (1995) identified three exposure
6 groups, but there is no unexposed group for comparison. Historical control data from the
7 Echeverria group were unavailable, precluding the derivation of PODs from the logistic
8 regression they reported.
5.1.3. Reference Concentration (RfC) Derivation, Including Application of Uncertainty
Factors
9 Adjusted LOAELs, ranked highest to lowest, are 56 mg/m3 (Echeverria et al.. 1995).
10 29 mg/m3 (Seeber. 1989). and 15 mg/m3 (Cavalleri et al.. 1994). which were selected as the
11 PODs, as described above. The PODs were reduced by the following UFs:
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-16 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
1. Human variation. The UF of 10 was applied for human variation for all of the studies
that were selected in derivation of the RfC. These studies are from occupationally
exposed subjects, who are generally healthier than the overall population, and thus
provide no data to determine the relative effects of susceptible population including
children, elderly, and/or people with compromised health. Additionally, no information
was presented in the human studies with which to examine variation among subjects.
The quantitative analyses that have been performed evaluating pharmacokinetic
variation between adults and children for tetrachloroethylene and its metabolites using
physiologically based pharmacokinetic (PBPK) models (Clewell and Andersen, 2004; Gentry
et al.. 2003; Pelekis et al.. 2001) indicate that validation of these results for various
life-stages and further refinement of the parameters in the model are necessary before
the results of such an analysis can be considered for use in risk evaluation.
2. Animal-to-human uncertainty. Human studies were used in the derivation of the RfC.
Consequently, this UF is not needed.
3. Subchronic-to-chronic uncertainty. A factor to address the potential for more severe or
additional toxicity from chronic or lifetime exposure to tetrachloroethylene was not used
for the principal studies. The PODs are based on studies involving chronic exposure, so
no extrapolation was necessary.
4. LOAEL-to-NOAEL uncertainty. A UF of 10 is generally applied when the POD is a
LOAEL due to a lack of a NOAEL. When NOAELs are used, a UF is not applied. For
all of the human studies and endpoints selected (Cavalleri et al.. 1994; Echeverria et al..
1995; Seeber. 1989). PODs were LOAELs and a UF of 10 was applied to these
endpoints.
5. Database uncertainty. A database UF of 10 has been applied to address the lack of data
to adequately characterize the hazard and dose-response in the human population. A
number of data gaps were identified from both the human and animal literature, including
the need for high quality epidemiologic studies of residential exposures including
children and the elderly, chronic animal studies (including in developing animals)
designed to define and characterize the exposure-response relationships for the observed
neurotoxicological effects, particularly, reaction time, cognitive and visual function.
Additionally, the available studies of immunologic and hematologic toxicity studies (e.g.,
(Emara et al.. 2010; Marth. 1987) are limited. The relative lack of data taken together
with the concern that other structurally related solvents have been associated with
immunotoxicity, particularly relating to autoimmune disease (Cooper et al.. 2009)
contributes to uncertainty in the database for tetrachloroethylene.
The available epidemiologic studies of residential exposures were judged to be limited
for developing an RfC (Altmann et al.. 1995; NYSDOH. 2010; Schreiber et al.. 2002;
Storm et al.. In Press) based on consideration of selection bias, residual confounding
(population comparability) and/or selection of neurological methods. Yet the residential
studies yielded the most sensitive neurotoxic endpoint associated with
tetrachloroethylene exposure, decrement in VCS. Because this specific endpoint was not
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-17 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
evaluated in any of the occupational studies, it cannot be concluded that VCS changes
would not occur at the higher exposures of the occupational studies. There were
impairments in CCI for one set of occupationally exposed subjects (Cavalleri etal.. 1994;
Gobba et al.. 1998). but this effect was not evaluated in other occupational studies. There
is also a lack of studies which evaluated the critical effects reaction time, cognitive and
visual functional deficits in populations exposed to tetrachloroethylene at lower than the
studied occupational exposure levels, including at residential levels. These data gaps,
and the lack of developmental and immune functional assessment therefore contribute to
the uncertainty in the tetrachloroethylene toxicity database.
These UFs were applied to each of the four endpoints from the three principal
neurotoxicological studies of occupational tetrachloroethylene exposure: color vision changes
(Cavalleri et al.. 1994). cognitive and reaction time changes (Echeverria et al.. 1995). and
neurobehavioral changes in cognitive performance tasks (Seeber. 1989). The UFs for each study
and endpoint are presented in Table 5-3 as well as in Figure 5-2. RfCs from the different
endpoints ranged from 0.015 to 0.056 mg/m3. A value of 0.04 mg/m3 is supported by these
multiple studies, as the midpoint of the range of available values (then rounded to one significant
figure), and is the recommended RfC for tetrachloroethylene.
5.1.4. Dose-Response Analyses for Comparison of Noncancer Effects Other Than Critical
Effects in Neurotoxicity
This section presents inhalation dose-response analyses for noncancer effects other than
the critical effect of neurotoxicity. The purpose of these analyses is twofold: (1) to provide a
quantitative characterization of the relative sensitivity of different organs/sy stems to
Figure 5-2. Reference concentration values for inhalation exposure to
tetrachloroethylene.
Ectwverris et al (1995)- Co^nim e laea-aits: adult, occupational - LOAEL
I rbfvfrrufrjit (I99sl- Reaction time rat»s«i Adult, occupation!! - LOAEL
<• 1»J m is « w
Scfbtr (19S9) S>ui ahthnviiiral measures; adult, occupational - LOAEL
Cat allei'i M aijlW) - Calor roufuuoa: adult. occupational - LOAEL
ie-q: a-n km imi im:
[Pl.RCj mg m*
9 Point of Departure
E2 L"Fa - Interspecies, aramaitohuman
8 ITj - Subphrenic to chxornc exposuia dmstren
0 til - LOAEL to NOAEL
0 tTjl - Intia species human vanabttav
BiTo - Database
*.' Candidate Reference C emcentratiaii
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-18 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
tetrachloroethylene, and (2), to provide information that may be useful for cumulative risk
assessment in which multiple chemicals have a common target organ/system other than the
central nervous system. Therefore, for each organ/system, —saiple reference concentrations"
(sRfCs) are calculated based on the same methodology as is used for the critical effect of
neurotoxicity. These sRfCs are based on an evaluation of studies identified in Section 4.10 as
suitable for dose-response analysis.
The method of analysis is the same as that described above for neurotoxicity, using the
NOAEL/LOAEL approach. Benchmark dose modeling was not performed because these sample
RfCs are meant for comparison purposes only (across organs/tissues or across chemicals). HECs
are derived using either (1) the RfC methodology for a Category 3 gas, extrathoracic effects,
adjusted for equivalent continuous exposure; or (2) the PBPK model with an appropriate dose
metric. In addition, the PBPK model being used to perform route-to-route extrapolation from
oral to inhalation exposure, so both inhalation and oral studies are considered together here. The
HEC is then treated as a POD to which the following uncertainty factors may be applied:
1. Human variation. The UF of 10 is applied for human variation to all PODs. The
rationale is the same as described above for neurotoxicity. Furthermore, there is some
indication that human variability (at least for one end point) may be substantially more
than that implied by the default UF. Kidney toxicity is thought to be associated with
metabolism of tetrachloroethylene along the glutathione (GSH) conjugation pathway. As
described in Section 3.5, PBPK model predictions for GSFt conjugation span a wide
range that may be due to uncertainty, variability, or both. Glutathione S-transferase
(GSTs) are known to be polymorphic in the human population, with some isoforms
exhibiting a substantial population of null phenotypes.
2. Animal-to-human uncertainty. The PODs from rats and mice are expressed as HECs
calculated using either the RfC methodology or the PBPK model. Therefore, the UF of
three is applied for animal-to-human uncertainty to the PODs from rats and mice to
account for potential pharmacodynamic differences. This factor is not applied to PODs
from human studies.
3. Subchronic-to-chronic uncertainty. When the POD is based on a study of subchronic or
shorter duration, then the UF of 10 is applied to address the potential for additional or
more severe toxicity from chronic or lifetime exposure.
4. LOAEL-to-NOAEL uncertainty. A UF of 10 is generally applied when a LOAEL is used
due to a lack of a NOAEL. This factor may be reduced to 3 if the effect is considered
minimally adverse at the response level observed.
5. Database uncertainty. A database UF of 10 is applied to all PODs. The rationale is the
same as described above for neurotoxicity.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-19 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
5.1.4.1. Sample Reference Concentrations (RfCs) for Kidney Toxicity
As discussed in Section 4, numerous studies have reported adverse effects in the kidney
from tetrachloroethylene. Five studies reporting kidney toxicity were identified in Section 4.10
as suitable for dose-response analysis. The only human study was Mutti et al. (1992), which
reported statistically significant increases in retinol binding protein (RBP), p2n-globulin, and
albumin in urine among dry cleaners as compared to matched controls. In addition, for
seven different urinary markers, the prevalence of individuals with abnormal values
(>95th percentile of controls) was four- to fivefold greater in the exposed group. This study was
in humans chronically exposed and was thus used to calculate a sRfC. Of the rodent studies
reporting nephrotoxicity, only JISA (1993) identified a chronic NOAEL, with the other
three rodent studies reporting subchronic (Jonker et al.. 1996) or chronic LOAELs (NCI. 1977;
NTP. 1986b).
Therefore, among the rodent studies, only JISA (1993). which reported effects in both
mice and rats, were used in sRfCs calculations. A summary of the PODs and UFs applied is in
Table 5-4. The resulting sRfCs range from 0.05-0.2 mg/m3 based on nuclear enlargement
(karyomegaly) in the proximal tubules of chronically exposed mice and rats (JISA. 1993) with a
slightly lower sRfC of 0.03 mg/m3 based on urinary markers of nephrotoxicity in occupationally
exposed humans (Mutti et al.. 1992).
5.1.4.2. Sample Reference Concentrations (RfCs) for Liver Toxicity
As discussed in Section 4, numerous studies have reported adverse effects in the liver
from tetrachloroethylene. Six studies, none in humans, reporting liver toxicity were identified in
Section 4.10 as suitable for dose-response analysis. Only JISA (1993) reported a chronic
NOAEL, so was carried forward for derivation of a sRfC. However, it is unclear whether the
reported effect of angiectasis, or enlargement of the blood vessels, is related to the other liver
effects of tetrachloroethylene, which generally involve hepatocytes. Therefore, two other studies
were utilized, one of which reported a chronic LOAEL for liver degeneration and necrosis (NTP.
1986b) and the other of which reported a NOAEL for liver weight increases after 6 week
exposures (Buben and O'Flahertv. 1985). The remaining studies either only reported a LOAEL
(Jonker et al.. 1996; Kiellstrand et al.. 1984). or reported a NOAEL for a very short duration (14
days. Berman et al.. 1995). and were therefore not considered further.
Therefore, JISA (1993). NTP (NTP. 1986b). and (Buben and O'Flahertv. 1985) were used
to calculate sRfCs. In addition, PBPK modeling was used to calculate the total rate of oxidative
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-20 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
metabolism in the liver as a dose metric for deriving the HECs.1 A summary of the PODs and
UFs applied is in Table 5-5. The resulting sRfCs range from 0.09 mg/m3 based increased
liver/body weight ratios after 6 week exposures (Buben and O'Flahertv. 1985) to 0.7 mg/m3
based on liver effects after chronic exposures (USA. 1993; NTP. 1986b). It should also be noted
that in the chronic studies, increased liver tumors were observed at the lowest doses tested.
Therefore, under chronic exposure conditions, cancer effects are likely to be more
important than noncancer effects in the liver.
5.1.4.3. Sample Reference Concentrations (RfCs) for Immunotoxicity and Hematologic
Toxicity
As discussed in Section 4, a number of studies have reported changes in hematologic or
immunologic parameters with tetrachloroethylene exposure. Two studies reporting hematologic
effects were identified in Section 4.10 as suitable for dose-response analysis. The human study
(Emara et al.. 2010) reported changes in various standard hematological measures in subjects
with mean blood levels of 1.685 mg/L. Application of the PBPK model gives an air
concentration estimate during exposure of 18 ppm corresponding to this blood level, assuming
constant concentration during exposure. Adjustment to equivalent continuous exposure gives an
1 The MOA for tetrachloroethylene-induced liver toxicity is not clear. It appears that TCA as the sole contributory
metabolite cannot explain tetrachloroethylene-induced hepatotoxicity (Buben and O'Flahertv. 1985; Clewell et al..
20051. It is not known whether reactive intermediates such as tetrachloroethylene oxide and trichloroacetyl chloride
are involved in induced liver toxicity. In consideration of these uncertainties, it appears more appropriate to use
total rate of oxidative metabolism as the dose-metric for tetrachloroethylene-induced liver toxicity. This quantity is
then scaled by body-weight to the 374th power so as to enable extrapolation of risk across species.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-21 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-4. Sample RfCs for kidney effects
Kidney endpoint
(species)
HEC in
mg/m3
(LOAEL/
NOAEL)
Uncertainty factors (UFs)
Sample
RfC
mg/m3
Reference
Composite
UF
ufa
UFh
UFS
ufd
ufl
Urinary markers of
nephrotoxicity (human)
34
(LOAEL)
1,000
1
10
1
10
10
0.03
Mutti et al.
CI992)
Nuclear enlargement in
proximal tubules (rat)
61 (NOAEL)
300
3
10
1
10
1
0.2
JISA (1993)
Nuclear enlargement in
proximal tubules (mouse)
14 (NOAEL)
300
3
10
1
10
1
0.05
JISA (1993)
Table 5-5. Sample RfCs for liver effects
HEC in
Uncertainty factors (UFs)
Liver endpoint (species)
mg/m
(LOAEL
/
NOAEL)
Composi
te UF
UF
A
UFh
UFS
ufd
ufl
Sample
RfC
mg/m3
Reference
Increased angiectasis
(mouse)
210
(NOAEL
)
300
3
10
1
10
1
0.7
iisa
(1993)
Increased liver
degeneration/necrosis
(mouse)
2,100
(LOAEL)
3,000
3
10
1
10
10
0.7
NTP
(1986b)
Increased liver/body weight
ratio (mouse)
270b
(NOAEL
)
3,000
3
10
10
10
1
0.09
Buben &
O'Flaherty
(1985)
"Calculated with PBPK model using the dose metric of liver oxidative metabolism.
bRoute-to-route extrapolation from oral exposure.
HEC of 6.4 ppm, or 43 mg/m3. This can be treated as a chronic LOAEL, given the 7-year mean
exposure duration (>10% of lifespan). The other study (Marth. 1987) reported reversible
hemolytic anemia in mice after 7 weeks drinking water exposure to 2 week old mice for 7 weeks.
Because only a LOAEL was identified, the exposures were subchronic, and the effect has not
been reproduced at such low exposure in other studies, Marth (1987) was not considered further
for sRfC derivation. However, it should be noted that the LOAEL identified was very
low—0.05 mg/kg-day—and may be a cause for additional concern about hematologic effects.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-22 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Therefore, Emara et al. (2010) was used to calculate a sRfC. A summary of the POD and UFs
applied is in Table 5-6. The result is a sRfC of 0.04 mg/m3.
5.1.4.4. Sample Reference Concentrations (RfCs) for Reproductive and Developmental
Toxicity
As discussed in Section 4, a number of studies have reported reproductive and
developmental effects from tetrachloroethylene exposure. Four studies, none in humans,
reporting reproductive or developmental effects were identified in Section 4.10 as suitable for
dose-response analysis. All of these studies reported NOAELs. The developmental studies were
all of appropriate duration for detecting those effects. The reproductive study (Beliles et al..
1980) was short term (5 days exposure), but was the only suitable study for reproductive toxicity
and assessment was limited to males. Therefore, all four studies were used to calculate sRfCs.
A summary of the PODs and UFs applied is in Table 5-7. For all these endpoints, the subchronic
to chronic UF was not used because the studies sufficiently covered the developmental window
or window of sperm development. The resulting sRfCs range from 0.4-0.7 mg/m3 for different
developmental effects (Carney et al.. 2006; Nelson et al.. 1980; Tinston. 1994). with an
intermediate value of 0.5 mg/m3 for reduced sperm quality (Beliles et al.. 1980).
5.1.4.5. Summary of Sample Reference Concentrations (RfCs) for Noncancer Endpoints
Other Than the Critical Effect
The lowest sRfCs for these noncancer endpoints are similar to the values calculated based
on the critical effect of neurotoxicity (see Figure 5-3), therefore supporting the selection of the
critical effect: 0.03 mg/m3 from Mutti et al. (1992) and 0.04 mg/m3 from Emara et al. (2010).
The other sRfCs are less than 20-fold greater than the RfC. This suggests that multiple effects
may begin to occur as exposure rises above those at which tetrachloroethylene begins to induce
neurotoxicity. These results also suggest that it is important to take into account effects from
tetrachloroethylene other than neurotoxicity when assessing the cumulative effects of multiple
exposures.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-23 DRAFT—DO NOT CITE OR QUOTE
-------
""J
O S
ri s:
o ^
O
^ s?
K ^
^ >!
^ s
¦si
o
s:
>!
§•
rs
st
>!
Si
5^
v»
On
to
Table 5-6. Sample RfCs for immunological and hematological effects
Immunotoxicity/hematotoxicity
endpoint (species)
MI C in
mg/m3
(LOAEL/
NOAEL)
Uncertainty factors (UFs)
Sample
RfC
mg/m3
Reference
Composite
UF
UFa
UFh
UFs
UFd
UFl
Reduced RBC, hemoglobin;
increased WBC, lymphocytes,
IgE (human)
43
(LOAEL)
1,000
1
10
1
10
10
0.04
Emara et al.
(2010)
Si
5
§•
>1
to
o
3
RBC = red blood cells.; WBC = white blood cells.
Table 5-7. Sample RfCs for reproductive and developmental effects
Reproductive/developmental
endpoint (species)
HEC in
mg/m3
(LOAEL/
NOAEL)
Uncertainty factors (UFs)
Sample
RfC
mg/m3
Reference
Composite
UF
UFa
UFh
UFs
UFd
UFl
Decreased weight gain; altered
behavior, brain acetylcholine (rat)
200
(NOAEL)
300
3
10
1
10
1
0.7
Nelson et al.
0980^)
Reduced sperm quality (mouse)
140
(NOAEL)
300
3
10
1
10
1
0.5
Beliles et al.
0980s)
Increased F2A pup deaths by Day
29; CNS depression in F1 and F2
122
(NOAEL)
300
3
10
1
10
1
0.4
Tinston et al.
(1994)
Decreased fetal and placental
weight; skeletal effects (rat)
110
(NOAEL)
300
3
10
1
10
1
0.4
Carney et al.
(2006)
CNS = central nervous system
O
O
-------
_ >g
o
0 ^
1 ^
& Si
^ &
k ^
hti. Co
% <=>
^ s
¦1 ?».
O
3:
>!
§•
Cb
ss
S
>!
•5
§.<§>
>!
N> G\
2 ^
to
~n
H
O
O
cm
CNS
UFcomp=1000
l Echeverria et al. (1995) [H]
^Fcomj^lOO^
l Seeber (1989) [H]
UFcomp=1000
t Cavalleri et al. (1994) [H]
Kidney
UFcomp=1000
_UFcom£^300_
( Mutti et al. (1992) [H]
©JISA (1993) [R]
UFcomp=300
¦0 JISA (1993) [
Liver
_UFcom£^300_
-© JISA (1993) [
UFcomp=3000
^Fcom£^300^
W NTP (1986) [
-© Buben & O'Flaherty (1985) [M]
Immunological/Hematological
Emara et al. (2010) [H]
Reproductive/Developmental
~
n
UFcomp=300
0
~—
UFcomp=300
O Cz
-© Nelson et al. (1980) [R]
© Beliles et al. (1980) [M]
~1 1—I I I II11
~1 1—I I I I 111
~1 1—I I I I 111
n—i—i i i 1111
~i—i—i i i 1111
~i—i—i i i 111|
10
-2
10
-1
1
10
10
10°
10
l/r 3
mg/m
Figure 5-3. Comparison of candidate RfCs (black squares) supporting the RfC (grey vertical line) and sample RfCs (open squares) for
effects other the critical effect (CNS toxicity). Black circles = study/endpoint LOAEL in terms of human equivalent
concentrations. Open circles = study/endpoint NOAEL in terms of human equivalent concentrations. Species in each study is
shown in brackets after the reference (mouse: M; rat: R; human: H).
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
5.1.5. Previous Inhalation Assessment
There is no previous IRIS RfC assessment for tetrachloroethylene.
5 2 ORAL REFERENCE DOSE (RfD)
Ideally, the studies of greatest duration of exposure and conducted via the oral route of
exposure have the most confidence for derivation of an RfD.1 An earlier assessment of
tetrachloroethylene oral noncancer toxicity by EPA, for example, identified liver toxicity in
Buben and O'Flaherty (1985) as the critical effect for developing an RfD (U.S. EPA. 1988).
However, the application of pharmacokinetic models for a route-to-route extrapolation of the
inhalation studies expands the database of studies suitable for RfD calculation. The California
EPA (2001). for example, carried out a route-to-route extrapolation of the human inhalation
studies of neurotoxic effects to develop a Public Health Goal for oral tetrachloroethylene
exposure, based on a route-to-route extrapolation from inhalation neurotoxicity studies.
5.2.1. Choice of Principal Study and Critical Effects
As discussed in Section 5.1.1, based on evidence that neurological effects were
associated with lower tetrachloroethylene concentrations, neurotoxicity is selected as the critical
noncancer health effect of tetrachloroethylene. The three principal studies are of inhalation
exposures. The nervous system is an expected target with lower oral tetrachloroethylene
exposures, because tetrachloroethylene and many metabolites produced from inhalation
exposures will also reach the target tissue via oral exposure. In addition, other organ systems
such as the liver and kidney are also common targets associated with both inhalation and either
oral routes of subchronic or chronic exposure. The similarity of effects in these organ systems
with either oral or inhalation exposure to tetrachloroethylene supports the use of route
extrapolation to compare PODs for oral and inhalation exposure. In addition, differences in first-
pass metabolism between oral and inhalation exposures can be adequately accounted for by the
PBPK model. For these reasons, the three inhalation neurotoxicity studies used to derive the
RfC are chosen as principal studies for the RfD.
1 The RfD is expressed in units of milligrams per kilogram body weight per day (mg/kg-day). In general, the RfD is
an estimate (with uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human population
(including sensitive subgroups) that is likely to be without an appreciable risk of deleterious noncancer effects
during a lifetime.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-26 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
5.2.2. Additional Analyses: Route-to-Route Extrapolation Using PBPK Modeling
The present analysis defines a POD using the traditional NOAEL/LOAEL approach. As
discussed in Section 5.1.2, dose-response modeling was not feasible with the principal studies.
This assessment has attempted to expand the database for derivation of an RfD using relevant
inhalation data and route-to-route extrapolation with the aid of a PBPK model (see Section 3.5).
Several factors support the use of route-to-route extrapolation for tetrachloroethylene.
Tetrachloroethylene has been shown to be rapidly and well absorbed by both the oral and
inhalation routes of exposure (ATSDR. 1997). Additionally, the metabolic pathways and
kinetics of excretion with oral exposure are similar to those of inhalation exposure (ATSDR.
1997). Furthermore, the data for oral administration indicate a pattern of effects similar to that of
inhalation exposure. PBPK modeling was also used with suitable studies in animals in order to
inform the process of extrapolating to human equivalent doses (HEDs). It is not clear if the
noncancer effects observed in humans are the result of tetrachloroethylene itself and/or one or
more metabolites. However, tetrachloroethylene in the blood can safely be presumed to be a step
in the toxicity pathway. Therefore, area under the curve (AUC) of blood tetrachloroethylene
concentration derived from PBPK modeling is considered the best surrogate for an internal dose.
The use of blood tetrachloroethylene provides some attempt to account for breathing rates and to
adjust for kinetic processes related to tetrachloroethylene ADME, and it is assumed to better
reflect tetrachloroethylene pharmacokinetics than use of default methodologies. Moreover,
based on the results of the harmonized PBPK model (Chiu and Ginsberg), the sensitivity to the
choice of dose metric for route-to-route extrapolation is low, with alternative dose metrics such
as GSH metabolism, oxidative metabolism, or trichloroacetic acid (TCA) in blood giving route-
to-route conversions within 1.4-fold of the conversion based on tetrachloroethylene in blood.
The harmonized PBPK model of Chiu and Ginsberg was used to derive the oral dose that
would result in the same tetrachloroethylene in blood AUC as that following a continuous
inhalation exposure from the three principal studies (Cavalleri etal.. 1994; Echeverria et al..
1995; Seeber. 1989). The route-to-route extrapolation starts with the estimation of the average
venous blood tetrachloroethylene AUC resulting from continuous inhalation exposure at the
adjusted LOAELs from the four neurological endpoints in the three principal studies (Cavalleri
et al.. 1994; Echeverria et al.. 1995; Seeber. 1989). The venous blood tetrachloroethylene AUC
at steady state resulting from continuous exposure to these tetrachloroethylene concentrations
was estimated to range from 4.5 to 17 mg-hr/L-day, according to the Chiu and Ginsberg
harmonized model. While the model utilizes data from some healthy adult volunteers, it cannot
be considered to address pharmacokinetic variation in the full human population. The oral
exposure scenario was also modeled as continuous, since at these exposure levels, the AUC of
tetrachloroethylene in blood is insensitive to the exposure pattern. The route-to-route
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-27 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
extrapolation oral ingestion values at the LOAELs were 2.6 mg/kg-day for Cavalleri et al.
(1994), 9.7 mg/kg-day for Echeverria et al. (1995) and 5.0 mg/kg-day for Seeber (1989). The
results are presented in Table 5-8.
5.2.3. Reference Dose (RfD) Derivation, Including Application of Uncertainty Factors
To address differences between study conditions and conditions of lifetime human
environmental exposure, the route-to-route extrapolated PODs of 2.6 mg/kg-day (Cavalleri et al..
1994). 9.7 mg/kg-day (Echeverria et al.. 1995) and 5.0 mg/kg-day (Seeber. 1989) are reduced by
UFs that consider specific areas of uncertainty. The application of uncertainty factors was
similar to that for the different endpoints used to derive the RfC. The following areas of
uncertainty were evaluated for this RfD:
1. Human variation. The UF of 10 is applied for human variation for all of the studies that
were selected in derivation of the RfC. As indicated in the RfC discussion the principal
studies selected do not include evaluation of potential sensitive populations including
children, elderly, and immune compromised individuals.
2. Animal-to-human uncertainty. Since the principal studies and critical endpoints were
from human studies, this factor was not applied.
3. Subchronic-to-chronic uncertainty. As with the RfC derivation, described in
Section 5.1.3, for the human studies, the PODs are based on studies involving chronic
exposure, so no extrapolation was necessary.
4. LOAEL-to-NOAEL uncertainty. The PODs from all the principal studies were LOAELs
so a 10-fold factor was applied to approach the range where a negligible response could
be expected.
5. Database uncertainty. A database UF of 10 has been applied to address the lack of data
to adequately characterize the hazard and dose-response in the human population as was
done for the derivation of the inhalation RfC. A number of data gaps are identified in
both the human and animal literature. Notable gaps in the literature are the need for high
quality epidemiologic studies of residential exposures, or suitable chronic animal studies
(including in developing animals) designed to define and characterize the exposure-
response relationships for the observed neurotoxicological effects, particularly, reaction
time deficits, cognitive and visual function. Briefly, neurotoxicological changes are
observed in residential studies (Altmann et al.. 1995; NYSDOH. 2010; Schreiber et al..
2002; Storm et al.. In Press) at exposures that range from 2-100 times lower than the
exposures in the principal occupational exposure studies (Cavalleri etal.. 1994;
Echeverria et al.. 1995; Seeber. 1989). The relative lack of data concerning immune and
hematological toxicities taken together with the concern that other structurally related
solvents have been associated with immunotoxicity, particularly relating to autoimmune
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-28 DRAFT—DO NOT CITE OR QUOTE
-------
a
o
<
s?
cs
>!
o
^3
§•
rs
st
3
>!
sT
St
*3
o
>1
>!
Oq
o
>!
Si
Si
s
O
Sfc
I
§•
K) On
2
Table 5-8. Application of uncertainty factors to neurological endpoints from three studies used to derive the
RfD
Neurological endpoint
Oral
human equivalent
dose", mg/kg-day
(NOAEL/LOAEL)
Uncertainty factors (UFs)
Candidate
RfD
mg/kg-day
Reference
Composite
UF
ufa
UFh
UFS
ufd
ufl
Cognitive domain
Visual reproduction, pattern memory,
pattern recognition—adult, occupational
9.7
(LOAEL)
1,000
1
10
1
10
10
0.0097
Echeverria
etal. (1995)
Digit symbol, cancellation, digit
reproduction, perceptual speed—adult,
occupational
5.0
(LOAEL)
1,000
1
10
1
10
10
0.0050
Seeber
(1989)
Reaction time domain
Reaction time in pattern memory, adult,
occupational
9.7
(LOAEL)
1,000
1
10
1
10
10
0.0097
Echeverria
etal. ri995^
Visual function domain
Color confusion—adults, occupational
2.6
(LOAEL)
1,000
1
10
1
10
10
0.0026
Cavalleri
etal. C1994)
to
VO
~n
H
O
O
Equivalent oral exposure from application of the PBPK model on the basis of equivalent AUC of blood tetrachloroethylene for humans.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
5. Disease (Cooper et al.. 20091 also contributes to uncertainty in the database for
tetrachl oroethy lene.
6. UFs for the different endpoints were applied similarly to that for the RfC. The PODs
from each neurological endpoint were derived from a route-to-route extrapolation using a
PBPK model to obtain oral exposure equivalents. Composite UFs applied was 1,000 for
the four critical endpoints. A summary for each endpoint can be found in Table 5-8 and
Figure 5-4.
In summary, an RfD for tetrachl oroethy lene was developed through a route-to-route
extrapolation from the PODs for each of the four endpoints from the three principal
neurotoxicological studies of occupational tetrachloroethylene exposure: color vision changes
(Cavalleri et al.. 1994). cognitive and reaction time changes (Echeverria et al.. 1995). and
neurobehavioral changes in cognitive performance tasks (Seeber. 1989). The oral exposure POD
equivalent to the continuous inhalation exposure NOAELs or LOAELs was estimated via PBPK
modeling. A composite UF for each of the four endpoints was 1,000. Dividing the POD by the
composite UF for each endpoint yields an RfDs ranging from 2.6 x 10 3 to 9.7 x 10 3 mg/kg-day.
From this range an RfD of 6 x 10 3 mg/kg-day is supported by these multiple studies, as a
midpoint of the range of available values (then rounded to one significant figure), and is the
recommended RfD for tetrachl oroethylene. This RfD is equivalent to a drinking water
concentration of 0.21 mg/L, assuming a body weight of 70 kg and a daily water consumption of
2L.
5.2.4. Dose-response Analyses for Noncancer Effects Other Than Critical Effect of
Neurotoxicity
This section presents oral dose-response analyses for noncancer effects other than the
critical effect of neurotoxicity. The purpose of these analyses is twofold: (1) to provide a
quantitative characterization of the relative sensitivity of different organs/sy stems to
tetrachloroethylene, and (2), to provide information that may be useful for cumulative risk
assessment in which multiple chemicals have a common target organ/system other than the
central nervous system. Therefore, for each organ/system, —saiple reference doses" (sRfDs) are
calculated based on the same methodology as is used for the critical effect of neurotoxicity.
These sRfDs are based on an evaluation of studies identified in Section 4.10 as suitable for dose-
response analysis.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-30 DRAFT—DO NOT CITE OR QUOTE
-------
J 5 KO
SI?
3 5"
® J* ~
as ^
& ~
>!
cs
^ s;
®L O s
t S 5
k. S a
^ 2S
-| 'Jf ^
"^ § J
o
U o-^
T- ns **
N> Q\
O '
to
4^
Echeverria et al (1995)-Cognitive measures; human -LOAEL, route-to-route
Echeverria et al (1995)- Reaction time measures; human - LOAEL, route-to-route
Seeber (1989)-Xeurobehavioral measures; human - LOAEL, route-to-route
Cavalleri et al (1994) - Color confusion; human - LOAEL, route-to-route
# Point of Departure
23 LT.\ - Interspecies; animal to human
H LTs - Subchronic to chronic exposure duration
0 UFl - LOAEL to NOAEL
SH LTh - Intraspecies; human variability
E3 LTp - Database
O Candidate Reference Dose
1E-03
1E-02
1E-01
1E+00
1E+01
[PERC] mg/kg-day
Figure 5-4. Reference dose values from principal studies following exposure to tetrachloroethylene.
L»J
w
H
a
o
-------
1 The method of analysis is the same as that described above for neurotoxicity, using the
2 NOAEL/LOAEL approach. Benchmark dose modeling was not performed because these sample
3 RfDs are meant for comparison purposes only (across organs/tissues or across chemicals). HEDs
4 are derived using either (1) mg/kg-day dose adjusted for equivalent continuous exposure; or
5 (2) the PBPK model with an appropriate dose metric. In addition, the PBPK model being used to
6 perform route-to-route extrapolation from inhalation to oral exposure, so both inhalation and oral
7 studies are considered together here. For each endpoint where PBPK modeling is used, the dose
8 metric used to derive the HED is the same as that used to derive the HEC. The HED is then
9 treated as a POD to which the following uncertainty factors may be applied:
10
11
12 1.
13
14 2.
15
16
17
18 3.
19
20
21 4.
22
23 5.
24
25
26
5.2.4.1. Sample Reference Doses (RfDs) for Kidney Toxicity
27 As discussed in Section 4, numerous studies have reported adverse effects in the kidney
28 from tetrachloroethylene. Five studies reporting kidney toxicity were identified in Section 4.10
29 as suitable for dose-response analysis. The only human study was Mutti et al. (1992). which
30 reported statistically significant increases in RBP, p2[j,-globulin, and albumin in urine among
31 chronically exposed dry cleaners as compared to matched controls. In addition, for
32 seven different urinary markers, the prevalence of individuals with abnormal values
33 (>95th percentile of controls) was four- to fivefold greater in the exposed group. This study was
34 considered adequate to derive a sRfD. Of the rodent studies reporting nephrotoxicity, only JISA
Human variation. The UF of 10 is applied for human variation to all PODs. The
rationale is the same as described above for neurotoxicity.
Animal-to-human uncertainty. The PODs from rats and mice are expressed as HEDs
calculated using the PBPK model. Therefore, the UF of three is applied for animal-to-
human uncertainty to the PODs from rats and mice to account for potential
pharmacodynamic differences. This factor is not applied to PODs from human studies.
Subchronic-to-chronic uncertainty. When the POD is based on a study of subchronic or
shorter duration, then the UF of 10 is applied to address the potential for more severe
toxicity from chronic or lifetime exposure.
LOAEL-to-NOAEL uncertainty. A UF of 10 is generally applied when a LOAEL is used
due to a lack of a NOAEL.
Database uncertainty. A database UF of 10 is applied to all PODs. The rationale is the
same as described above for neurotoxicity.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-32 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
(1993) identified a chronic NOAEL, with the other three rodent studies reporting subchronic
(Jonker et al.. 1996) or chronic LOAELs (NCI. 1977; NTP, 1986b).
Therefore, among the rodent studies, only JISA (1993). which reported effects in both
mice and rats, was carried forward to calculate sRfDs. Because all the studies are inhalation
studies, route-to-route extrapolation was performed using the PBPK model with the AUC of
tetrachloroethylene in venous blood dose metric. A summary of the extrapolated PODs and UFs
applied is in Table 5-9. The resulting sRfDs range from 0.007-0.03 mg/kg-day, based on
nuclear enlargement in the proximal tubules of chronically exposed mice and rats (JISA. 1993).
with a slightly lower sRfD of 0.005 mg/kg-day based on urinary markers of nephrotoxicity in
occupationally exposed humans (Mutti et al.. 1992).
5.2.4.2. Sample Reference Doses (RfDs) for Liver Toxicity
As discussed in Section 4, numerous studies have reported adverse effects in the liver
from tetrachloroethylene. Six studies, none in humans, reporting liver toxicity were identified in
Section 4.10 as suitable for dose-response analysis. Only JISA (1993) reported a chronic
NOAEL, so was carried forward for derivation of a sRfD. However, it is unclear whether the
reported effect of angiectasis, or enlargement of the blood vessels, is related to the other liver
effects of tetrachloroethylene, which generally involve hepatocytes. Therefore, two other studies
were included at this stage, one of which reported a chronic LOAEL for liver degeneration and
necrosis (NTP. 1986b) and the other of which reported a NOAEL for liver weight increases after
6 week exposures (Buben and O'Flahertv. 1985). The remaining studies either only reported a
LOAEL (Jonker et al.. 1996; Kiellstrand et al.. 1984). or reported a NOAEL for a very short
duration (14 days. Berman et al.. 1995). and were therefore not considered further.
Therefore, JISA (1993). NTP (1986b). and (Buben and O'Flahertv. 1985) were used to
calculate sRfDs. In addition, PBPK modeling was applied using the liver oxidative metabolism
dose metric to derive the HEDs. A summary of the PODs and UFs applied is in Table 5-10. The
resulting sRfDs range from 0.01 mg/kg-day based increased liver/body weight ratios after
6 week exposures (Buben and O'Flahertv. 1985) to 0.08 mg/kg-day based on liver effects after
chronic exposures (JISA. 1993; NTP. 1986b). It should also be noted that in the chronic studies,
increased liver tumors were observed at the lowest doses tested. Therefore, under chronic
exposure conditions in this organ, liver cancers are likely to be more important than noncancer
effects in the liver.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-33 DRAFT—DO NOT CITE OR QUOTE
-------
""J
0
1 ^
si s
^ s?
K ^
^ >!
% s
^ s
¦I"? a.
o
s:
>!
§•
rs
st
5
>!
Si
Table 5-9. Sample RfDs for kidney effects
Kidney endpoint (species)
111 1) ' in
mg/kg-day
(LOAEL/NO
AEL)
Uncertainty factors (UFs)
Sample
RfD
mg/kg-day
Reference
Composite
UF
UFa
UFh
UFs
UFd
UFl
Urinary markers of
nephrotoxicity (human)
5 .4 (LOAEL)
1,000
1
10
1
10
10
0.005
Mutti et al.
(1992)
Nuclear enlargement in
proximal tubules (rat)
9.5 (NO AEL)
300
3
10
1
10
1
0.03
JISA (1993)
Nuclear enlargement in
proximal tubules (mouse)
2.2 (NOAEL)
300
3
10
1
10
1
0.007
JISA (1993)
a -r
>1
to 0\
2 ^
3
aCalculated with PBPK model using the dose metric of AUC of tetrachloroethylene in venous blood.
Table 5-10. Sample RfDs for liver effects
U>
~n
H
O
O
Liver endpoint (species)
111 1) ' in
mg/kg-day
(LOAEL/
NOAEL)
Uncertainty factors (UFs)
Sample
RfD
mg/kg-day
Reference
Composite
UF
UFa
UFh
UFs
UFd
UFl
Increased angiectasis
(mouse)
24.5 (NOAEL)
300
3
10
1
10
1
0.08
JISA (1993)
Increased liver
degenerati on/necrosi s
(mouse)
252 (LOAEL)
3,000
3
10
1
10
10
0.08
NTP (1986b)
Increased liver/body
weight ratio (mouse)
32 (NOAEL)
3,000
3
10
10
10
1
0.01
Bub en &
O'Flaherty
0985s)
-------
""J
O S
ri s:
o ^
O
^ s?
K ^
^ >!
^ s
1^
~5 £
^3 S
2. £^
S' §¦*
Ci O
•
>1
to
o
aCalculated with PBPK model using the dose metric of liver oxidative metabolism.
Table 5-11. Sample RfDs for immunological and hematological effects
Immunotoxicity/
hematotoxicity endpoint
(species)
HI D' in
mg/kg-day
(LOAEL/
NOAEL)
Uncertainty factors (UFs)
Sample
RfD
mg/kg-day
Reference
Composite
UF
ufa
UFh
UFS
ufd
ufl
Reduced RBC, hemoglobin;
increased WBC, lymphocytes,
IgE (human)
6.8
(LOAEL)
1,000
1
10
1
10
10
0.007
Emara et al.
(2010)
On
to
aCalculated with PBPK model using the dose metric of AUC of tetrachloroethylene in venous blood.
RBC =red blood cells; WBC = white blood cells.
Table 5-12. Sample RfDs for reproductive and developmental effects
Reproductive/developmental
endpoint (species)
HEDa in
mg/kg-day
(LOAEL/N
OAEL)
Uncertainty factors (UFs)
Sample
RfD
mg/kg-day
Reference
Composite
UF
UFa
UFh
UFS
UFd
UFl
Reduced sperm quality
(mouse)
22 (NOAEL)
300
3
10
3
10
1
0.07
Beliles
et al. (1980")
LtJ
Calculated with PBPK model using the dose metric of AUC of tetrachloroethylene in venous blood.
~n
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
5.2.4.3. Sample Reference Doses (RfDs) for Immunotoxicity and Hematologic Toxicity
As discussed in Section 4, a number of studies have reported changes in hematologic or
immunologic parameters with tetrachloroethylene exposure. Two studies reporting hematologic
effects were identified in Section 4.10 as suitable for dose-response analysis. The human study
(Emara et al.. 2010) reported changes in various standard hematological measures in subjects
with mean blood levels of 1.685 mg/L. Application of the PBPK model provides an estimated
HED of 6.8 mg/kg-day. This was treated as a chronic LOAEL, given the 7 year mean exposure
duration (>10% of lifespan), and is carried forward to calculate a sRfD. The other study (Marth.
1987) reported reversible hemolytic anemia in mice after 7 weeks drinking water exposure to
2-week old mice for 7 weeks. Although Marth (1987) was not considered further, as
summarized in Section 5. 1.4.3, it should be noted that the LOAEL identified was very
low—0.05 mg/kg-day—and may be a cause for additional concern about hematologic effects.
Therefore, Emara et al. (2010) was used to calculate a sRfD. A summary of the POD and
UFs applied is in Table 5-11. The result is a sRfD of 0.007 mg/kg-day.
5.2.4.4. Sample Reference Doses (RfDs) for Reproductive and Developmental Toxicity
As discussed in Section 4, a number of studies have reported reproductive and
developmental effects from tetrachloroethylene exposure. Four studies, none in humans,
reporting reproductive or developmental effects were identified in Section 4.10 as suitable for
dose-response analysis. All of these studies reported NOAELs. The developmental studies were
all of appropriate duration for detecting those effects. The reproductive study (Beliles et al..
1980) was short term (5 days exposure), but was the only suitable study for reproductive toxicity.
The PBPK model does not include gestational, fetal, or neonate compartments, so none of the
inhalation studies could be converted to oral equivalents. However, the reproductive study was
performed in mature male mice, for which the PBPK model could be used.
Therefore, only Beliles et al. (1980) was used to calculate a sRfD. A summary of the
POD and UFs applied is in Table 5-12. The subchronic to chronic UF was not used because the
study period sufficiently covered the window of sperm production. The resulting sRfD is
0.07 mg/kg-day for reduced sperm quality (Beliles et al.. 1980).
5.2.4.5. Summary of Sample Reference Doses (RfDs) for Noncancer Endpoints Other
Than the Critical Effect
The lowest sRfDs for these noncancer endpoints are similar to the values calculated
based on the critical effect of neurotoxicity (see Figure 5-5): 0.005 mg/kg-day from Mutti et al.
(1992). and 0.007 mg/kg-day from both JISA (1993) and Emara et al. (2010). All of the other
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-36 DRAFT—DO NOT CITE OR QUOTE
-------
1 sRfDs are within about 10-fold of the recommended RfD. This suggests that multiple effects
2 may occur at about the same exposures at which tetrachloroethylene begins to induce
3 neurotoxicity. These results also suggest that it is important to take into account effects from
4 tetrachloroethylene other than neurotoxicity when assessing the cumulative effects of multiple
5 exposures.
5.2.5. Previous Oral Assessment
6 EPA previously reported an RfD of 1 x 10 2 mg/kg-day (U.S. EPA. 1988). based on an
7 continuous equivalent NOAEL of 14 mg/kg-day in Buben and O'Flaherty (1985). and a
8 composite UF of 1,000 (10 for extrapolation from the rat to humans, 10 for human variation, and
9 10 for extrapolating to chronic exposure conditions).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-37 DRAFT—DO NOT CITE OR QUOTE
-------
""J
3 2 m
S. £ 3
Ci sj >i'
O * ~
as ^
>!
Si
% g
Si- ^ si
S o S
S5 5
*
°S o
^ s
¦I ^ a,
"a § |
S. ^
If ^ $T
>i
tO 0\
2 S3
u>
oo
~n
H
O
O
Q
M—
m
11111 1—i—i i 11111
CNS
UFcomj^lOOO
l Echeverria et al. (1995) [H]
UFcomj^lOOO
I Seeber (1989) [H]
UFcomj^lOOO
l Cav alleri et al. (1994) [H]
Kidney
_UFcom£^100^
UFcomp=300
I Mutti et al. (1992) [H]
-^0 J ISA (1993) [R]
UFcomp=300
-© JISA (1993) [
Liver
^^Fcom£^30^
-© JISA (1993) [
_UFcom£^300^
UFcomp=3000
w NTP (1986) [M]
Buben & O'Flaherty (1985) [M]
Immunological/Hematological
UFcomp=1QQQ
I Emara et al. (2010) [H]
Re p rod ucti ve/D e ve lo p me nta I
UFcomj^SOO
-© Beliles et al. (1980) [M]
"I—I I I I I 11 1 1—I I I I 111 1 1—I I I I I I I 1 1—I I I I I 11 1 1—I I I I I II
10
10
10
10
10
10J
m
g/kg/d
Figure 5-5. Comparison of candidate RfDs (black squares) supporting the RfD (grey vertical line) and sample RfDs (open squares)
for effects other the critical effect (CNS toxicity). Black circles = study/endpoint LOAEL in terms of human equivalent dose.
Open circles = study/endpoint NOAEL in terms of human equivalent dose. Species in each study is shown in brackets after
the reference (mouse: M; rat: R; human: H).
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
5 3 UNCERTAINTIES IN INHALATION REFERENCE CONCENTRATION (RfC)
AND ORAL REFERENCE DOSE (RfD)
The following discussion further characterizes uncertainties associated with the RfC or
RfD for tetrachloroethylene. As presented earlier in this section (see also Sections 5.1.2, 5.1.3,
5.2.2, and 5.2.3), the uncertainty factor approach, following EPA practices (U.S. EPA. 1994.
2002). was applied to PODs consisting of LOAELs from three epidemiologic studies of
neurological effects. Factors accounting for a number of uncertainties in the analyses were
adopted to account for extrapolating the POD, the starting point in the analysis: (1) to a no-
adverse-effect concentration or dose (NOAEL), given insufficient information to characterize
minimally adverse effect levels in the principal studies with benchmark dose modeling; (2) to a
diverse population of varying susceptibilities; and (3) to account for database deficiencies.
These extrapolations are carried out with default approaches given the paucity of experimental
tetrachloroethylene data to inform individual steps. As further explained below, limited
information is available on human variation in blood tetrachloroethylene concentration and can
provide rough estimates of the degree of human variation. Evaluation of a tetrachloroethylene
exposure dose or concentration likely to be without an appreciable risk of chronic adverse health
effects over a lifetime and associated uncertainties relies on chemical-specific data to describe
dose-response curves, on the breadth of the database for evaluating toxicity in a number of
organs, and on characteristics of these data.
A broad range of animal toxicology and human epidemiologic data is available for the
hazard assessment of tetrachloroethylene, as described throughout Section 4. These studies
include short-term and long-term bioassays in rats and mice; neurotoxicology studies in humans,
rats, mice, and gerbils; prenatal developmental toxicity studies in rats, mice, and rabbits and a
two-generation reproduction study in rats; and numerous supporting genotoxicity and
metabolism studies. Toxicity associated with inhalation exposure to tetrachloroethylene is
observed in the liver, kidney, central nervous system, reproductive organs, and the developing
fetus. Liver, kidney, and neurodevelopmental effects are observed with oral exposure.
Nevertheless, critical data gaps have been identified and uncertainties associated with data
deficiencies are more fully discussed below.
The neurotoxic effects observed in a residential population (Altmann etal.. 1995) are
similar to those observed in occupational populations exposed at higher mean
tetrachloroethylene concentration (Echeverria et al.. 1994; Seeber. 1989). Schreiber et al.
(NYSDOH. 2010; 2002) and Storm et al.fln Press) observed visual effects (visual contrast
sensitivity) among residents colocated near dry-cleaning establishments; however, these studies
had significant limitations related to subject selection and test methodology. Effects in the CNS
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-39 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
and in other organ systems (liver, kidney, reproductive, and developmental) are observed at
similar average tetrachloroethylene concentrations in occupational populations and at higher
average tetrachloroethylene concentrations in animals than in the three neurotoxicological
studies selected as principal studies (Cavalleri etal.. 1994; Echeverria et al.. 1995; Seeber. 1989).
As more fully discussed in Sections 5.1 and 5.2, uncertainties in other studies of neurotoxicity
differ from those in the three principal studies. For both occupational and residential
populations, studies do not describe a NOAEL and human variation is not well characterized in
study subjects. Uncertainties associated with the occupational studies include the following:
(1) potential for neurobehavioral effects at lower exposures and (2) exposure pattern differences
between occupational and residential studies with peaks characterizing occupational exposures.
For animal studies, uncertainties are associated with extrapolating from experiments in
genetically inbred rodents of high concentration and subchronic duration to infer a concentration
that is likely to be without appreciable risk of adverse health effects over a lifetime to a diverse
human population.
5.3.1. Point of Departure
A POD based on a LOAEL or NOAEL is, in part, a reflection of the particular exposure
concentrations or doses at which a study was conducted. It lacks characterization of the dose-
response curve and for this reason is less informative than a POD obtained from dose-response
modeling. With respect to neurotoxicity of tetrachloroethylene, the PODs are all LOAELs
because benchmark dose-response modeling was not feasible.
5.3.2. Extrapolation from Laboratory Animal Studies to Humans
Extrapolating from animals to humans embodies has further issues and uncertainties.
While this extrapolation was not necessary for the critical effects it was necessary for
comparison to some of the other organ toxicities. First, the effect and its magnitude associated
with the concentration at the POD in rodents is extrapolated to human response.
Pharmacokinetic models are useful to examine species differences in ADME. This was possible
for liver toxicity where limited MOA information suggests oxidative metabolism as important to
toxicity. The use of PBPK modeling for interspecies extrapolation with the dose metric of liver
oxidative metabolism increased the POD for liver effects by more than 10-fold as compared to
the use of applied dose. On the other hand, use of the AUC of tetrachloroethylene in venous
blood as a dose metric for other endpoints had a negligible impact as compared to the use of
applied dose. In the case of tetrachloroethylene-induced kidney effects, available data suggest
GSH conjugation to be involved. As described in Section 3.5, PBPK model-derived estimates of
dose metrics related to GSH conjugation of tetrachloroethylene span a very wide (> 1,000-fold)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-40 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
range when performing interspecies extrapolation, due to uncertainty, variability, or both.
Therefore, this dose metric was not used.
5.3.3. Human Variation
Heterogeneity among humans is another uncertainty associated with extrapolating doses
from animals to humans. Uncertainty related to human variation needs consideration, also, in
extrapolating dose from a subset or smaller sized population, say of one sex or a narrow range of
life-stages typical of occupational epidemiologic studies, to a larger, more diverse population.
Subjects in the epidemiologic studies comprise adults, and some characterization of the response
of children to tetrachloroethylene exposure was found in limited data for a similar neurological
(visual system) parameter (Schreiber et al.. 2002) and in a larger number of subjects (NYSDOH.
2005; Storm et al.. In Press) using other visually based testing paradigms. Additionally, in a
postnatal neurotoxicity study in mice (Fredriksson et al.. 1993). persistent neurological effects
(i.e., increased locomotion and total activity, and decreased rearing behavior at 60 days of age,
measured 43 days after exposure ceased) were observed at an oral dose of 5 mg/kg-day, with no
NOAEL.
In the absence of tetrachloroethylene-specific data on human variation, a factor of 10 was
used to account for uncertainty associated with human variation. Human variation may be larger
or smaller; however, tetrachloroethylene-specific data to examine the potential magnitude of
over- or under-estimation are few. As described in Section 3.5, the residual difference between
the PBPK model predictions and individual measurements of blood tetrachloroethylene in
humans had a geometric standard deviation of about twofold, suggesting that the ratio between a
median and 95th percentile measurement would be about threefold. This is consistent with
EPA's standard division of the human variability UF into threefold for toxicokinetics and
threefold for toxicodynamics. However, the available human toxicokinetic data are in healthy
adult volunteers, and may underestimate the degree of variability in the full population and
across life-stages. Limited quantitative analyses have been performed evaluating
pharmacokinetic variation between adults and children for tetrachloroethylene and its
metabolites using PBPK models (Clewell and Andersen. 2004; Gentry et al.. 2003; Pelekis et al..
2001). However, the authors indicated that validation of these results for various life-stages and
further refinement of the parameters in the model are necessary before the results of such an
analysis can be considered for use in risk assessment. In addition, as described in Section 3.5,
PBPK model predictions for GSH conjugation span a wide range that may be due to uncertainty,
variability, or both. GSTs are known to be polymorphic in the human population, with some
isoforms exhibiting a substantial population of null phenotypes. Therefore, human variability
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-41 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
associated with GSH conjugation may be substantially more than implied by the default UF.
With respect to toxicodynamics, no data are available to inform the degree of human variability.
5.3.4. Database Uncertainties
The following critical data gaps have been identified: uncertainties associated with
database deficiencies on neurological, developmental, and immunological effects.
As described above in Section 5.1.3, the three occupational studies (Cavalleri et al.. 1994;
Echeverria et al.. 1995; Seeber. 1989) used to derive the RfC evaluated neurotoxicity following
occupational exposures with PODs 3- to 100-fold higher than those identified from residential
studies (Altmann et al.. 1995; NYSDOH. 2010; Schreiber et al.. 2002; Storm et al.. In Press). In
comparison to the occupational studies, the available residential studies were judged to more
limited for developing an RfC, based on consideration of the study design (population
comparability) and/or selection of neurological methods (see Table 5.2). However, they provide
human evidence of neurotoxicity following tetrachloroethylene exposure in a residential setting,
with reaction time deficits, visual system dysfunction and cognitive performance deficits.
In addition, data characterizing dose-response relationships and chronic visuospatial
functional deficits and the cognitive effects of tetrachloroethylene exposure under controlled
laboratory conditions are lacking. Data from acute studies in animals (Oshiro et al.. 2008;
Umezu et al.. 1997; Warren et al.. 1996) suggest that cognitive function is affected by exposure
to tetrachloroethylene. These studies do not address the exposure-response relationship for
subchronic and chronic tetrachloroethylene exposures on cognitive functional deficits observed
in humans (e.g.. Altmann et al.. 1995; Echeverria et al.. 1995; Seeber. 1989). Even more
importantly, there is a lack of cognitive testing in both developmentally exposed animals and
adult animals following exposures to tetrachloroethylene that are longer than acute durations of
exposure. Visual system dysfunction and processing of visuospatial information are sensitive
endpoints in human studies. The exposure-response relationship of these functional deficits
could be evaluated more definitively with studies using homologous methods that examine
retinal and visual function in experimental animals. However, there has been a limited
evaluation of chronic exposure to tetrachloroethylene on visual function in rodents, with the
exception of the evoked potential studies by Mattsson et al. (1998). These types of studies could
help determine whether there are both peripheral and central effects of tetrachloroethylene
exposure on visual perception, and they could be used as an animal model to better define the
exposure-response relationships in humans.
Finally, additional data are needed to assess the potential hematological and
immunological effects of tetrachloroethylene. In humans, (Emara et al.. 2010) reported changes
in various standard hematological measures in subjects with mean tetrachloroethylene blood
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-42 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
levels of 1.685 mg/L. In addition, reversible hemolytic anemia was observed in female mice
exposed to very low drinking water levels of tetrachloroethylene (0.05 mg/kg-day) beginning at
2 weeks of age in one series of studies (Marth. 1987; Marth et al.. 1985; 1989). Although
additional corroborating studies are lacking, the observation of an effect at a very low exposure
level raises additional concern about hematological and immunological effects. The fact that
other solvents [e.g., toluene, and the structurally similar solvent trichloroethylene (Cooper et al..
2009)1 have been associated with immunotoxicity contributes further concern about this gap in
the database for tetrachloroethylene.
5 4 CANCER DOSE-RESPONSE ASSESSMENT
The following dose-response assessment was developed following the Guidelines for
Carcinogen Risk Assessment (U.S. EPA. 2005a). As discussed in Section 4.10.2,
tetrachloroethylene is characterized as —ikely to be carcinogenic in humans by all routes of
exposure," based on some epidemiologic evidence and conclusive evidence in mice and rats. No
available epidemiologic studies of cancer were found to be suitable for dose-response
assessment. Therefore, the following dose-response assessment is based on data from rodent
bioassays. Because the MO As for tetrachloroethylene carcinogenicity are not known, the tumors
reported in rodent bioassays are considered relevant to humans and a low-dose linear
extrapolation is used to estimate human cancer risk from rodent dose-response data.
5.4.1. Choice of Study/Data with Rationale and Justification
As discussed in Section 4, the several chronic exposure studies in rats and mice include an oral
gavage study in mice and female rats by National Cancer Institute (NCI. 1977) and
two inhalation studies in mice and rats (JISA. 1993; NTP. 1986b). These studies established that
the administration of tetrachloroethylene, either by ingestion or by inhalation to sexually mature
rats and mice, results in increased incidence of tumors. Mouse liver tumors (hepatocellular
adenomas and carcinomas) and rat mononuclear cell leukemia (MCL) were reported in both
sexes in two lifetime inhalation bioassays employing different rodent strains, and mouse liver
tumors were also reported in both sexes in an oral bioassay (NCI. 1977). Tumors reported in a
single inhalation bioassay include kidney and testicular interstitial cell tumors in male F344 rats
(NTP. 1986a). brain gliomas in male and female F344 rats (NTP. 1986a). and hemangiomas or
hemangiosarcomas in male Crj:BDFl mice (JISA. 1993).
This analysis considers all three bioassays but focuses primarily on the JISA (1993) study
results. The NCI (1977) oral gavage study in Osborne-Mendel rats was considered to be
inconclusive because of the high incidence of respiratory disease, and high mortality with
tetrachloroethylene exposure. Lesions indicative of pneumonia were observed in almost all rats
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-43 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
at necropsy. A high incidence of toxic nephropathy was evident in tetrachloroethylene-exposed
male and female rats. Early mortality was also seen in tetrachloroethylene-exposed animals;
50% of the high dose males and females had died by Weeks 44 and 66, respectively. Regarding
the NCI (1977) gavage study in mice, several issues contribute to judging the results to be less
useful for quantitative risk assessment than the inhalation studies. First, dosing lasted 78 weeks
rather than 104 weeks as in the inhalation studies. Thus, in making direct comparisons, it might
be expected that the observed tumor incidence in the NCI (1977) study would underestimate the
incidence associated with 104 weeks of exposure. Second, the dosing schedule was variable, and
doses were increased by 100 mg/kg-day in the low-dose group and by 200 mg/kg-day in the
high-dose group after 11 weeks of study. Consequently, while time-weighted averages and
PBPK modeling provide means for estimating the effective level of exposure, the actual
correspondence of exposure with the observed effects is less clear. Further, mortality was
significantly increased in both treated groups over that of controls, suggesting that the maximum
tolerated dose had been exceeded. Therefore, dose-response modeling of the NCI (1977) rat and
mouse bioassay data was not conducted.
The JISA (1993) bioassay was used for dose-response modeling of rodent cancer
endpoints also seen with higher exposures in the earlier NTP (1986b) bioassay. The lower
exposure of both mice and rats in the JISA bioassay and the use of three, rather than
two, exposure groups provides a stronger basis for deriving dose-response relationships for risk
assessment purposes, insofar as all other aspects of these studies can be considered comparable.
For mice, the lowest and middose exposure concentrations in the JISA (1993) study were 10- and
twofold lower, respectively, than the lower exposure concentration (100 ppm) in the NTP
(1986b) inhalation study. For rats, the low-exposure concentration in the JISA (1993) study was
fourfold lower than in the NTP study (200 ppm). The JISA (1993) bioassay was also used for
dose-response modeling of the increased hemangiomas and hemangiosarcomas primarily in
spleen, liver, skin and adipose tissue of male mice, since it was the only bioassay that reported
this tumor type. Therefore, for most endpoints including liver tumors, mononuclear cell
leukemias and hemangiosarcomas, the JISA (1993) study was used for dose-response modeling.
The NTP (1986b) study was utilized for modeling the increased incidence in renal cancers, brain
cancers and testicular tumors with treatment reported only in this bioassay. The sections below
summarize the rodent tumor findings and additional considerations for data set selection.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-44 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
5.4.2. Dose-Response Data
5.4.2.1. Liver Tumors in Mice
All three bioassays showed increases in hepatocellular tumors in male and female mice.
Table 5-13 summarizes these incidence patterns. Because hepatic adenomas and carcinomas are
considered part of the same continuum of tumor development, and adenomas may be
differentiated from carcinomas only on the basis of size, this analysis emphasizes the combined
incidence of these two tumor types. Historical data from the Japan Bioassay Research Center
(JBRC), where the JISA (1993) study was conducted, indicate that the liver tumor incidences in
the control group were fairly typical for this laboratory (see Table 5-14). Specifically, the
incidence in controls was 28% for males and 6% for females; the averages for the laboratory
were 23 and 2% and the upper bounds were 42 and 8%, respectively, for carcinomas.1
The liver tumor results of the two inhalation studies are reasonably concordant for both
male and female mice when adjusted for background tumor incidence (see Figure 5-6). The
incidence among male mice in the JISA (1993) study did not follow a clearly monotonic pattern,
with a higher response in the lowest dose group than that in the next higher dose group.
However, when considering the degree of expected variability given the number of animals in
each dose group, this pattern appeared consistent with the overall supralinear dose-response
patterns for the male and female mice in the NTP (1986b) and JISA (1993) studies.
The NCI (1977) study, in addition to the dosing and duration limitations noted
above, only reported hepatocellular carcinomas but not adenomas. This was consistent with
other NCI study reports of that time. Because, as stated above, hepatic adenomas and
carcinomas are considered part of the same continuum of tumor development, the other two
bioassays provide a more complete evaluation of hepatocarcinogenesis associated with
tetrachloroethylene exposure.
1 Combined historical incidence of adenomas or carcinomas was not available. Presumably the incidence of
carcinomas slightly underestimates the overall incidence of adenomas or carcinomas.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-45 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Table 5-13. Tumor incidence in mice exposed to tetrachloroethylene
Bioassay
Doses/exposures
Sex
Body
weight"
kg
Survival-adjusted tumor
incidenceb (%)
Administered
Continuous
equivalent
Hepatocellular adenomas or carcinomas
NCI (1977)°
B6C3Fi mice
Gavage:
5 d/wk,
78 wk
Vehicle control
450 mg/kg-day
900
0e mg/kg-day
332
663
Male
0.030
2/20
32/48
27/45
(10)
(67)
(60)
Vehicle control
300 mg/kg-dayd
600
0e
239 mg/kg-day
478
Female
0.025
0/20
19/48
19/45
(0)
(40)
(42)
NTP (1986b)
B6C3Fi mice
Inhalation:
6 hr/d,
5 d/wk,
104 wk
0 ppm
100
200
0 ppm
18
36
Male
0.037
17/49
31/47
41/50
(35)
(70)
(82)
0 ppm
100
200
0 ppm
18
36
Female
0.032
4/45
17/42
38/48
(9)
(40)
(79)
JISA (1993)
Crj:BDFl mice
inhalation:
6 hr/d,
5 d/wk,
104 wk
0 ppm
10
50
250
0 ppm
1.8
9.0
45
Male
0.048
13/46
21/49
19/48
40/49
(28)
(43)
(40)
(82)
0 ppm
10
50
250
0 ppm
1.8
9.0
45
Female
0.035
3/50
3/47
7/48
33/49
(6)
(6)
(15)
(67)
Hemangiosarcomas", liver or spleen
JISA (1993)
0 ppm
10
50
250
0 ppm
1.8
9.0
45
Male
0.048
4/46
2/49
7/48
11/49
(4)
(2)
(13)
(18)
Note: Data sets carried through dose-response modeling shown in bold.
"Average body weight reached during adulthood.
bAnimals dying before the first appearance of the tumor of interest but no later than week 52 were omitted from the
totals because these animals were presumed not to have adequate time on study to develop tumors.
°No adenomas were reported in this study.
dGavage doses listed were increased after 11 weeks by 100 mg/kg-day in each low-dose group or by 200 mg/kg-day
in each high-dose group. Animals surviving the 78-week exposure period were observed until the week 90 study
termination. Lifetime average daily (administered) doses (LADDs) were calculated as follows:
LADD (mg/kg-day) = Cumulative administered dose (mg/kg)/(total days on study)
= {[(initial dose rate x 11 weeks) + (later dose rate x 67 weeks)]/90 weeks}ftet x 5/7
(days)
"These tumors were reported as hemangioendotheliomas in the JIS A (1993) report. The term has been updated to
hemangiosarcoma. Note that these incidences do not match those tabulated in Tables 11, 12 of the JISA report
summary. The incidences reported here represent a tabulation of hemangioendotheliomas in liver or spleen from
the individual animal data provided in the JISA report.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-46 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-14. Historical control data of the Japan Bioassay Research Center,
Crj/BDFl mouse, 104-week studies
Tumor types
Inhalation, feeding, and drinking
studies (19 studies)
Inhalation studies only (9 studies)
Total incidence
(%)
Range (%)
Total incidence
(%)
Range(%)
Male mice
Liver
hepatocellular adenoma
hepatocellular
carcinoma
165/947 (17.4)
215/947 (22.7)
4.0-34.0
2.0-42.0
92/448 (20.5)
105/448 (23.4)
10.0-30.6
10.0-36.7
Spleen
hemangioma3
hcmangiosarcoma1
17/946 (1.8)
30/946 (3.2)
0-10.0
0-8.0
8/448 (1.8)
12/448 (2.7)
0-8.0
0-6.0
Female mice
Liver
hepatocellular adenoma
hepatocellular
carcinoma
50/949 (5.3)
22/949 (2.3)
2.0-10.0
0-8.0
18/449 (4.0)
14/449 (3.1)
2.0-6.0
0-8.0
Spleen
hemangioma3
hemangiosarcoma3
8/949 (0.9)
3/949 (0.3)
0-6.0
0-2.0
5/449 (1.1)
3/449 (0.7)
0-6.0
0-2.0
1
2 aThe terms —iamangioendothelioma: benign" and -hemangioendothelioma" in the original study have been changed
3 to -hemangioma" and -hcmangiosarcoma." respectively.
4
5 Source: Attachment to letter dated September 5, 2001, from K. Nagano, Japan Bioassay Research Center, Japan
6 Industrial Safety and Health Association, to R. McGaughy, U.S. EPA. Available from hotline.iris@epa.gov.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-47 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
^.°-9
NTP, 1986
J ISA, 1993
NTP, 1986
JiSA, 1993
<5 0.7
¦a
"0 0.5
~Q 0.5
a> 0.4
W 0.2
female mice
male mice
Continuous equivalent concentration (ppm)
Continuous equivalent concentration (ppm)
Figure 5-6. Mouse liver tumor responses (hepatocellular adenomas or
carcinomas), as additional risk, for two chronic inhalation bioassays (see
Table 5-13), plotted against continuous equivalent concentration (ppm), for
male and female mice.
5.4.2.2. Other Tumor Sites in Male Mice
In addition to elevations in hepatocellular adenomas and carcinomas, the JISA (1993) study
demonstrated increases in hemangiomas and hemangiosarcomas. These tumors were seen
primarily in spleen and liver, with several instances also reported in subcutaneous skin and in
adipose tissue, in mid- and high-dose male mice (Cochran-Armitage trend test
[two-sided], = 0.008). The incidence of spleen hemangiosarcomas in control and low-dose male
mice—2/46 and 2/49, respectively, each about 4%—was similar to the JBRC historical control
incidence for spleen only (3.2%, range 0-8%; see Table 5-14). The increase in this tumor type
with tetrachloroethylene exposure was not replicated in the NCI (1977) or NTP (1986b) studies.
In the NTP male mice, hemangiomas or hemangiosarcomas were only reported in liver; the
incidences were as follows: controls, 3/49 (6%); low dose, 2/49 (4%); high dose, 2/50 (4%),
within the range of NTP historical controls incidence for all sites, 2-8% (average 4.4%)
(http://ntp.niehs.nih.gov/ntp/research/database searches/historical controls/path/m inhar.txt).
The reasons the two bioassays differ with regard to identifying increases in hemangiomas and
hemangiosarcomas have not been elucidated; differences may be due to the strain of mouse used
or other factors. For this endpoint, therefore, the JISA (1993) study was selected for
dose-response modeling.
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/11 5-48 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
5.4.2.3. Mononuclear Cell Leukemia in Rats
The NTP (1986b) and JISA (1993) studies demonstrated increased MCL incidences for
male and female F344/N or F344/DuCrj rats (see Table 5-15). Although the NCI study, in
Osborne-Mendel rats, did not demonstrate any MCL increases, this study is considered
inconclusive because of low survival, and for other reasons noted above in Section 5.4.1.
The responses in the NTP (1986b) study were approximately twofold higher than for the
corresponding groups in the JISA (1993) study in all groups, including controls. Control groups
for both laboratories were consistent with their respective historical controls (see Table 5-16 for
the JISA historical controls). Like the hepatocellular tumor results in mice (see Section 5.4.2.1),
the MCL results from the NTP and JISA studies were plotted in terms of additional risk versus
administered concentration to evaluate relative increases in tumor incidence (see Figure 5-7).
The NTP and JISA studies are consistent for male rats at the administered concentration of
Table 5-15. Incidence of mononuclear cell leukemia, kidney tumors, and
brain gliomas in rats exposed to tetrachloroethylene by inhalation
Bioassay
Exposure concentration (ppm)
Sex
Body
weight8 (kg)
Su rvival-adj u sted
tumor incidenceb
%
Administered
Continuous
equivalent
Mononuclear cell leukemia
NTP (1986b)
0
0
Male
0.44
28/50
(56)
F344/N rats
200
36
37/48
(77)
inhalation
400
71
37/50
(74)
6 hr/d,
5 d/wk,
0
0
Female
0.32
18/50
(36)
200
36
30/50
(58)
104 wk
400
71
29/50
(60)
JISA (1993)
0
0
Male
0.45
11/50
(22)
F344/DuCrj rats
50
8.9
14/50
(28)
inhalation
200
36
22/50
(44)
6 hr/d,
600
110
27/50
(54)
5 d/wk,
0
0
Female
0.3
10/50
(20)
104 wk
50
8.9
17/50
(34)
200
36
16/50
(32)
600
110
19/50
(38)
Kidney: tubular cell adenoma or adenocarcinoma
NTP (1986b)
0
200
400
0
36
71
Male
0.44
1/49
3/47
4/50
(2)
(6)
(8)
Brain gliomas
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-49 DRAFT—DO NOT CITE OR QUOTE
-------
NTP CI986b")
0
0
Male
0.44
1/50
(2)
200
36
0/48
(0)
400
71
4/50
(8)
Testicular interstitial cell tumors
NTP CI986b")
0
0
Male
0.44
35/50
(70)
200
36
39/47
(83)
400
71
41/50
(82)
1
2 Note: Data sets carried through dose-response analysis shown in bold.
3 "Average body weight reached during adulthood.
4 bAnimals dying before the first appearance of the tumor of interest but no later than week 52 were omitted from the
5 totals because these animals were presumed to have had inadequate time on study to develop these tumors.
6
7 Sources: NTP (1986b) and JISA (19931.
8
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-50 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-16. Historical control data of the Japan Bioassay Research Center,
F344/DuCrj (Fischer) rat, 104-week studies
1
Tumor types
Inhalation, feeding, and drinking
studies (23 studies)
Inhalation studies only
(11 studies)
Total incidence (%)
Range (%)
Total incidence (%)
Range (%)
Male rats
Mononuclear cell
leukemia
147/1,149(12.8)
6.0-22.0
76/549 (13.8)
6.0-22.0
Kidney
Renal cell adenoma
Renal cell carcinoma
2/1,149 (0.2)
2/1,149 (0.2)
0-2.0
0-2.0
1/549 (0.2)
2/549 (0.4)
0-2.0
0-2.0
Female rats
Mononuclear cell
leukemia
147/1,048 (14.0)
2.0-26.0
68/448 (15.2)
8.0-20.0
Kidney
Renal cell adenoma
Renal cell carcinoma
1/1,048 (0.1)
0/1,048 (0.0)
0-2.0
NA
1/448 (0.2)
0/448 (0.0)
0-2.0
NA
2
3 Source: Attachment to letter dated September 5, 2001, from K. Nagano, Japan Bioassay Research Center, Japan
4 Industrial Safety and Health Association, to R. McGaughy, U.S. EPA. Available from hotline.iris@epa.gov.
5
6
female rats
Continuous equivalent concentration (ppm)
Continuous equivalent concentration (ppm)
Figure 5-7. Rat mononuclear cell leukemia responses (minus control) in
two chronic bioassays (see Table 5-15), plotted against continuous equivalent
exposure (ppm) for (a) male and (b) female rats.
8
9
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/11 5-51 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
200 ppm (36-ppm continuous equivalent) in terms of the relative increases in tumors over
background incidences. For female rats, the dose-response patterns are less similar. A higher
overall response is seen in the NTP study. However, the JISA female rats have a steeper
increase at the lowest exposure level (50 ppm administered concentration, 9 ppm continuous
equivalent) than would be expected based on the NTP study, which did not include that exposure
level. Both studies suggest some degree of saturation of effects in the range of exposures
considered (see Figure 5-7).
Overall, the NTP and JISA studies show concordant MCL responses for both male and
female F344 rats. F344 rats were used in both studies, so residual differences could be
attributable to the specific lines of animals used at each laboratory and to laboratory-specific
procedures. As discussed in Section 5.4.1, the JISA study rather than the NTP study was
selected for dose-response modeling because it provides data on tumor incidences at lower
exposure and the use of three exposures provides a strong basis for dose-response analyses.
5.4.2.4. Other Tumor Sites in Rats
Additional tumor findings in rats included a significant increase in the NTP bioassay of
two rare tumor types, kidney tumors in males and brain gliomas in both sexes of exposed F344/N
rats. The NTP (1986b) bioassay also reported increases in the rate of testicular interstitial cell
tumors, a tumor type of high incidence in unexposed male F344 rats. Table 5-15 summarizes the
incidence data for these tumor sites.
The potential significance of the NTP brain tumor findings is supported by their relative
rarity (evidenced by a low historical control incidence) and earlier occurrence with increasing
tetrachloroethylene exposure, indicating an effect of exposure on latency. In males,
tetrachloroethylene-induced brain tumors were seen beginning at week 88 compared with
week 99 in controls. Female brain tumors were first seen at 75 weeks in tetrachloroethylene-
exposed animals compared with 104 weeks in control group females. Additionally, the nervous
system is known to be a target of tetrachloroethylene exposure in humans and animals (see
Sections 4.5.3 and 5.1.1). Therefore, although the overall incidences are low relative to other
tumor sites, and the finding was not replicated in the JISA study, the rarity of rat brain tumors in
control animals and the additional data suggesting biological plausibility support dose-response
modeling of this tumor type.
The evidence for kidney tubule cell adenomas and adenocarcinomas differed slightly between
the two bioassays (see Table 5-15). The JISA study showed no apparent trend among incidences
compared with either concurrent or historical controls (see Table 5-16). In contrast, the elevation
in exposed male rats in the NTP study, while not statistically significant when compared with
concurrent controls, was significant when compared using a trend test with the historical control
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-52 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
rate for the same facility (p = 0.0002, Cochran-Armitage, two-sided trend test). The
investigators noted the relative rarity of these tumors, with incidences of 1/249 among historical
controls for the study facility, and of about 0.2% in 1968 untreated controls in the NTP program
overall. Further support for the significance of the kidney tumors comes from evidence that the
related chemical trichloroethylene induces this tumor type in humans and in male rats (U.S.
EPA. 2009). Additional biological plausibility for this endpoint includes toxicokinetic data that
nephrotoxic and mutagenic metabolites are formed in the kidney following tetrachloroethylene
exposure. Therefore, although the overall incidences are low relative to other tumor sites, the
rarity of rat kidney tumors in control animals and the additional data suggesting biological
plausibility support dose-response modeling of this tumor type. The NTP (1986b) study was
better suited for modeling because it had a stronger trend, and was therefore selected for dose-
response modeling.
The NTP (1986b) study also reported an increase in the rate of testicular
interstitial cell tumors, a tumor type of high incidence in unexposed F344 rats. The reported
incidence of testicular interstitial cell tumors in male rates exposed to 0, 200 or 400 ppm
tetrachloroethylene was 36/50, 39/49, and 41/50, respectively. A higher incidence (47/50, or
92%) was seen in control rats in the JISA (1993) study than in the NTP (1986b) study. In the
JISA study, exposure to 50, 200, or 600 ppm tetrachloroethylene resulted in incidences of 47/50,
46/50, 45/50, and 48/50, respectively. Support for the significance of the testicular interstitial
cell tumors comes from evidence that the related chemical trichloroethylene induces this tumor
type in rats. Trichloroethylene did not induce increases in testicular interstitial cell tumors in the
F344 rat in a bioassay with a reported incidence of 47/48 (98%) in the vehicle control. However,
increases were seen in male Marshall rats, in which the incidence was 16/46, 17/46, 21/33, and
32/39 in untreated, vehicle control, 500, or 1,000 mg/kg-day trichloroethylene, respectively.
Therefore, although the overall increases in incidence are low relative to other tumor sites, the
additional data suggesting biological plausibility support dose-response modeling of this tumor
type.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-53 DRAFT—DO NOT CITE OR QUOTE
-------
Administered dose in
inhalation/oral animal
bioassay
^Animal PBPK model
Lifetime average daily dose metric
Preferred Dose Metrics Alternative Dose Metrics
AUCof
Rate of
Rate of
tetra-
AUCof
kidney
liver
chloro-
TCAin
GSH
oxidation
ethylene
blood
conjuga-
in blood
tion
i Fit dose response model
to observed response
POD in units of lifetime
average daily dose
metric
^ BMRtPOD
Slope Factor in units of
risk/(lifetime average
__dail^dosejrietric^_
Slope Factors as
risk/(Human Equivalent
nfetim^dail^dosenne^
If dose metric is
rate of oxidation or
of conjugation,
apply BW3/4 scaling.
Otherwise assume
equal AUCs.
^ Human PBPK model
Slope Factor or Unit Risk as risk/(Human
Equivalent continuous inhalation or oral
__environmental_exjoosu^
Figure 5-8. Sequence of steps for extrapolating from tetrachloroethylene
bioassays in animals to human-equivalent exposures expected to be
This document is a draft for review purposes only and does not constitute Agency policy.
1 5-54 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
associated with comparable cancer risk (combined interspecies and route-to-
route extrapolation). See Table 5-17 for units.
5.4.3. Dose Adjustments and Extrapolation Methods
This section provides details of the dose-response modeling carried out for developing cancer
risk values. The steps include estimation of dose metrics using relevant PBPK modeling (see
Section 3; Chiu et aD. suitable adjustment to continuous daily exposures from intermittent
bioassay exposures, dose-response modeling in the range of observation, interspecies
extrapolation, extrapolation to low exposures, and route-to-extrapolation. An overview of these
steps is provided in Figure 5-8. The schematic also addresses route-to-route extrapolation using
the Chiu and Ginsberg PBPK model, since after the slope factor is expressed in terms of risk per
unit of internal human dose, the PBPK model can be used to estimate the risk per unit of oral or
inhalation exposure, regardless of the route of administration in the original study.
5.4.3.1. Estimation of Dose Metrics for Dose-Response Modeling
Several factors inform the criteria for selection of dose metrics in this assessment: the
association of the metric with the toxic moiety relevant to the endpoint under consideration, the
availability of data and models for estimating that metric, and whether the resulting estimate is
sufficiently robust. When PBPK modeling is used, it is generally preferable to use a single
model for estimating all the dose metrics for dose-response modeling.
5.4.3.1.1. Hepatocellular tumors
Several metabolites of tetrachloroethylene are carcinogenic in mice, and it is thought that
the hepatocarcinogenicity of the parent compound is mediated through the action of one or more
of its metabolites. Oxidative metabolism is thought to predominate in the liver, and TCA is the
major resultant urinary excretion product. As discussed in Section 3, TCA appears to be formed
from spontaneous decomposition of trichloroacetyl chloride, which is known to bind to
macromolecules. Dichloroacetic acid (DCA) may be formed from dechlorination of TCA, but
DCA produced from this pathway is likely to be rapidly metabolized in the liver and not detected
in blood or urine. DCA that has been detected in urine is thought to be the result of kidney-
specific P-lyase metabolism of the results of GSH conjugation of tetrachloroethylene, and DCA
produced from this pathway is presumed to not play a role in liver toxicity or cancer. The
potential role of GST conjugates of tetrachloroethylene in liver carcinogenicity, although
unknown, is presumed to be less important that the role of oxidative metabolites.
The focus of most hypotheses with respect to contributors to tetrachloroethylene
hepatocarcinogenicity has been on TCA and, to a lesser extent, DCA. Data supporting the
conclusion that TCA and DCA, alone and in combination, are hepatocarcinogenic in rodents is
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-55 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
summarized in Tables 4-17, 4-18, and 4-19. In mice, TCA significantly increased the incidence
of liver tumors in male and female B6C3Fi mice exposed via drinking water for 52-104 weeks
(2002; 1990; 2004; DeAngelo et al.. 2008; Herren-Freund etal.. 1987; Pereira. 1996; Pereira and
Phelps. 1996). Incidence of tumors increased with increasing TCA concentrations (Bull et al..
2002; 1990; DeAngelo et al.. 2008; Pereira. 1996). These results were obtained under conditions
where the background incidence of tumors in control animals was generally low. The
development of tumors in animals exposed to TCA progressed rapidly, as evidenced by
significant numbers of tumors in less-than-lifetime studies of 82 weeks or less. Positive
evidence for tumor promotion by TCA (following exposure to known tumor initiators) has been
reported for liver tumors in B6C3Fi mice (Pereira et al.. 2001; Pereira et al.. 1997) and for
gamma-glutamyltransferase-positive foci in livers of partially hepatectomized Sprague-Dawley
rats (Parnell et al.. 1988). DCA also causes liver cancer in mice (1990; 2004; Daniel et al.. 1992;
DeAngelo et al.. 1999; Herren-Freund et al.. 1987). DCA and TCA are also hepatocarcinogenic
in mice when coadministered in the drinking water for 52 weeks (Bull et al.. 2004). Treatment-
related liver tumors were observed in male F344/N rats exposed via drinking water to DCA
(DeAngelo et al.. 1996) but not TCA (DeAngelo et al.. 1997) for 60 or 104 weeks. The
carcinogenicity of TCA and DCA has not been evaluated in female rats or in other species of
experimental animals.
Data on tumor phenotype support the view that TCA may not be the sole tumorigenic
metabolite of tetrachloroethylene, but also do not provide definitive evidence testing any
particular hypothesis. For instance, liver tumor genotypes (e.g., with regard to H-ras codon
61 mutation) and phenotypes (e.g., with regard to c-Jun staining) appear to differ among tumors
induced by TCA, DCA, the combination of TCA and DCA, and the structurally related
compound trichloroethylene (Bull et al.. 2002). Bull et al. (2002) suggest that for
trichloroethylene, the data are not consistent with the hypothesis that TCA is the sole active
moiety, but a similar experiment has not been conducted for tetrachloroethylene. However, by
analogy, it is possible that TCA and DCA, in combination with each other (and with other
reactive intermediates produced during the oxidative metabolism of tetrachloroethylene) may
contribute to the production of liver tumors. This appears to be the case for noncancer effects, as
the spectrum of endpoints caused by tetrachloroethylene includes effects broader than that
produced by TCA, and including fatty degeneration, focal necrosis and regenerative repair, some
of which may play a role in liver carcinogenesis (see Section 4.3.5).
The hepatocarcinogenic potencies of TCA and tetrachloroethylene have not been
directly compared in a single rodent bioassay. Appendix C presents a comparative quantitative
analysis of the carcinogenicity of TCA (including that predicted using PBPK modeling to be
produced from tetrachloroethylene) with the carcinogenicity of tetrachloroethylene. Statistically,
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-56 DRAFT—DO NOT CITE OR QUOTE
-------
1 this analysis did not reject the hypothesis of equivalent carcinogenic potencies of TCA and the
2 internal dose of TCA resulting from tetrachloroethylene exposure. However, power calculations
3 show that even if TCA only accounted for half of the potency of tetrachloroethylene, the
4 available data are unlikely to reject such a hypothesis. In addition, several factors, including the
5 much higher control incidence of liver tumors and the relatively high body weights of the
6 animals in the TCA bioassay, limit the direct comparability of the tetrachloroethylene and TCA
7 bioassay data. Therefore, this analysis is only of limited utility in elucidating the contribution of
8 TCA to tetrachloroethylene hepatocarcinogenic potency.
9 In consideration of these uncertainties, total rate of oxidative metabolism in the liver is
10 the most relevant dose-metric for tetrachloroethylene-induced liver toxicity. AUC for TCA in
11 the liver is also presented as a plausible alternative dose metric. The PBPK-derived estimates of
12 liver total oxidative metabolism and TCA AUC corresponding to the JISA bioassay exposures
13 for male and female mice are provided in Table 5-17.
Table 5-17. Summary of PBPK-derived dose metric estimates used for
dose-response analysis of rodent tumor data
14
Study
group
Administered
concentration
ppm
Liver total oxidative
metabolism
mg/kg/4-daya
Tetrachloroethylene
AUC in blood
mg-hr/L-db
TCA AUC in liver
mg-hr/L-dc
Total GSH
metabolism
mg/kg/4-dayd
Males
Females
Males
Females
Males
Females
Males
Females
Mice,
JISA
(1993)
0
10
50
250
0
2.25
8.25
33.6
0
2.13
7.75
31.6
0
4.11
22.3
116
0
4.18
22.6
117
0
78.5
280
1120
0
77.0
272
1090
Not used
Rats,
JISA
(1993)
0
50
200
600
Not used
0
20.0
80.9
247
0
20.1
81.4
248
Not used
Not used
Rats,
NTP
(1986b)
0
200
400
Not used
0
81.0
164
0
81.3
164
Not used
0.00
0.303
0.615
15
16 aPrimary dose metric for mouse hepatocellular tumors.
17 bPrimary dose metric for mouse hemangiomas or hemangiosarcomas, rat MCLs, rat kidney tumors, rat brain
18 gliomas, and rat testicular tumors.
19 Alternative dose metric for mouse hepatocellular tumors.
20 Alternative dose metric for rat kidney tumors.
21
22
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-57 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
5.4.3.1.2. Mononuclear Cell Leukemia
Tetrachloroethylene causes mononuclear cell leukemia in rats. Regarding the metabolites that
potentially contribute to MCL development, a role for GSH-derived intermediates was posited
based on findings for the related compound trichloroethylene in bovine species. However S-
(l,2,2,-trichlorovinyl)-Z-cysteine (TCVC), a GSH-derived metabolite of tetrachloroethylene,
induced no kidney or bone marrow effects when administered to two calves as a single dose
(Lock et al.. 1996). Aside from this evaluation of bone marrow toxicity of TCVC in the juvenile
cow, a species of unknown sensitivity to tetrachloroethylene-induced leukemia, other studies
aimed at elucidating the active metabolites contributing to leukemic effects have not been
reported. In particular, no such studies are available in the F344 rat, the species and strain in
which leukemic effects have been consistently observed in both sexes. It is thus concluded that
the specific active moiety(ies) by which tetrachloroethylene induces this type of tumor are not
known.
In summary, because considerable uncertainty surrounds the identification of the
causative chemical species, AUC of the parent compound in the blood is considered a viable
dose metric for MCL, and has the advantage of being a more proximal dose than administered
dose. The estimates of tetrachloroethylene AUC in blood corresponding to the JISA bioassay
exposures for male and female rats are provided in Table 5-17.
5.4.3.1.3. Kidney tumors
Tetrachloroethylene causes tubular toxicity in mice and rats, and is associated with small
increases in the incidences of kidney tumors reported in multiple strains of rats (JISA. 1993;
NTP. 1986b). These effects, including kidney cancer, are thought to be associated with
tetrachloroethylene metabolism by GSH conjugation, based on the production in the kidney of
nephrotoxic and genotoxic metabolites from this pathway (Lash and Parker. 2001). As noted in
Section 3, the PBPK model by Chiu and Ginsberg allows calculation of this dose metric. GSH
conjugation occurs in the kidney as well as in the liver from where the metabolic products may
be transported to the kidney. Therefore, the most appropriate dose metric for kidney toxicity
would be the total rate of metabolism of tetrachloroethylene via the GSH conjugation pathway.
However, overall the estimates of GSH conjugation in Chiu and Ginsberg were highly
uncertain and/or variable, and to a very different extent across species (also see Section 3).
Uncertainty in this estimate was the least, roughly twofold, in rats. In mice, the range of
estimates based on the different optimization runs was about 10-fold. In the human, the range of
predicted estimates spanned several orders of magnitude. In particular, two local maxima were
seen for the posterior modes, each of which the fit to the data was good and substantially similar.
However, the model predictions corresponding to each estimate differed by 3,000-fold. It was
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-58 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
not clear as to whether this 3,000-fold spread represented uncertainty or variability in the form of
a bimodal distribution for human GSH conjugation or both (see Section 3 for a discussion of
plausible reasons for a multimodal distribution).
In view of this large uncertainty/variability, and the inability to differentiate uncertainty
from variability, it appears more prudent to use AUC of the parent compound in the blood as a
preferred dose metric for kidney toxicity. This has the advantage of being a more proximal dose
to the kidney than administered dose. Total rate of metabolism of tetrachloroethylene via the
GSH conjugation pathway is also used as an alternative dose metric. PBPK-derived estimates of
tetrachloroethylene AUC in blood and total GSH metabolism corresponding to the JISA (1993)
male rat exposures are provided in Table 5-17.
5.4.3.1.4. Other dose metrics
No data are available concerning the metabolites that may contribute to the induction of
other rodent tumor types, including hemangiosarcomas or hemangiomas in male mice, kidney
tumors and testicular interstitial cell tumors in male rats or brain gliomas in male and female rats.
It is concluded that the specific active moiety(ies), mechanisms or modes of action by which
tetrachloroethylene induces these rodent tumor are not known. Accordingly, AUC of
tetrachloroethylene in the blood was used for these tumors because it is more proximal to the
target tissues than administered dose (see Table 5-17 for dose estimates used for dose-response
modeling).
In addition, all tumor sites considered for modeling were also modeled using
administered inhalation concentration, for comparison purposes. These concentrations (in ppm)
were adjusted for continuous exposure by averaging the five 6-hour daily exposures over the full
week, by multiplying by 6 hours/24 hours x 5 days/7 days (0.179) to yield equivalent continuous
concentrations. Tables 5-5 and 5-7 provide these adjusted concentrations.
5.4.3.1.5. Uncertainties in PBPK modeling and dose metrics
A detailed discussion of uncertainties in the dosimetric estimates, derived using a PBPK
model that considered all the available tetrachloroethylene PK data in the literature, was
provided in Sections 3.5.1.2.2 and 3.5.1.2.3. A full Bayesian analysis of the
uncertainty/variability was not performed. Nonetheless, the range of posterior modes provided
for the various dose metrics in Section 3.5.1.2.2 provides an estimate of the range of uncertainty
associated with each dose metric, which in turn results in a range of human unit risk estimates
associated with each dose metric used for any given end point in Tables 5-18 and 5-20.
In particular, the predictions for GSH conjugation in humans were found to be highly
uncertain. In the rat, the ranges of chain-specific posterior modes for GSH conjugation spanned
up to twofold, and in mice up to 10-fold. However, in humans, the ranges spanned several
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-59 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
orders of magnitude, reflecting the two —cl uters" of posterior modes with estimates of GSH
conjugation clearance differing by up to 3,000-fold. Tetrachloroethylene AUC was associated
with a twofold pharmacokinetic uncertainty/variability. The range in estimates of
tetrachloroethylene oxidation in humans was found to be largely dominated by a twofold
interindividual variability.
In terms of the selection of dose metric, tetrachloroethylene is metabolized to several
intermediates with carcinogenic potential. Although much data exist for TCA, they are
inadequate to support the conclusion that TCA alone is able to explain the hepatocarcinogenicity
associated with tetrachloroethylene exposure. Whether total oxidative metabolism, total GSH
metabolism, or tetrachloroethylene AUC in blood—either as measures of a precursor or
intermediate or as surrogates directly proportional to the toxic agent(s)—are adequate indicators
of potential risk is unclear. A role for the parent compound has not been ruled out, nor is it clear
whether the specific active moiety(ies) are proportional to administered concentration.
5.4.3.2. Extrapolation Methods
5.4.3.2.1. Dose-Response models and extrapolation to low doses
As discussed in Section 4.10.2, the available body of MO A information is not sufficient
to derive biologically based quantitative models for low-dose extrapolation. No key events in
the tumor development process for tetrachloroethylene have been identified that would
determine the overall dynamics of such a model, nor are there experimental data specific to
tetrachloroethylene describing any of the underlying toxicodynamic processes, such as cell
replication rates.
The multistage model has been used by EPA in the vast majority of quantitative cancer
assessments, initially because of its parallelism to the multistage carcinogenic process. A benefit
of the multistage model is its flexibility in fitting a broad array of dose-response patterns,
including allowing linearity at low dose. Occasionally the multistage model does not fit the
available data, in which case alternate models should be considered. The related multistage-
Weibull model has been the preferred model when individual data are available for time-to-
tumor modeling, which incorporates more of the information about response than does the
simpler dichotomous response model. Use of this decision scheme has contributed to greater
consistency among cancer risk assessments.
The multistage model is given by:
P(d) = q0 + (I - qo) x [ 1 -expi-X, = i q, x d)\ (5-1)
where:
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-60 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
d = exposure level (including internal dose metric) and
q, = parameters estimated in fitting the model, q, > 0; n is degree of the model
The multistage model in BMDS [Benchmark Dose Software, version 2.1.1 (U.S. EPA.
2009a)] was used initially to fit all data sets. Using the method of maximum likelihood, all
feasible orders of the multistage model up to the number of dose groups (n) less one were
evaluated for fit. Model fits with goodness-of-fit^-values >0.05 are generally considered
acceptable, with good visual fit and evaluation of standardized residuals for the control group
and points near the benchmark dose (the dose corresponding to a predetermined increase above
control levels, or BMD) also important. Among the model fits satisfying these criteria, the most
parsimonious model fit was generally selected.
Two tumor sites with statistically significantly decreased time to tumor were noted: brain
gliomas in NTP male rats and MCL in the NTP female rats, especially for the most severe stage
of leukemia observed (Stage 3). The multistage-Weibull model, given by the following
equation, was also used to evaluate the importance of decreased time to tumor and intercurrent
mortality in interpreting these responses.
q„ z = parameters estimated in fitting the model; q, > 0, z > 1; n is degree of the model,
The multistage-Weibull model is the same as the multistage model when z = 0. MSW (U.S.
EPA. 2010) was used for all multistage-Weibull model fits.
Following dose-response modeling in the range of observation, the cancer risk values for
extrapolation to low doses were derived from the lower bound on the concentration (BMCL)
associated with a level of risk from the low end of the observed range, usually 10% extra risk.
Extra risk has been used consistently throughout EPA risk assessments and is given by:
P(d,t) = q0 + (1 - go) x [ 1 -expi-^i i :nq, x d) x f]
(5-2)
where:
d = exposure level (or dose metric)
t = time to observation of the tumor
Extra risk = \P(d) - .P(O)] / [1 - .P(O)]
(5-3)
where:
P(d) = estimated response at exposure d and
P{0) = estimated response in the control group
6/24/11
This document is a draft for review purposes only and does not constitute Agency policy.
5-61 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
The slope factor (risk per mg/kg-day for oral exposure, risk per dose metric unit for PBPK-
modeled dose metrics) and risk per unit concentration (risk per mg/L for drinking water
exposure, or per |ig/m3 for inhalation exposure) are estimated using linear extrapolation from the
PODs because of the lack of information supporting another extrapolation approach (U.S. EPA.
2005a). by dividing the risk level by its associated BMCL:
Risk/(unit of exposure) = Extra risk/BMCL. (5-4)
5.4.3.2.2. Uncertainties in low-dose extrapolation approach
The MOA is a key consideration in clarifying how risks should be estimated for low-dose
exposure. However, MOA data are lacking or limited for all candidate cancer endpoints for
tetrachloroethylene (i.e., rat MCL, brain, testicular and kidney tumors, mouse hepatocellular
tumors and hemangiosarcomas). When the MOA cannot be clearly defined, EPA uses a linear
approach to estimate low-exposure risk, based on the following broad and long-held scientific
assumptions, which supported the development of the EPA's Guidelines for Carcinogen Risk
Assessment (U.S. EPA. 2005a).
• 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 incidence of various tumors.
• A broadening of the dose-response curve in the human population (less rapid fall-off with
dose) and, accordingly, a greater potential for risks from low-dose exposures (see Lutz et
al.. 2005; Zeise et al.. 1987) would result for two reasons. First, even if there is a
threshold concentration at the cellular level, that threshold is likely to be different among
different individuals. Secondly, greater variability in response to exposures in the
heterogeneous human population would be anticipated than in controlled laboratory
species and conditions (due to, e.g., genetic variability, disease states, age).
• The use of linear extrapolation provides consistency across assessments as well as
plausible upper-bound risk estimates that are believed to be health-protective (U.S. EPA.
2005a).
•
The overall uncertainty in low-dose risk estimation could be reduced to some degree if
the MOA for tetrachloroethylene were known with a high degree of confidence. However, even
in such a case, incorporation of MOA into dose-response modeling might not be straightforward
and might not significantly reduce the uncertainty about low-dose extrapolation. This is because
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-62 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
in addition to the MO A, other factors, such as human response variability, may strongly
influence the dose-response function in humans.
A number of different biological motivations have been put forward to support functional
forms that might be used to estimate risks from low-dose exposure to carcinogens or other toxic
substances. For cancer, the most prominent class of models, including the multistage model used
in this assessment, treats tumorigenesis as a multievent process and characterizes the probability
of accumulation of a series of changes (conceptualized as mutations or other events) that,
together, will result in formation of a malignant tumor.
The concept of a distribution of individual thresholds is a second approach used to
motivate functional forms for dose-response modeling. Such models assume that there is an
—individual threshold" for each member of the human population, and interindividual variation in
these thresholds determines the dose-response curve for a population. A recent National
Research Council report on risk assessment issues for TCE (NRC. 2006) included a discussion of
models based on distributions of thresholds. That report noted that if one assumes a normal or
logit distribution for individual thresholds this leads to a probit or logistic dose-response function
for the population and suggests that a variety of other distributions for thresholds would also lead
to sigmoidal shaped dose-response functions. The NRC report expressed the view that,
—A hough linear extrapolation has been advocated as an intentionally conservative approach to
protect public health, there are some theoretical reasons to think that sublinear nonthreshold
dose-response models may be more relevant for human exposure to toxicants, regardless of the
mode of action" (p. 319). On the other hand, the same report also noted that a very broad class
of dose-response functions can be obtained using distributions of thresholds models: -4n fact any
monotonic dose-response model, including the linearized multistage model, can be defined
solely in terms of a tolerance distribution without resorting to mechanistic arguments. These
considerations suggest that one must consider both the role of mode of action and the role of
response variability among humans in determining the likely shape of the dose-response
function" (p. 323).
The discussion from NRC (2006) emphasizes some key points in risk assessment.
Variability in the human population will have an important influence on the shapes of the dose-
response relationships for that population. This is distinct from the amount of variability that
may be observed in inbred animal strains. As noted in the NRC report, —0e might expect these
individual tolerances to vary extensively in humans depending on genetics, coincident exposures,
nutritional status, and various other susceptibility factors..." (p. 320). Thus, if a distribution-of-
thresholds approach is considered for a carcinogen risk assessment, application would depend on
the ability of modeling to reflect the degree of variability in response in human populations. By
design, most cancer bioassays are conducted in inbred rodent strains; accordingly, the parameters
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-63 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
provided by curve fits of distribution-of-thresholds models to bioassay data would not be
predicted to reflect the dose-response patterns in diverse human populations. It is important to
note that the NRC text has no recommendation for an approach where a tolerance distribution
model for humans is estimated by a statistical fit to rodent bioassay data.
The question of whether a tolerance distribution model is indeed an appropriate basis for
a risk assessment also warrants consideration. Low-dose linearity can arise in other contexts
distinct from effects of population variability and may be directly appropriate to a MO A. Low-
dose linearity can also arise due to additivity of a chemical's effect on top of background
chemical exposures and biological processes. In the case of chemicals such as
tetrachloroethylene, basic biological data do not exist to support the appropriateness of an
individual threshold model above models having inherent low-dose linearity. However, if
distribution of thresholds modeling were supported, it would need to be developed based on an
examination of predicted variability within in human population.
Given the current state of scientific knowledge about tetrachloroethylene carcinogenicity,
the straight line based risk estimates presented above form the preferred recommendation for
estimating a plausible upper-bound estimate of potential human risks from tetrachloroethylene.
This approach is supported by both general scientific considerations, including those supporting
the Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005a). as well as chemical-specific
findings. The former include the scientific principles articulated above—the expectation that a
chemical functions additively to background exposures, diseases, and processes; that variability
within the human population would broaden the dose-response curve and may eliminate
population thresholds if present; and that the approach provides consistency across assessments
facilitating direct comparison of the derived risk values.
5.4.3.2.3. Extrapolation to human equivalent environmental exposure
For extrapolation of risk to humans, this assessment used two approaches that were
dependent on the relevant dose metric: the EPA RfC methodology (U.S. EPA. 1994). which
applies when chemical-specific pharmacokinetic data are lacking, and EPA's cross-species
scaling methodology (Rhomberg. 1992). The discussions below include a consideration of
uncertainties inherent in each of these approaches.
5.4.3.2.3.1. Internal dose metrics
Because of the availability of PBPK modeling to estimate a plausible dose metric either
in terms of specific metabolites or metabolic pathways or blood concentration of the parent
compound in both laboratory rodents and humans, extrapolation to human equivalent
environmental exposure entailed the steps as shown in Figure 5-8. First, consistent with the
2005 cancer guidelines (U.S. EPA. 2005a). EPA's methodology for cross-species scaling
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-64 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
(Rhomberg. 1992) was considered when toxicological equivalence for the relevant tumor sites
was addressed, in order to convert the slope factor to units of risk per unit of human equivalent
internal dose metric. Then the slope factor was converted to units of risk per unit of human
equivalent environmental exposure by using the relationship between continuous human
exposure and internal dose metric estimated via the human PBPK model. These last
two considerations are further described below.1
EPA's cross-species scaling methodology, grounded in general principles of allometric
variation of biologic processes, was used for describing toxicological equivalence because of the
extensive empirical evidence supporting it (Crump et al., 1987). Briefly, in the absence of
adequate information to the contrary, the methodology determines toxicological equivalence
across species through equal average lifetime concentrations or AUCs of the carcinogen. One
typical application of this methodology is to oral exposures in mg/kg-day in the absence of
pharmacokinetic or pharmacodynamic information. However, the same principles apply to the
parent compound and metabolites (Rhomberg, 1992).
For the orally administered dose, the correspondence of equal AUCs is equivalent to
considering the exposures in terms of mg/kg3/4-day, and is achieved by multiplying animal
exposures by (BWanimai/BWhuman)1'4, based on the principle that clearance on average scales
allometrically according to BW 1 4 across species (U.S. EPA. 2005a). Note that this equivalence
across species entails the cross-species correspondence of internal doses in terms of AUCs or
mg/kg3/4-day, which is implicit in the frequent default case, i.e., oral carcinogens without
chemical-specific pharmacokinetic data. In other words, each time a carcinogen is scaled from
animals to humans on the basis of mg/kg3/4-day, an implicit assumption is that internal doses are
equipotent in terms of mg/kg3/4-day (—crcs-species scaling"), not mg/kg-day (—bodyweight
scaling").
Accordingly, when pharmacokinetic data are available that relate administered
concentration to enzymatically derived metabolites of the carcinogen, this methodology is still
applicable; internal doses, as a fraction of administered dose, should still tend to produce
equivalent effects when considered in terms of AUCs (when clearance of a specific metabolite is
specifically modeled) or mg/kg3/4-day (when rate of metabolism is calculated) because
metabolites are also subject to scale-affected clearance processes. The equivalence of
considering equal AUCs of a metabolite to scaling the rate of metabolism by BW3'4 can be easily
1 Typically, the POD would be expressed in terms of a human equivalent exposure. However, in this case, it is
expressed in terms of the internal dose metric. This is because the relationship between exposure and internal dose
may be nonlinear at the POD, even if the relationship between risk and internal dose is assumed to be linear below
the POD. Therefore, the slope factor is first expressed in terms of internal dose, reflecting the assumption of low-
dose linearity in internal dose. Then, provided the slope factor is applied at exposures well below the POD, where
the relationship between exposure and internal dose is linear, it can be converted to a risk per unit exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-65 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
understood if one assumes clearance rates for the metabolite scales allometrically according to
adjusted by the fraction metabolized. There is a wide body of empirical evidence that metabolic
rates associated with enzymatic processes scale with body weight to the 3/4 power (U.S. EPA,
BW 1 4 (U.S. EPA. 2005a) or if one thinks of the scaling as applied to the administered dose
(Rhomberg. 1992)). Furthermore, when this scaling is applied to an internal dose expressed as a
rate of production of metabolite(s), it is applicable regardless of the route of exposure. As an
example, in EPA's trichloroethylene assessment (external draft), the human equivalent risk for
liver and kidney effects was estimated using BW3'4 scaling of the daily rate of the toxicologically
relevant metabolic pathway (U.S. EPA. 2009).
As discussed earlier in this subsection, rates of liver oxidative metabolism and total GSH
metabolism are considered plausible dose-metrics for the liver and kidney, respectively. In order
to estimate equivalent toxic effects in humans using the cross-species scaling methodology,
tetrachloroethylene metabolized via either of these pathways was scaled using BW3 4 so that the
dose metric was expressed as mg/kg3/4-day. As explained earlier, the AUC of TCA in the liver,
the predominant metabolite along the oxidative pathway, is also presented as a plausible dose
metric for liver cancer. No additional scaling was needed as the average concentration of TCA
so determined was assumed to be equipotent when applied continuously over a lifetime in either
species. Likewise, AUC of tetrachloroethylene, used as the preferred dose metric for MCL and
kidney tumors, was not scaled further to extrapolate to humans.
Note that the involvement of reactive metabolites cleared nonenzymatically through
which all other metabolites may follow has been hypothesized in many cases, and scaling by BW
as opposed to BW3 4 has been proposed to be more appropriate in such cases. However, scaling
by BW was not considered pertinent for tetrachloroethylene because the possible reactive
metabolites cleared nonenzymatically have not been identified and because the majority of the
metabolites formed are thought to be sufficiently stable to be cleared enzymatically.
In the last step of the extrapolation to risk per human equivalent exposure, the slope
factors in terms of internal dose metrics (associated with parent or metabolites) were converted
to slope factors or unit risks in terms of human equivalent environmental inhalation and oral
exposures using pharmacokinetic modeling. See footnote c in Tables 5-10 and 5-12 for the
inhalation and oral conversion factors. For animals, the study-specific body weights were used
(see Tables 5-5 and 5-7), and for humans the default of 70 kg was used.
In summary, an adjustment for cross-species scaling (BW3/4) was applied to address
toxicological equivalence of internal doses between each rodent species and humans for
two dose metrics, total liver oxidative metabolism and total GSH metabolism, consistent with the
2005 Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005a). It is assumed that, without
data to the contrary, equal risks result from equivalent constant exposures. While the true
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-66 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
correspondence of equipotent tetrachloroethylene exposures across species is unknown, the use
of BW3 4 scaling is expected neither to over- or underestimate human risk, based on allometry
(Rhomberg. 1992).
5.4.3.2.3.2. Administered inhaled concentration as dose metric
For those sites for which pharmacokinetic-adjusted doses were not available or not
otherwise relevant, EPA's default RfC methodology was used (U.S. EPA. 1994).
Tetrachloroethylene is considered a Category 3 gas because it is water soluble and perfusion
limited, and it has systemic (extrarespiratory) effects. Because the ratio of blood/air partition
coefficients for the experimental animal species relative to humans is greater than or equal to one
(for F344 rats, 15.1/14.7= 1.03; for B6C3Fi mice, 18.6/14.7= 1.3), a default value of one was
used for this ratio (U.S. EPA. 1994). Consequently, when administered inhalation
concentrations were used as the dose metric, the concentrations were considered equipotent
across species for extrapolating risk to humans. Therefore, no further extrapolation was
necessary with the resulting PODs in the units of human equivalent environmental exposure
levels.
In summary, for MCL, hemangiomas or hemangiosarcomas, and brain, testicular, and
kidney tumors, tetrachloroethylene AUC in blood was judged to be more proximal than
administered tetrachloroethylene concentration to the adverse effect, and therefore more relevant
for estimating unit risks. Also based on allometry, average daily AUCs are expected to be
equipotent across species without any additional scaling involved. The true correspondence is
unknown, and risk may be higher or lower in humans than in rodents to an unknown degree.
5.4.4. Cancer Risk Values
Human cancer risk was assessed using four different sex-species animal data sets and a
PBPK model for interspecies and route-to-route extrapolation. In all cases, linear extrapolation
from the PODs was carried out because of the lack of information supporting another
extrapolation approach (U.S. EPA. 2005a). For each dataset, multistage modeling using
preferred and alternative (if available) PBPK model-based dose metrics was conducted, in
addition to multistage modeling using administered concentration. The NRC (2010) peer review
recommended more extensive quantitative evaluation of the uncertainty due to different forms of
dose-response models. Moreover, NRC (2010) agreed that for several datasets, the multistage
model does not fit the data at lower doses, noting evidence of supralinearity in the underlying
dose-response relationship. NRC (2010) also noted that in such cases, low-dose linear
extrapolation is not conservative and the external review draft Toxicological Review did not
present the full ranges of variation and uncertainty in relation to model choice.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-67 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Therefore, for the JISA (1993) datasets, additional analyses were conducted using
administered concentration and the range of dichotomous models included in BMDS. In
addition to the multistage model, these include the Gamma, Weibull, LogLogistic, LogProbit,
dichotomous Hill, Probit, and Logistic models. For the dichotomous Hill model, the slope was
fixed at 1, making it equivalent to a Michaelis-Menten model. Statistically, the slope needed to
be fixed so that goodness of fit statistics could be derived given the number of dose groups (three
exposed plus one control). Biologically, the Michaelis-Menten model is a natural choice for
saturable biological processes, such as enzyme kinetics, that are not accounted for in the selected
dose metrics. Hereafter, the dichotomous Hill model with slope fixed at 1 is referred to as the
Michaelis-Menten model.
The results of the suite of models were evaluated for goodness-of-fit. For datasets
exhibiting supralinearity, models that led to both a better fit to the supralinear shape and a stable
BMDL were considered for further application using PBPK model-based dose metrics. The
choice of result best representing an upper bound estimate of human carcinogenic potency
among the derived values considered a number of factors, as described in Section 5.4.4.2. These
factors include the magnitude and robustness of the response, the role of metabolism, the
carcinogenic MO As, and the dose-response model fit and resulting low-dose extrapolation
predictions.
The sections below provide the results of the dose-response modeling using the male and
female mouse and rat data from the JISA (1993) inhalation bioassay and male and female rat
data from the NTP (1986b) inhalation bioassay. Route-to-route extrapolation for estimating
human cancer risk via oral exposure to tetrachloroethylene is then presented. Finally,
quantitative and qualitative uncertainties underlying the risk estimation process are discussed.
5.4.4.1. Dose-Response Modeling Results
5.4.4.1.1. Hepatocellular tumors, male mice
In accordance with standard practice in the absence of MO A data supporting a particular
dose-response model form, multistage modeling of the JISA bioassay data was carried out, using
the preferred and alternative dose metrics of total liver oxidative metabolism and TCA AUC in
liver. Modeling for both dose metrics generated fits for one-, two-, and three-stage models
(details in Appendix D). All model fits had adequate goodness-of-fit ^-values (p > 0.05), and
overall adequate fit given the nonmonotonicity in the observed dose-response range (with
standardized residuals within ± 2). There was no statistical improvement (by likelihood ratio
test) in adding higher order terms to the first-order term and a one-stage model was selected (see
Figure 5-9 for the fit using total oxidative metabolism).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-68 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Extrapolation to humans using total oxidative metabolism led to a BMDio of 2.9, and its
lower bound benchmark dose (BMDLio) was 1.4-fold lower at 2.1 mg/kg3/4-day liver oxidative
metabolism (see Figure 5-9). Linear extrapolation from the POD to low internal dose, followed
--3
by conversion to human exposures led to a human equivalent unit risk of 1.8 x 10 per ppm.
Extrapolation to humans using TCA AUC in liver led to a human equivalent internal dose
POD (BMCLio) of 69 mg-hr/L-day TCA in blood. The corresponding central tendency estimates
was approximately 1.5-fold higher, at 97 mg-hr/L-day. Linear extrapolation from the POD to
low internal dose, followed by conversion to human exposures led to a human equivalent unit
-3
risk of 1.5 x 10 per ppm, slightly lower than the estimate using total liver oxidative
metabolism.
Dose-response modeling of the male mouse liver tumor data using administered exposure
fit the data points similarly to when using total oxidative metabolism or TCA AUC in liver
(details in Appendix D). The result was directly interpretable as a human equivalent POD
(BMCLio), at 2.7 ppm tetrachloroethylene in air. The corresponding central tendency estimate
was nearly twofold higher, at 3.9 ppm. Linear extrapolation from this POD led to a human
-3
equivalent unit risk of 37 x 10 per ppm, more than an order of magnitude higher than using
either PBPK-estimated dose metric.
The NRC (2010) peer review recommended more extensive quantitative evaluation of the
uncertainty due to different forms of dose-response models. The analysis was conducted using
administered concentration and the range of dichotomous models included in BMDS. Among
the models fitted, five models fit worse than the multistage (Gamma, Weibull, LogLogistic,
LogProbit, and Michaelis-Menten), and two models fit better than the multistage (Probit and
Logistic). However, the multistage model had the lowest residual for the control group,
indicating that the alternative models were no better than the multistage model in addressing the
supralinear shape in this dataset. Nonetheless, the estimated BMCLi0s from the better fitting
models were less than twofold different than that using the multistage model.
Therefore, due the limited sensitivity to the selection of dose-response models and the
finding that none of the alternative models was clearly superior to the standard multistage model
for addressing this dataset's supralinearity at the lower doses, the multistage model results were
carried forward to support cancer risk estimates (Table 5-18). Due to the data supporting
oxidative metabolism as being involved in hepatocellular tumors, the estimates carried forward
were those using total oxidative metabolism as the dose metric (preferred), and those using TCA
AUC in liver as the dose metric (alternative). The remaining analyses (Tables 5-19 and 5-20)
using administered concentration using multistage and other dose-response models are retained
only to better characterize the range of results from different dose-response models.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-69 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
5.4.4.1.2. Hepatocellular tumors, female mice
As was done for the male mouse hepatocellular tumors, in accordance with standard
practice in the absence of MO A data supporting a particular dose-response model form,
multistage modeling of the JISA bioassay data was carried out, using the preferred and
alternative dose metrics of total liver oxidative metabolism and TCA AUC in liver. Modeling
for both dose metrics included one-, two-, and three-stage models. Adequate fits were obtained
with all three models, with adequate goodness-of-fit^-values (p > 0.05), and overall adequate
visual fit (see details in Appendix D). The second order term led to a statistically significant
improvement in fit, but there was no statistical improvement with the third order term, as it was
estimated to be zero. Therefore a two-stage model was selected for both dose metrics (see
Figure 5-10 for the fit using total oxidative metabolism).
Extrapolation to humans using total liver oxidative metabolism led to a human equivalent
internal dose POD (BMCLio) of 3.9 mg/kgy4-day liver oxidative metabolism. The corresponding
central tendency estimate was 2.2-fold higher, at 8.4 mg/kgy4-day. Linear extrapolation from the
POD to low internal dose, followed by conversion to human exposures led to a human equivalent
-3
unit risk of 0.92 x 10 per ppm.
Extrapolation to humans using TCA AUC in liver led to a human equivalent POD
(BMCLio) of 139 mg-hr/L-day TCA in blood. The corresponding central tendency estimate was
approximately 2.1-fold higher, at 292 mg-hr/L-day. Linear extrapolation from the POD to low
internal dose, followed by conversion to human exposures led to a human equivalent unit risk of
-3
0.73 x 10 per ppm, slightly lower than the estimate using total liver oxidative metabolism.
Dose-response modeling using administered exposure fit the data points similarly to
when using total oxidative metabolism or TCA AUC in liver (details in Appendix D). The result
was directly interpretable as a human equivalent POD (BMCLio), at 3.8 ppm tetrachloroethylene
in air. The corresponding central tendency estimate was approximately twofold higher, at
-3
5.0 ppm. Linear extrapolation from this POD led to a human equivalent unit risk of 27 x 10
per ppm, more than an order of magnitude higher than using either PBPK-estimated dose metric.
The NRC (2010) peer review recommended more extensive quantitative evaluation of the
uncertainty due to different forms of dose-response models. The analysis was conducted using
administered concentration using the range of dichotomous models included in BMDS. All the
models (Gamma, Weibull, LogLogistic, Michaelis-Menten, LogProbit, Probit, and Logistic) fit
similar to or better than the multistage. The estimated BMCLi0s from the better fitting models
were less than threefold different than that using the standard multistage model.
Therefore, due to the limited sensitivity to the selection of dose-response models, the
multistage model results were carried forward to support cancer risk estimates (see Table 5-18).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-70 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Due to the data supporting oxidative metabolism in hepatocellular tumors, the estimates carried
forward were those using total oxidative metabolism as the dose metric (preferred), and those
using TCA AUC in liver as the dose metric (alternative). The remaining analyses (see
Tables 5-19 and 5-20) using administered concentration using multistage and other dose-
response models are retained only to better characterize the range of results from different dose-
response models.
5.4.4.1.3. Hemangiosarcomas, male mice
Hemangiosarcomas were also observed in the JISA male mice, in liver, spleen, fat, and
subcutaneous skin. Because these tumors differ etiologically from the hepatocellular adenomas
and carcinomas, they were modeled separately. In accordance with standard practice in the
absence of MO A data supporting a particular dose-response model form, multistage modeling of
the JISA bioassay data was carried out, using the preferred dose metric of tetrachloroethylene
AUC in blood, including fits for one-, two-, and three-stage models (details in Appendix D). A
one-stage model was found to be sufficient, with an adequate goodness-of-fit /> value (p = 0.38),
and overall adequate visual fit (see Figure 5-11). There was no statistical improvement in fitting
higher order models, as all the higher order parameters were estimated to be zero.
Multistage CancerModel with 0.95 Confidence Level
Multistage Cancer
Linear extrapolation
0.9
0.7
-o
0)
"G
0)
0.6
<
c
o
0.5
o
2
Ll_
0.4
0.3
0.2
45
0
5
10
15
20
25
30
35
40
dose
Figure 5-9: Dose-response modeling of male mouse hepatocellular tumors
associated with inhalation exposure to tetrachloroethylene, in terms of liver
total oxidative metabolites; response data from JISA (1993). Details in
Appendix D.
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/11 5-71 DRAFT—DO NOT CITE OR QUOTE
-------
M uitistage Cancer Model with 0.95 Confidence Level
Multistage Cancer
Linear extrapolation
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
P.MP
10
BMDL
0
5
15
20
25
30
dose
Figure 5-10. Dose-response modeling of female mouse hepatocellular tumors
associated with inhalation exposure to tetrachloroethylene, in terms of liver
total oxidative metabolites; response data from JISA (1993). Details in
Appendix D.
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/11 5-72 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-18. Human equivalent unit risks, derived using PBPK-derived dose
metrics and multistage model; tumor incidence data from JISA (1993) and
NTP (1986b)
Study Group
Tumor type
(multistage model with
all dose groups unless
otherwise specified)
Human Equivalents
PODa, in internal dose units
SFxlO"3
/internal
dose
unitb
IURxlO"3
/ppm (PBPK
range)0
Primary dose metrics
Male mice
JISA ri993^
Hepatocellular adenomas
or carcinomas
BMDio
BMDLio
2.9
2.1
Total liver oxidative
metabolism, mg/kg0 75-d
49
1.8
(1.6-1.8)
Hemangiomas,
hemangiosarcomas,
BMDio
BMDLio
63
34
PCE AUC in blood,
mg-hr/L-d
2.9
5.9
(5.9-6.9)
Female mice
JISA ri993^
Hepatocellular adenomas
or carcinomas
BMDjo
BMDLio
8.4
4.0
Total liver oxidative
metabolism, mg/kg0 75-d
25
0.90
(0.84-0.93)
Male rats
jisa (1993)
MCL
BMDio
BMDLio
46
30
PCE AUC in blood,
mg-hr/L-d
3.4
6.8
(6.8-8.0)
Female and male
rats JISA M993^
MCL (Michaelis-
Menten)
BMDio
BMDLio
20
5.0
PCE AUC in blood,
mg-hr/L-d
20
40
(40-47)
Male rats
NTP (1986b)
MCL
BMDio
BMDLio
136
61
PCE AUC in blood,
mg-hr/L-d
1.6
3.3
(3.3-3.9)
MCL (control and low
dose groups only)
BMDio
BMDLio
11
5.2
PCE AUC in blood,
mg-hr/L-d
19
39
(39-45)
MCL (Michaelis-
Menten)
BMDio
BMDLio
17
3.0
PCE AUC in blood,
mg-hr/L-d
33
68
(67-71)
Alternate Dose
Metrics
Male mice
JISA ri993^
Kidney tumors
BMDio
BMDLio
246
110
PCE AUC in blood,
mg-hr/L-d
0.90
1.8
(1.8-2.1)
Brain gliomas
BMDio
BMDLio
400
192
PCE AUC in blood,
mg-hr/L-d
0.62
1.3
(1.2-1.5)
Female mice
JISA (1993)
Testicular interstitial cell
tumors
BMDio
BMDLio
31
14
PCE AUC in blood,
mg-hr/L-d
7.1
14
(14-17)
MCL
BMDio
BMDLio
28
15
PCE AUC in blood,
mg-hr/L-d
6.6
13
(13-16)
Total risk for any of
above four tumor types
BMDio
BMDLio
14
8.2
PCE AUC in blood,
mg-hr/L-d
12
25
(25-29)
Male rats
NTP Q 986b)
Male mice
JISA (1993)
Hepatocellular aden-
omas or carcinomas
BMDio
BMDLio
97
69
TCA AUC in liver,
mg-hr/L-d
1.5
1.5
(1.4-1.5)
Hepatocellular aden-
omas or carcinomas
BMDio
BMDLio
292
141
TCA AUC in liver,
mg-hr/L-d
0.72
0.72
(0.68-0.74)
Female mice
JISA (1993)
Kidney tumors
bmd05
bmdl05
0.46
0.21
Total GSH metabolism,
mg/kg0.75-d
243
100
(0.047-110)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-73 DRAFT—DO NOT CITE OR QUOTE
-------
1
Table 5-18. Human equivalent unit risks, derived using PBPK-derived dose
metrics and multistage model; tumor incidence data from JISA (1993) and
NTP (1986b) (continued)
SF = Slope Factor; IUR = Inhalation Unit Risk; MCL= Mononuclear cell leukemias
aPODs were estimated at the indicated BMRs in terms of extra risk; i.e., BMDL10 = lower bound for the level of the
internal dose metric associated with 10% extra risk. Dose metric units are in the first column, and include cross-
species scaling to a human equivalent internal dose metric. See Appendix D for dose-response modeling details.
bSlope Factor = BMR/BMDLBmr in units of risk per dose metric unit (as given in the first column).
Inhalation unit risk (IUR) is given by the product of the slope factor in units of risk per dose metric unit and an
inhalation dose metric conversion factor (DMCFppm): IUR = BMR/BMDLBMr x DMCFppm, where the DMCFppm is
derived from the PBPK model. The DMCFppm for each dose metric is shown below:
Dose metric
DMCFppm
Overall posterior mode
Range of posterior modes
Total liver oxidative metabolism
0.0363
0.0339-0.0372
Tetrachloroethylene blood AUC
2.03
2.01-2.36
TCA AUC in liver
1.02
0.956-1.04
Total GSH metabolism
0.428
0.00019-0.44
Values in bold correspond to using the overall posterior mode, and are carried forward for consideration as the
recommended IUR. The difference between the overall and alternative posterior modes is negligible (relative to
other uncertainties) except for the Total GSH metabolism dose metric.
dSee Section 5.4.4.1.3 for calculation.
2
3
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-74 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-19. Dose-response summary and unit risk estimates using
continuous equivalent administered tetrachloroethylene levels as dose metric,
from NTP (1986b) and JISA (1993)
Study group
Tumor type
(multistage model and all
dose groups unless
otherwise specified)
POD (ppm)
Unit riska'b
x 10 3/ppm
Male mice
JISA (19931
Hepatocellular adenomas
or carcinomas
BMCio
BMCLjo
3.9
2.7
37
Hemangiomas or
hemangiosarcomas
BMCio
BMCL10
24
13
7.5
Overall risk of either
tumor type above0
BMCio
BMCL10
3.3
2.4
42
Female mice
JISA (1993)
Hepatocellular adenomas
or carcinomas
BMCio
BMCL10
5.0
3.8
27
Male rats
JISA (19931
MCL
BMCjo
BMCL10
21
13
7.6
MCL (Michaelis-Menten
model)
BMCjo
BMCLjo
8.6
2.2
45
Female rats
JISA (19931
MCL
BMCjo
BMCLjo
60
27
3.7
MCL (control and low
dose groups only)
BMCjo
BMCLjo
4.2
2.3
43
Female and male
rats JISA (19931
MCL
BMCjo
BMCLjo
32
21
4.8
MCL (Michaelis-Menten
model)
BMCjo
BMCLjo
7.7
1.4
71
Male rats
NTP (1986b1
Kidney tumors
BMCjo
BMCLjo
110
50
2.0
Brain gliomas
BMCjo
BMCLjo
180
73
1.4
Testicular interstitial cell
tumors
BMCjo
BMCLjo
13
6.1
16
Mononuclear cell
leukemia
BMCjo
BMCLjo
12
6.5
15
Overall risk for any of
above four tumor types0
BMCjo
BMCLjo
5.7
3.5
29
1 MCL = Mononuclear cell leukemia.
2 aUsing dose coefficients in terms of administered ppm of tetrachloroethylene adjusted to equivalent continuous
3 exposure, consistent with RfC methodology (U.S. EPA. 19941. and the multistage model, extra risk.
4 ''Unit risks, which are approximations for extrapolation to lower doses, should not be used with exposures greater
5 than the POD from which they were derived without considering the curvature of the dose-response function (see
6 Appendix D for modeling details).
7 °Overall risk estimated using maximum likelihood method. See Appendix D.3.1 for details.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-75 DRAFT—DO NOT CITE OR QUOTE
-------
1 Data source: See Tables 5-5, 5-7, Appendix D.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-76 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-20. Range of outputs from fitting different BMDS models using
continuous equivalent administered tetrachloroethylene levels as dose metric,
from JISA (1973)"
Study group
Tumor type
(all dose groups unless
otherwise specified)
Range of PODs (ppm)
Range of
0.1/BMCL10
x 10~3/ppm
Male mice
JISA (19931
Hepatocellular adenomas
or carcinomas
BMCio
BMCL10
2.5-11
0.4-4.8
21-250
Hemangiomas or
hemangiosarcomas
BMCio
BMCL10
16-32
4.1-22
4.5-24
Hepatocellular adenomas
or carcinomas
BMCio
BMCL10
5.0-13
3.8-11
9.4-27
Female mice
JISA (19931
MCL
BMCio
BMCLjo
6.9-30
0.062 - 22
4.5 - 1600
Male rats
JISA (19931
MCL
BMCio
BMCLio
4.9 - 88
0-37
2.7 - +co
Female and male
rats JISA (19931
Mononuclear cell
leukemia
BMCio
BMCLio
4.5-42
0.001-31
3.3-71
1
2 MCL = Mononuclear cell leukemia
3 aUsing dose coefficients in terms of administered ppm of tetrachloroethylene adjusted to equivalent continuous
4 exposure, consistent with RfC methodology (U.S. EPA. 19941. and extra risk. Range from use of different dose-
5 response models (Gamma, Weibull, LogLogistic, LogProbit, Michaelis-Menten, Probit, Logistic, and Multistage)
6 using all dose groups, only including models with goodness-of-fit /^-values >0.1.
7
8
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-77 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Multistage Cancer Model with 0.95 Confidence Level
0.4
Multistage Cancer
Linear extrapolation
0.35
0.3
0.25
0.2
0.15
.1
0.05
0
BMDL
BMD
0 20 40 60 80 100 120
dose
Figure 5-11. Dose-response modeling of male mouse hemangiomas or
hemangiosarcomas associated with inhalation exposure to
tetrachloroethylene, in terms of tetrachloroethylene AUC in blood; response
data from JISA (1993). Details in Appendix D.
Extrapolation to humans led to an internal dose POD (BMCLio) of 34 mg-hr/L-day
tetrachloroethylene in blood (see Table 5-18). The corresponding central tendency estimate was
nearly twofold higher, at 63 mg-hr/L-day. Linear extrapolation from the POD to low internal
dose, followed by conversion to human exposures led to a human equivalent unit risk of
-3
5.9 x 10 per ppm.
Dose-response modeling using administered exposure fit the data points similarly to when
using tetrachloroethylene AUC in blood (details in Appendix D). The result was directly
interpretable as a human equivalent POD (BMCLio), at 13 ppm tetrachloroethylene in air (see
Table 5-19). The corresponding central tendency estimate was approximately twofold higher, at
-3
24 ppm. Linear extrapolation from this POD led to a human equivalent unit risk of 7.5 x 10
per ppm, slightly higher than using tetrachloroethylene AUC in blood.
These results raise some concern that total cancer risk based on the male mice data may
be underestimated by considering only one site. Methods for estimating overall risk from sites
with very different dose metrics are not currently available. However, when an analysis using
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/11 5-78 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
administered concentration as the dose metric for both sites was carried out, using a method
based on maximum likelihood estimation1, the overall risk was estimated to be only slightly
higher than that using hepatocellular tumors alone (see Table 5-19). The analysis yielded an
overall risk value of 0.042 per ppm, compared with the unit risk of 0.037 based on hepatocellular
tumors alone. On the other hand, using administered concentration for the hepatocellular tumors
may substantially overestimate human equivalent risk as compared to that estimated by using
total liver metabolism, under the assumption that oxidative metabolism is likely an important
component of this process. See Appendix D. 1.1.3 for a summary of the calculations.
The NRC (2010) peer review recommended more extensive quantitative evaluation of the
uncertainty due to different forms of dose-response models. The analysis was conducted using
administered concentration using the range of dichotomous models included in BMDS. All of
the models had similar or worse fits than the multistage (Gamma, Weibull, LogLogistic,
LogProbit, and Michaelis-Menten, Probit, and Logistic). The estimated BMCLi0s ranged from
3.2-fold less to 1.7-fold more than that using the multistage model.
Therefore, due to the limited sensitivity to the selection of dose-response models, the
multistage model result was carried forward to support cancer risk estimates (Table 5-18). Due
to the lack of data on the active moiety for this endpoints, the result carried forward used AUC of
tetrachloroethylene in blood as the preferred dose metric. The remaining analyses (Table 5-19)
using administered concentration using multistage and other dose-response models are retained
only to better characterize the range of results from different dose-response models.
5.4.4.1.4. Mononuclear cell leukemia (MCL), male rat
In accordance with standard practice in the absence of MO A data supporting a particular
dose-response model form, multistage modeling of the JISA bioassay data was carried out
considering fits for one-, two-, and three-stage models (details in Appendix D). Using the
preferred dose metric of tetrachloroethylene AUC in blood, a one-stage model had a goodness-
of-fit/>value = 0.52, generally considered adequate, and the standardized residuals were within
the recommended limit of ±2 units (see Figure 5-12a). There was no statistical improvement in
1 An approach suggested in the EPA cancer guidelines to characterize total risk from multiple tumor sites would be
to estimate cancer risk from tumor-bearing animals. EPA traditionally used this approach until Science and
Judgment in Risk Assessment (NRC. 19941 made a case that this approach would tend to underestimate composite
risk when tumor types occur in a statistically independent manner— that is, that the occurrence of a
hemangiosarcoma, say, would not be dependent on whether there was a hepatocellular tumor. This assumption
cannot currently be verified and if not correct could lead to an overestimate of risk from combining across tumor
sites. However, NRC (1994) argued that a general assumption of statistical independence of tumor-type occurrences
within animals was not likely to introduce substantial error in assessing carcinogenic potency from rodent bioassay
data.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-79 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
fitting higher order models, as all the higher order parameters were estimated to be zero.
Extrapolation to humans led to an internal dose POD (BMCLio) of 30 mg-hr/L-day
tetrachloroethylene in blood (see Table 5-18). The corresponding central tendency estimate was
less than twofold higher, at 46 mg-hr/L-day. Linear extrapolation from the POD to low
exposures, followed by conversion to human exposures led to a human equivalent unit risk of
-3
6.8 x 10 per ppm.
Dose-response modeling using administered exposure fit the data points similarly to that
using tetrachloroethylene AUC in blood (details in Appendix D). The result was directly
interpretable as a human equivalent POD (BMCLio), at 13 ppm tetrachloroethylene in air (see
Table 5-19). The corresponding central tendency estimate was approximately twofold higher, at
-3
21 ppm. Linear extrapolation from this POD led to a human equivalent unit risk of 7.6 x 10
per ppm, very similar to that using tetrachloroethylene AUC in blood.
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/11 5-80 DRAFT—DO NOT CITE OR QUOTE
-------
Multistage Cancer Model with 0.95 Confidence Level
BMDL BMP
1
2
3
0 50 100 150 200 250
a. One-degree multistage model fit to male
rat MCL data, all dose groups.
Dichotomous-Hill Model with 0.95 Confidence Level
4
5
6
0 20 40 60 80 100 120
c. Michaelis-Menten model fit to male rat
MCL data, all dose groups.
Dichotomous-Hill Model with 0.95 Confidence Level
11
12
13
14
15
16
17
18
b. One-degree multistage model fit to female
rat MCL data, all dose groups.
Multistage Cancer Model with 0.95 Confidence Level
Multistage Can
extrapolation
d. Multistage model fit to female rat MCL
data, control and lowest dose group only.
9
10
e. Michaelis-Menten model fit to female and
male rat MCL data, all dose groups
Figure 5-12. Dose-response modeling of female and male rat MCLs
associated with inhalation exposure to tetrachloroethylene, in terms of
tetrachloroethylene AUC in blood; response data from JISA (1993). Details
in Appendix D.
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/11 5-81 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
To address NRC (2010) peer review comments, additional dose-response models were
evaluated for this dataset in order to obtain a better model fit, particularly at lower doses where
the dataset exhibited some supralinearity (see Appendix D, Table D-l 1). The analysis was
conducted using administered concentration using the range of dichotomous models included in
BMDS. Among the models fitted, five models fit better than the multistage (Gamma, Weibull,
LogLogistic, LogProbit, and Michaelis-Menten), with two models leading to worse fits than the
multistage (Probit and Logistic). Visually, the Michaelis-Menten model better captured the
supralinear dose-response shape of the data. Because of the better dose-response fit, the
Michaelis-Menten model was preferred over the standard multistage model for this dataset using
administered concentration. The human equivalent POD (BMCLio) was 2.2 ppm
tetrachloroethylene in air (see Table 5-19), with the corresponding central tendency estimate 3.8-
fold higher at 8.6 ppm. Linear extrapolation from this POD led to a human equivalent unit risk of
-3
45 x 10 per ppm, sixfold higher than using the multistage model.
Based on this analysis, the Michaelis-Menten model was also fitted using the preferred
dose metric of tetrachloroethylene AUC in blood (the analysis was not conducted using other
dose-response models because of the near proportionality between this dose metric and
administered tetrachloroethylene). Extrapolation to humans led to an internal dose POD
(BMCLio) of 5 mg-hr/L-day tetrachloroethylene in blood (see Table 5-18). The corresponding
central tendency estimates was about 4-fold higher, at 20 mg-hr/L-day. Linear extrapolation
from the POD to low exposures, followed by conversion to human exposures led to a human
-3
equivalent unit risk of 40 x 10 per ppm.
Therefore, two approaches were carried forward to support cancer risk estimates: the
standard approach using the multistage model and a better fitting approach using the Michaelis-
Menten model, both on the basis AUC of tetrachloroethylene in blood. The remaining analyses
using administered concentration using these and (less preferred) alternative approaches are
retained only to better characterize the range of results from different dose-response models.
5.4.4.1.5. Mononuclear cell leukemia (MCL), female rat
In accordance with standard practice in the absence of MO A data supporting a particular
dose-response model form, multistage modeling of the JISA bioassay data was carried out
considering fits for one-, two-, and three-stage models (details in Appendix D). Using the
preferred dose metric of tetrachloroethylene AUC in blood, a one-stage model had a goodness-
of-fit/>value (p = 0.34) generally considered adequate, and the standardized residuals were
within the recommended limit of two units (see Figure 5-12b). There was no statistical
improvement in fitting higher order models, as all the higher order parameters were estimated to
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-82 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
be zero. Extrapolation to humans led to an internal dose POD (BMCLio) of 61 mg-hr/L-day
tetrachloroethylene in blood. The corresponding central tendency estimate was about twofold
higher, at 136 mg-hr/L-day. Linear extrapolation from the POD to low exposures, followed by
-3
conversion to human exposures led to a human equivalent unit risk of 3.4 x 10 per ppm.
Dose-response modeling using administered exposure fit the data points similarly to that
using tetrachloroethylene AUC in blood (details in Appendix D). The result was directly
interpretable as a human equivalent POD (BMCLio), at 27 ppm tetrachloroethylene in air. The
corresponding central tendency estimate was approximately twofold higher, at 60 ppm. Linear
-3
extrapolation from this POD led to a human equivalent unit risk of 3.7 x 10 per ppm,
essentially the same as using tetrachloroethylene AUC in blood.
To address NRC (2010) peer review comments, additional options were evaluated for this
dataset in order to obtain a better model fit, particularly at lower doses the dataset exhibited some
supralinearity (see Appendix D, Table D-6). These analyses were conducted using administered
concentration, due to its close proportionality with AUC of tetrachloroethylene in blood. This
case was the most extreme among the supralinear datasets, with the multistage model estimate of
the control incidence markedly above the data, and the estimate of the lowest dose group
markedly below the data. Briefly, use of a wider range of dose-response models (as suggested
by NRC. 2010) for the full dataset was considered first. When those attempts proved
unsuccessful, incorporation of historical controls and exclusion of higher exposure groups was
also considered. These approaches are described in more detail below.
First, the range of dichotomous models included in BMDS was considered. Among the
models fitted, four models fit better than the multistage (Gamma, Weibull, LogLogistic, and
Michaelis-Menten), two models fit similarly to the multistage (Probit and Logistic), and one
model fit worse than the multistage (LogProbit). However, for the better fitting models, the
predicted response rate became virtually infinite in slope approaching zero dose. Thus, no
BMCLio could be estimated (see Appendix D), indicating that the statistical uncertainty is too
great to support the BMC estimates. While data are lacking to inform the dose-response
relationship below 50 ppm in female rats, these fits are consistent with the possibility that a
response plateau extends below the lowest observed response. Therefore, none of these options
were successful in both improving upon the multistage model fit and estimating a BMCL.
The next strategy for obtaining an adequate fit to the female rat MCL data involved
focusing model fitting on the low exposure range. First, the sensitivity of the fit to the use of
historical controls was examined in an attempt to constrain the estimated control response at a
level representative of previously observed values. Thus, the concurrent control was replaced
with the overall historical control incidence for inhalation studies in this laboratory (66/448
among control female rats in inhalation studies; see Table 5-16), and all models above were
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-83 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
fitted. None of these fits was both adequate and an improvement on the fits obtained with
concurrent controls (results not shown).
Next, exposure groups were excluded from analysis, starting with the highest exposure
group (600 ppm). All models used above were considered, as was the use of either the
concurrent or historical controls. All model fits were essentially the same as when using the full
data set (see Appendix D). Consequently, the next highest exposure group's data (200 ppm)
were also excluded. Only the multistage model was fit to the two remaining data points (control
and 50 ppm) because the other models use more parameters and need more data points. The
BMCLio was 2.3 ppm, and the BMCio was about twofold higher at 4.9 ppm (see Figure 5-12d;
details in Appendix D). Linear extrapolation from the POD to low exposures, followed by
-3
conversion to human exposures led to a human equivalent unit risk of 43 x 10 per ppm. In
sum, dose-response modeling of the full female rat MCL data set was only superior to the
multistage model for models that could not provide a lower bound estimate for a POD. The only
method that both led to a better fit to the control data and provided a lower bound BMC estimate
for a POD was use of just the concurrent control and lowest female rat exposure group. This
analysis is therefore consistent with the suggestion by the NRC (2010) that use of the multistage
model for the full datasets is not likely to provide a conservative upper bound estimate of risk for
this dataset, and may therefore underestimate risk.
Based on this analysis, the multistage model was also fitted to only the concurrent control
and lowest exposure group using the preferred dose metric of tetrachloroethylene AUC in blood
(the analysis was not conducted using other models because of the near proportionality between
this dose metric and administered tetrachloroethylene). Extrapolation to humans led to an
internal dose POD (BMCLio) of 5.2 mg-hr/L-day tetrachloroethylene in blood. The
corresponding central tendency estimates was about twofold higher, at 11 mg-hr/L-day. Linear
extrapolation from the POD to low exposures, followed by conversion to human exposures led to
-3
a human equivalent unit risk of 39 x 10 per ppm, essentially the same as the result using
administered concentration.
Therefore, two approaches were carried forward to support cancer risk estimates: the
standard approach using the multistage model and the full dataset, and the only available better-
fitting approach using the multistage model and only the control and lowest dose group data,
both on the basis AUC of tetrachloroethylene in blood. However, neither method fully captures
the potential extent of supralinearity into the region below the lowest dose. The remaining
analyses using administered concentration using these and (less preferred) alternative approaches
are retained only to better characterize the range of results from different dose-response models.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-84 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
5.4.4.1.6. Mononuclear cell leukemia (MCL), combined female and male rat
The MCL data for male rats and especially female rats were challenging to fit because of
the apparent supralinearity at lower doses. It was hypothesized that the male and female MCL
responses reflect the same underlying dose-response to tetrachloroethylene. The presence of a
supralinear shape to the dose-response for both male and female rats, in both the NTP (1986b)
and JISA (1993) bioassays (see Figure 5-7), and the similar background MCL rates between
sexes in the JISA rats, are consistent with this hypothesis. Combining the datasets would
increase statistical power and thus perhaps better stabilize the BMDL estimates while being able
to fit the supralinear shape.
Two analyses were conducted to evaluate the consistency of the two JISA datasets. A
test described by Stiteler et al (1993) evaluates whether two datasets are consistent with an
underlying dose-response model. In this case, the Michaelis-Menten model was used, given its
relative success at fitting both datasets in the low-dose region. The test involves comparing the
maximum log-likelihoods for the separate and combined datasets. The resulting />value was
0.54, indicating insufficient reason to conclude that the datasets differ from one underlying
model. The other analysis used a logistic regression to test whether the datasets differed
significantally between males and females. The advantage of this approach is that it does not
require assuming a specific functional form to represent the dose response relationship. This
analysis yielded ap-walue of 0.197, indicating no significant relationship of sex in the pattern of
responses. See Appendix D for more details of both analyses.
The analysis began with fitting all dichotomous models to the combined male and female
MCL data on the basis of administered concentration. As compared to the sex-specific analyses,
only the Michaelis-Menten model provided an overall improved fit to all dose groups relative to
the multistage model (see Figure 5-12d). The resulting BMCio was 7.7 ppm, and the BMCLio
was about sixfold lower at 1.4 ppm (see Table 5-19). Thus, combining the male and female rat
MCL data generated a result with slightly greater statistical uncertainty (shown in the wider
confidence interval) than POD estimates for the sex-specific results. Linear extrapolation from
-3
this POD to low exposures led to a human equivalent unit risk of 71 x 10 per ppm.
Based on this analysis, the Michaelis-Menten model was also fitted using the preferred
dose metric of tetrachloroethylene AUC in blood (the analysis was not conducted using other
models because of the near proportionality between this dose metric and administered
tetrachloroethylene). The result was a human equivalent POD (BMCLio) of 3.0 mg-hr/L-day.
The corresponding central tendency estimate was approximately sixfold higher, at 17 mg-hr/L-
-3
day. Linear extrapolation from this POD led to a human equivalent unit risk of 68 x 10 per
ppm, essentially the same as the estimates using administered tetrachloroethylene.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-85 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Therefore, the approaches carried forward to support cancer risk estimates was the
Michaelis-Menten model on the basis of AUC of tetrachloroethylene in blood. The remaining
analyses using administered concentration are retained only to better characterize the range of
results from different dose-response models.
5.4.4.1.7. Other tumors in male rats
As discussed in Section 5.4.1., tumors occurred at multiple sites in male rats exposed to
tetrachloroethylene in the NTP (1986b) bioassay. While the design of NTP study is less suitable
than the JISA study for developing risk estimates, due to the higher exposures and the fewer dose
groups, dose-response modeling of these data was conducted to address variability in responses
across animal strains and bioassays. Estimates were developed for the risk of each tumor type
individually, as well as for the risk of any combination of tumor types. Because these analyses
are considered less preferred alternatives to those based on the JISA study, additional analyses
with respect to dose-response model selection were not conducted for these data.
5.4.4.1.7.1. Kidney tumors, male rat
As discussed in Section 5.4.3.3 regarding selection of dose-metrics, metabolism of
tetrachloroethylene via the GSH conjugation pathway was calculated as a dose-metric relevant
for effects in the kidney. Multistage modeling of the NTP bioassay was carried out in units of
tetrachloroethylene conjugated with GSH per kg body weight to the 3/4 power per day
considering fits for one- and two-stage models. A one-stage model was found to be sufficient,
with an adequate goodness-of-fitp-value (p = 0.75) and overall adequate visual fit (see
Figure 5-13, details not provided). There was no statistical improvement in fitting higher order
models, as all the higher order parameters were estimated to be zero. Extrapolation to humans
led to an internal POD (BMDLio) of 0.21 mg/kg° 75-day in blood (see Table 5-18). The
corresponding central tendency estimate was about twofold higher, at 0.46 mg/kg°'75-day. Linear
extrapolation from the POD to low internal dose, followed by conversion to human exposures
-3
led to a human equivalent unit risk of 100 x 10 per ppm. However, using the range of posterior
modes for the PBPK model predictions led to human equivalent risks of 0.047 xl0 3toll0x
-3
10 , a range of more than 2000-fold. In view of this large range (much larger than the range for
any of the other
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-86 DRAFT—DO NOT CITE OR QUOTE
-------
Multistat Cancer Model with 0.95Confictenee Level
Multistage Csncer Mcdel with 0.95 Confidence Level
1
2
3
Multistage Cancer
Li near extrapolation —
a. One-stage model fit to kidney tumors.
Multistage Cancer
Linear extrapolation
dose
BMP
b. One-stage model fit to brain gliomas.
Multistage Cancer Model with 0.95Confictence Level
Multistage Cancer Model with 0.95 Confidence Level
4
5
6
7
Multistags Cancer
Linear extrapolaticn
BMDL BMP
c. One-stage model fit to testicular
interstitial cell tumors..
Multistage Cancer
Li near extrapol st\on
BMDL . . BMP.
d. One-stage model fit to MCLs
Figure 5-13. Dose-response modeling of male rat tumors—kidney, brain
gliomas, interstitial cell tumors, MCLs—associated with inhalation exposure
to tetrachloroethylene, in terms of tetrachloroethylene AUC in blood;
response data from NTP (1986b). Details in Appendix D.
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/11 5-87 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
endpoints), and the inability to discern from the toxicokinetic data whether this spread
represented uncertainty or variability or both (see Section 3; Chiu and Ginsberg). AUC of the
parent compound in the blood was preferred as the dose metric for kidney toxicity, while
carrying forward the results of using the GSH conjugation dose metric for comparison.
Thus, multistage modeling of the kidney tumor data was also carried out in units of
tetrachloroethylene AUC in blood, considering fits for one-, two-, and three-stage models. A
one-stage model had an adequate goodness-of-fitp-walue (p = 0.74) and overall adequate visual
fit. There was no statistical improvement in fitting higher order models, as all the higher order
parameters were estimated to be zero (modeling results not shown). Extrapolation to humans led
to an internal POD (BMDLio) of 110 mg-hr/L-day tetrachloroethylene in blood (see Table 5-18).
The corresponding central tendency estimate was about twofold higher, at 246 mg-hr/L-day.
Linear extrapolation from the POD to low internal dose, followed by conversion to human
-3
exposures led to a human equivalent unit risk of 1.8 x 10 per ppm.
Dose-response modeling using administered exposure fit the data points similarly to when
tetrachloroethylene AUC in blood was used (details in Appendix D). The result was directly
interpretable as a human equivalent POD (BMCLio), at 50 ppm tetrachloroethylene in air (see
Table 5-19). The corresponding central tendency estimate was approximately twofold higher, at
110 ppm. Linear extrapolation from this POD led to a human equivalent unit risk of
-3
2.1 x 10 per ppm, essentially the same as the estimate using tetrachloroethylene AUC in blood.
Two multistage model results were carried forward to support cancer risk estimates (Table
5-18): that using AUC of tetrachloroethylene in blood as the dose metric (preferred), and those
using GSH conjugation metabolism as the dose metric (alternative). For the alternative dose
metric, it is also noted that the range of PBPK model-based estimates is carried forward to
characterize the impact of uncertainty in GSH conjugation metabolism in humans.
5.4.4.1.7.2. Brain tumors, male rat
Multistage modeling of the NTP bioassay data for brain gliomas in male rats was carried
out in units of tetrachloroethylene AUC in blood, considering fits for one- and two-stage models.
A one-stage model was found to be sufficient, with adequate goodness-of-fit p-w alue (p = 0,11)
and overall adequate visual fit (see Figure 5-13, details not shown). There was no statistical
improvement in fitting higher order models, as all the higher order parameters were estimated to
be zero.
Extrapolation to humans led to an internal POD (BMDLio) of 190 mg-hr/L-day
tetrachloroethylene in blood (see Table 5-19). The corresponding central tendency estimate was
less than twofold higher, at 400 mg-hr/L-day. Linear extrapolation from the POD to low internal
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-88 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
dose, followed by conversion to human exposures led to a human equivalent unit risk of
1.3 x 10~3 per ppm.
Dose-response modeling using administered exposure fit the data points similarly to when
tetrachloroethylene AUC in blood was used (details in Appendix D). The result was directly
interpretable as a human equivalent POD (BMCLio), at 73 ppm tetrachloroethylene in air (see
Table 5-19). The corresponding central tendency estimate was about twofold higher, at
180 ppm. Linear extrapolation from this POD led to a human equivalent unit risk of
1.4 x 10 3 per ppm, essentially the same as the estimate using tetrachloroethylene AUC in blood.
The multistage modeling result using tetrachloroethylene AUC in blood was carried
forward to support cancer risk estimates (Table 5-18).
5.4.4.1.7.3. Testicular tumors, male rat
Multistage modeling of the NTP bioassay data for testicular tumors was carried out in
units of tetrachloroethylene AUC in blood, considering fits for one- and two-stage models. A
one-stage model had an adequate goodness-of-fitp-walue (p = 0.40) and overall adequate visual
fit (see Figure 13c; details not shown). There was no statistical improvement in fitting higher
order models, as all the higher order parameters were estimated to be zero.
Extrapolation to humans led to an internal POD (BMDLio) of 14 mg-hr/L-day
tetrachloroethylene in blood (see Table 5-18). The corresponding central tendency estimate was
about twofold higher, at 31 mg-hr/L-day. Linear extrapolation from the POD to low internal
dose, followed by conversion to human exposures led to a human equivalent unit risk of
-3
14 x 10 per ppm.
Dose-response modeling using administered concentration fit the data points similarly to
when tetrachloroethylene AUC in blood was used (details in Appendix D). The result was
directly interpretable as a human equivalent POD (BMCLio), at 6.1 ppm tetrachloroethylene in
air (see Table 5-19). The corresponding central tendency estimate was approximately twofold
higher, at 13 ppm. Linear extrapolation from this POD led to a human equivalent unit risk of 16
-3
x 10 per ppm, the same as the higher estimate using tetrachloroethylene AUC in blood.
The multistage modeling result using tetrachloroethylene AUC in blood was carried
forward to support cancer risk estimates (Table 5-18).
5.4.4.1.7.4. Mononuclear cell leukemia, male rat
Multistage modeling of the NTP bioassay data for male rat MCL was carried out in units
of tetrachloroethylene AUC in blood, considering fits for one- and two-stage models. A
one-stage model had an adequate goodness-of-fit p-w alue (p = 0,18) and overall adequate visual
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-89 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
fit (see Figure 5-13d; details not shown). There was no statistical improvement in fitting higher
order models, as all the higher order parameters were estimated to be zero. Extrapolation to
humans led to an internal POD (BMDLio) of 15 mg-hr/L-day tetrachloroethylene in blood, and a
corresponding central tendency estimate about twofold higher, at 28 mg-hr/L-day. Linear
extrapolation from the POD to low internal dose, followed by conversion to human exposures
-3
led to a human equivalent unit risk of 13 x 10 per ppm.
Dose-response modeling using administered exposure fit the data points similarly to when
tetrachloroethylene AUC in blood was used (details in Appendix D). The result was directly
interpretable as a human equivalent POD (BMCLio), at 6.5 ppm tetrachloroethylene in air (see
Table 5-19). The corresponding central tendency estimate was approximately twofold higher, at
-3
12 ppm. Linear extrapolation from this POD led to a human equivalent unit risk of 15 x 10 per
ppm, essentially the same as the estimate using tetrachloroethylene AUC in blood.
The multistage modeling result using tetrachloroethylene AUC in blood was carried
forward to support cancer risk estimates (Table 5-18).
5.4.4.1.7.5. Total risk estimate for NTP (1986b) male rats
The increased incidences of kidney, brain, and testicular interstitial cell tumors seen in
the NTP (1986b) male rats led to unit risks which ranged from about 1 x 10"3 to about 14 x 10"3
per ppm, all lower than the unit risk based on male rats in the JISA (1993) study using the
Michaelis-Menten model. In order to compare the results of both studies more equitably, the
overall impact of these multiple tumor types, or the risk of developing any combination of the
four tumor types, was estimated. First, the tumor types were judged likely to occur
independently of each other, or not only in the presence of one of the other tumor types. The
individual risk estimates developed above were combined for an overall estimate of risk of any
combination of these four tumor types, using the approach based on maximum likelihood
estimation described in Section 5.4.4.1.3.
In terms of tetrachloroethylene AUC, the POD (BMDLio) was 8.2 mg-hr/L-day
tetrachloroethylene in blood (see Table 5-18). The corresponding central tendency estimate was
almost twofold higher, at 14 mg-hr/L-day. Linear extrapolation from the POD to low internal
dose, followed by conversion to human exposures led to a human equivalent unit risk of
-3
25 x 10 per ppm.
Using administered exposure, the estimated overall risk was similar to when
tetrachloroethylene AUC in blood was used (details in Appendix D). The result was directly
interpretable as a human equivalent POD (BMCLio), at 3.5 ppm tetrachloroethylene in air (see
Table 5-19). The corresponding central tendency estimate was approximately twofold higher, at
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-90 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
6.1 ppm. Linear extrapolation from this POD led to a human equivalent unit risk of
-3
29 x 10 per ppm, essentially the same as the higher estimate using tetrachloroethylene AUC in
blood.
The combined overall risk using tetrachloroethylene AUC in blood as the dose metric for
each tumor type was carried forward to support cancer risk estimates (Table 5-18). Overall, the
combined unit risk estimate was less than twofold higher than the highest individual unit risk.
While this bioassay is less ideal for low dose extrapolation than the JISA bioassay, it is still
notable that the combined risk estimate supports the JISA study results, less than threefold lower
-3
than the highest JISA study estimate of -70 x 10 per ppm.
5.4.4.1.8. Summary and discussion of site-specific dose-response modeling
The standard approach of applying the multistage model to the candidate data sets, using
PBPK model-based dose metrics, yielded results that were considered adequate according to
several criteria, including goodness-of-fit^-values > 0.05 and standardized residuals within ±2.
However, the NRC (2010) peer review report recommended a more extensive quantitative
evaluation of uncertainty due to different forms of dose-response models. In particular, NRC
(2010) agreed that for several datasets, the multistage model does not fit the data at lower doses
owing to the supralinear shape in the data. Furthermore, they noted that lack of significance in
goodness-of-fit tests can result from a small number of animals in each dose group, and use of
such tests to justify a selection of a dose-response model can be misleading. Therefore, for the
datasets from JISA (1993), additional analyses were performed to examine whether alternative
dose-response models better accounted for datasets that exhibited supralinearity, and to more
generally characterize the range that would result from applying different dose-response models.
The discussion here focuses on the JISA (1993) data, since these were selected as the primary
source of dose-response data.
For mouse hepatocellular tumors and hemangiomas and hemangiosarcomas, the
alternative analyses did not lead to better fits and did not suggest a wide range of possible results
from alternative dose-response models. Therefore, for those datasets, the results from the
standard multistage approach were carried forward for consideration (in some cases including an
alternative dose metric in addition to the preferred one).
For male and female rat MCLs, some of the analyses yielded model results that
substantially improved fit to the datasets' supralinearity. For male rat MCLs, the preferred result
carried forward used the Michaelis-Menten model, with the standard multistage approach also
carried forward as an alternative for comparison. However, application of the range of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-91 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
alternative dose-response models led to a wide (>300-fold) range of BMCL estimates, indicating
that the data have difficulty supporting a robust statistical lower bound on the BMC.
For female rat MCLs, the only approach that was successful in both addressing the
supralinearity and estimating a BMCL was multistage modeling of only the control and low dose
group. Moreover, for this dataset, the standard multistage approach using the entire dataset had
the most pronounced inaccuracy with respect to the supralinearity in the data. These two results
- the multistage model using the full dataset and using only the control and low dose group -
were carried forward because they were the best available for this dataset. The fit to the full
dataset likely substantially overestimates the BMD, due to the markedly high estimate for the
control incidence and the markedly low estimate for the low dose incidence. However, while the
fit to only the control and low dose groups leads to a good fit to those data, it cannot
quantitatively address the possibility that the supralinearity extends below the lowest dose group.
Finally, application of the range of alternative dose-response models led to an unbounded range
of BMCL estimates, with some models unable to estimate a statistical lower bound on the BMC.
Because of these difficulties in fitting the individual rat MCL datasets, a subsequent
analysis was performed using the combined male and female datasets. There are no biological
data suggesting that the male and female rats would not reflect the same underlying toxicological
dose-response, and statistical tests indicated that these data could be combined. In fitting the
range of available dichotomous models, it was found that the Michaelis-Menten model led to the
best fit, and was able to account for the supralinearity in the full dataset. Moreover, the range of
alternative dose-response models all lead to stable estimates for the lower bound on the BMD.
Therefore, the analysis using the Michaels-Menten model on the full, combined male and female
rat MCL dataset was carried forward for consideration.
Figure 5-14 shows the relative magnitudes of the unit risks associated with each tumor
site. Also shown are the unit risks estimated using alternate dose metrics, including administered
concentration, the range of estimates based on alternative PBPK model parameters, and the range
of estimates based on the range of dose-response models available in BMDS. Finally, this figure
also includes estimates based on dose-response modeling of the NTP (1986b) bioassay. In terms
of preferred dose metrics (see Section 5.4.3.2.3.1), the unit risks, rounded to one significant
-3
figure, ranged from 0.9 to 70 x 10 per ppm, about an 80-fold range.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-92 DRAFT—DO NOT CITE OR QUOTE
-------
Male mice, hepatocellular tumors
Male mice, hemangiomas or hemangiosarcomas
Female mice, hepatocellular tumors
Male rats, MCL
Female rats, MCL
Female and male rats, MCL
Male rats (NTP), kidney tumors
Male rats (NTP): brain, kidney, testes, MCL
Human equivalent PCE unit risk estimates (per ppm)
per mg/m
10 6 10 5 10 4 10 3 10 2 10 1
-x
AdmVto
multra
Adminis!
° TCA AUC in livel
O Total GSH metab.
A Total liver oxidative me^ab."
a (filled) primary dose metiV
multistage, all dose group
(filled) primaiV dose metric,
other analyses
x
-5
-4
3
-2
-1
Figure 5-14. Comparison of inhalation unit risks for tetrachloroethylene
derived from rodent bioassays using PBPK-based dose metrics and
administered concentration. Symbols represent results using the posterior
mode PBPK model results, with filled symbols representing the preferred
dose metrics (Tables 5-18 and 5-19). Red-filled symbols use the multistage
model with all dose groups; green-filled symbols use a different dose-
response approach in response to NRC (2010) comments. Solid error bars
show the range of estimates using the range of posterior modes for the
human PBPK model-based conversion to a human equivalent unit risk
(Table 5-18). Dashed error bars show the range of unit risk estimates (based
on administered concentration) using alternative dose-response models with
goodness-of-fit p-values > 0.10 (Table 5-20).
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
6/24/11 5-93 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
5.4.4.2. Choice of Data Set and Associated Uncertainties
The choice of data set for best representing an upper bound estimate of human
carcinogenic potency involves a number of factors, including the magnitude and robustness of
the response, the role of metabolism, the carcinogenic MO As, the dose-response model fit, and
the resulting low-dose extrapolation predictions.
The highest magnitude and most robust responses for tetrachloroethylene carcinogenicity in
rodents are the increased incidences of liver tumors (hepatocellular adenomas and carcinomas)in
both sexes of mice and of MCL in both sexes of rats, with biologically and statistically
significant increases over background (see Section 5.4.2). In mice, hemangiosarcomas in
liver,spleen, fat, and subcutaneous skin were reported in males in the JISA study, but the
incidences (in terms of additional risk) were lower than those for hepatocellular adenomas and
carcinomas, and not reported in other studies. Additional tumor findings in rats included
significant increases in the NTP bioassay of testicular interstitial cell tumors and kidney tumors
in males, and brain gliomas in males and females. The incidences (in terms of additional risk)
were lower than those for MCL, and not reported in other studies. Therefore, on the basis of this
factor, mouse liver tumors and rat MCLs carry the greatest weight since these endpoints are
biologically and statistically significant, and reproducible across sexes and bioassays.
In terms of the role of metabolism, the specific toxic moieties have not been identified for
any endpoint. However, for mouse liver tumors and rat kidney tumors there are data that identify
the likely metabolic pathway involved—oxidation and GSH conjugation, respectively. For
oxidation, toxicokinetic data and modeling indicate that this pathway represents a greater
fraction of tetrachloroethylene disposition in mice than in humans, a difference that can be
accounted for quantitatively through use of the PBPK model. Therefore, this factor leads to
decreasing the weight accorded to mouse liver tumors, but the extent of the difference can be
carried through quantitatively and addressed in the comparison of resulting low-dose
extrapolation predictions. For rat kidney tumors, the range of estimates for GSH conjugation is
very wide, with some estimates based on this dose metric being higher than those based on the
AUC of tetrachloroethylene in blood, which was selected as the preferred surrogate dose metric.
Therefore, it is unclear whether the weight accorded to rat kidney tumors should be increased or
decreased, as the toxicokinetic data are inadequate to quantify the extent of interspecies
differences. For the endpoints other than mouse liver and rat kidney tumors, toxicokinetic data
are not informative as to the choice of data set that may best reflect human carcinogenic potency
In terms of MO A, only for rat kidney tumors and mouse liver tumors are there any
concrete hypotheses. For rat kidney tumors, the hypothesized modes of action include
mutagenicity, peroxisome proliferation, o^-globulin nephropathy, and cytotoxicity not
associated with o^-globulin accumulation. For mouse liver tumors, the MO A hypotheses
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-94 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
concern mutagenicity, epigenetic effects (especially DNA hypomethylation), oxidative stress,
and receptor activation (focusing on a hypothesized PPARa activation MO A). However, the
available evidence is insufficient to support the conclusion that either rat kidney or mouse liver
tumors are mediated solely by one of these hypothesized modes of action. In addition, no data
are available concerning the mechanisms that may contribute to the induction of other rodent
tumors (including MCL, brain gliomas, or testicular interstitial cell tumors in exposed rats and
hemangiosarcomas in exposed mice). Furthermore, no mechanistic hypotheses have been
advanced for the human cancers suggested to be increased with tetrachloroethylene exposure in
epidemiologic studies, including bladder cancer, non-Hodgkin lymphoma and multiple myeloma.
Although target organ concordance is not a prerequisite for evaluating the implications of animal
study results for humans (U.S. EPA. 2005a). it is notable that the leukemias (in both sexes of
rats) support the observation of lymphopoietic cancers in individuals employed as dry cleaners
and degreasers, and the liver tumors (in both sexes of mice) support the observation of liver
tumors in dry cleaners (see Section 4.10.1.1.2). Overall, the MO As involved in the
carcinogenicity of tetrachloroethylene and its metabolites are not known, and mechanistic data
are not informative as to the choice of data set that may best reflect human carcinogenic potency.
The next factor involves the dose-response modeling results. There are a number of
uncertainties associated with the dose-response modeling. First, there is some uncertainty with
respect to the dose-response model fits. This is particularly true with respect to nonmonotonic
and/or supralinear data sets. As discussed extensively in Section 5.4.4.1, for such datasets, a
number of alternative analyses were performed in an attempt to obtain better model fits. The
datasets that did not exhibit supralinearity were all fit well by the multistage model, and carry the
greatest weight from this perspective. These include the female mouse hepatocellular tumors,
male mouse hemangiosarcomas, and all the NTP (1986b) datasets. For the male mouse
hepatocellular tumors, none of the alternative analyses were successful in obtaining better model
fits to the supralinear dose response shape, so these data carry somewhat less weight from this
perspective. The most challenging datasets were the rat MCL data from JISA (1993), which
necessitated trying multiple approaches. Among those results, the results of the male MCL data
and the combined male and female MCL data carry the greatest weight, since the Michaelis-
Menten model both fit the supralinear shape and resulted in a stable BMCL estimate. Less
weight is accorded to results of the female MCL data, which necessitated use of only the control
and lowest dose group. Another indicator related to the dose-response fit is the statistical
uncertainty at the POD. For the selected dose-response models this uncertainty is quite modest
at around twofold or less for all data sets except the combined male and female MCL fits, which
had statistical uncertainty at the POD of around fivefold. In addition, for the male MCL fits, the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-95 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
use of some alternative dose-response models led to poorly bounded BMCs, suggesting that this
dataset may carry somewhat less weight due to its more limited ability to bound the BMC.
The final factor involves the resulting low-dose extrapolation predictions. As shown in
Figure 5-14, the dose-response analysis of the combined male and female MCL data resulted in
the highest unit risk estimated using a preferred dose metric, and carry the greatest weight from
this perspective. The sex-specific rat MCL JISA (1993) results carry the next greatest weight, as
they are about twofold less than the estimate based on the combined male and female MCL data
from JISA (1993). For studies that reported multiple tumors [brain, kidney, testes, and MCL in
NTP (1986b), rats; and hepatocellular tumors and hemangiomas or hemangiosarcomas in JISA
(1993) mice], there is concern that total cancer risk is underestimated by considering only one
site. Estimates of total tumor risk from male rats in the NTP (1986b) bioassay were less than
threefold lower than the most sensitive result based on male and female MCL data (JISA, 1993),
and thus carry the next greatest weight from this perspective.
Given these significant gaps in the scientific knowledge regarding the metabolites and
mechanisms contributing to tetrachloroethylene-induced cancer, the factors given the strongest
consideration in selection among the available data set were the magnitude and robustness of the
response, the dose-response model fit, and the resulting low-dose extrapolation predictions.
Based on these factors, the dose-response analyses using the Michaelis-Menten model of the
combined male and female rat MCL from the JISA study were selected. These data showed a
strong and robust observed response; the dose response modeling was able to fit the dataset's
supralinearity as well as estimate a reasonable BMDL; and it is the most sensitive unit risk
estimate using a preferred dose metric. Therefore, this analysis is accorded the greatest overall
weight among the available choices. Supporting this selection are two analyses given slightly
less weight: the Michaelis-Menten model-based analysis of the male MCL from the JISA
bioassay, and the analysis of the total tumor risk among four sites from male rats in the NTP
bioassay. Each of these results is also based on strong and robust observed responses and fits
that accounted for any supralinearity, and lead to only slightly less sensitive unit risk estimates.
However, the male MCL data from JISA (1993) led to a much wider range of BMDL estimates
when a range of alternative dose-response models were applied; and the NTP (1986b) data are
based on fewer dose groups and on several endpoints that were not reproduced in other
bioassays. Finally, the results from the analysis of only the control and low dose group from the
female MCL JISA (1993) data were of similar sensitivity, but where based on dose-response
modeling that could not account for any supralinearity below the lowest dose, and thus were
accorded less overall weight.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-96 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
5.4.4.3. Recommended Inhalation Unit Risk
Human inhalation cancer risk has been assessed using several different gender-species
animal tumor data sets and a newly developed —harinnized" PBPK model. These results, and
their uncertainties, have been discussed above and are summarized in Figure 5-14. Based on
consideration of the factors discussed above, the combined male and female rat MCL data
provide strongest basis for deriving a unit risk, defined as a plausible upper-bound excess
lifetime cancer risk estimated to result from continuous exposure to tetrachloroethylene unit risk.
From Table 5-18, the recommended inhalation unit risk value is 7 x 10~2 per ppm, or
1 x 10~5 per jug/m3, rounding to one significant digit. Estimates of male-only rat MCL from
JISA (1993) and total tumor risk from brain, kidney, testes, and MCL in NTP (1986b) rats were
also strongly supported, and were less than threefold lower than the preferred estimate. Lower
estimates that were less strongly supported include results for other tumor sites using preferred
dose metrics and some alternative dose metrics. A higher estimate results from using the male
rat kidney tumors using the high estimate of human GSH conjugation. However, none of these
lower or higher estimates were considered as strongly supported for estimating a plausible upper
bound excess lifetime cancer risk as the one selected. The recommended unit risk should not be
used with exposures exceeding 1 ppm, or 10 mg/m3 (the equivalent ambient exposures
corresponding to the POD for male and female rat MCLs), because above this exposure level the
dose-response relationship is not linear and the unit risk would tend to overestimate risk.
5.4.4.4. Recommended Oral Slope Factor
The oral slope factor was developed from inhalation data because the only available oral
bioassay had several limitations for extrapolating to lifetime risk in humans (see also
Section 5.4.1). First, the study was conducted by gavage at relatively high doses. Human
exposures are less likely to occur in boluses, and high doses are associated at least with saturable
metabolism processes which may involve a different profile of toxicological processes than those
prevalent at more likely environmental exposure levels. Also, the animals were dosed for only
approximately 75% of the more usual 2-year period, making the oral study less useful for
estimating lifetime risk. Route-to-route extrapolation from the inhalation PODs developed from
the JISA study (see Table 5-18) was carried out using the human pharmacokinetic models
described in Section 3.5. The total tumor risks from multiple sites (brain, kidney, testes, and
MCL in rats and hepatocellular tumors and hemangiomas or hemangiosarcomas in mice) were
estimated using the same methods as was done for the inhalation unit risk estimates, with results
-3 -3
of 20 x 10 per mg/kg-day for rats in NTP (1986b) and 18x10 per mg/kg-day for mice in
JISA (1993). Table 5-20 and Figure 5-15 summarize all of the resulting slope factors.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-97 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
The same rationale supporting selection of the estimates based on combined male and
female rat leukemias for the inhalation unit risk applies to the oral slope factor. The
recommended slope factor is 6 x 10~2 per mg/kg-day, rounding to one significant digit.
Estimates of male-only rat MCL from JISA (1993) and total tumor risk from brain, kidney,
testes, and MCL in NTP (1986b) rats were also strongly supported, and were less than threefold
lower than the preferred estimate. Lower estimates that were less strongly supported include
results for other tumor sites using preferred dose metrics and some alternative dose metrics. A
higher estimates results from using the male rat kidney tumors using the high estimate of human
GSH conjugation. However, none of these lower or higher estimates were considered as strongly
supported for estimating a plausible upper bound excess lifetime cancer risk as the one selected.
The recommended slope factor should not be used with exposures exceeding 2 mg/kg-day (the
equivalent ambient exposure corresponding to the POD for male and female rat MCLs), because
above this exposure level the dose-response relationship is not linear and the slope factor would
tend to overestimate risk.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-98 DRAFT—DO NOT CITE OR QUOTE
-------
° PCE AUC in blood ^
5
o
m
D TCA AUC in liver
•
O Total GSH metab.
t—
Q)
Cl
O
fi
u
A Total liver oxidative metab.
m (filled) primary dose metric, i-a
multistage, all dose groups A
(filled) primary dose metric,
-------
Table 5-21. Human equivalent oral slope factors, derived using primary dose
metrics and multistage model; tumor incidence data from JISA (1993) and
NTP (1986b)
Study Group
Tumor type
(multistage model with
all dose groups unless
otherwise specified)
Human Equivalents
PODa, in internal dose units
SFxlO"3
/internal
dose
unitb
OSFxlO"3/
mg/kg-day
(PBPK
range)0
Primary dose metrics
Male mice
JISA (1993)
Hepatocellular adenomas
or carcinomas
BMDio
BMDLio
2.9
2.1
Total liver oxidative
metabolism, mg/kg0 75-d
49
2.1
(1.6-2.6)
Hemangiomas,
hemangiosarcomas,
BMDio
BMDLio
63
34
PCE AUC in blood,
mg-hr/L-d
2.9
5.1
(4.6-5.3)
Female mice
JISA (1993)
Hepatocellular adenomas
or carcinomas
BMDjo
BMDLio
8.4
4.0
Total liver oxidative
metabolism, mg/kg0 75-d
25
1.1
(0.84-1.3)
Male rats
JISA (1993)
MCL
BMDio
BMDLio
46
30
PCE AUC in blood,
mg-hr/L-d
3.4
5.9
(5.3-6.1)
MCL (Michaelis-
Menten)
BMDio
BMDLio
20
5.0
PCE AUC in blood,
mg-hr/L-d
20
35
(31-36)
Female rats
JISA (1993)
MCL
BMDio
BMDLio
136
61
PCE AUC in blood,
mg-hr/L-d
1.6
2.8
(2.5-2.9)
MCL (control and low
dose groups only)
BMDio
BMDLio
11
5.2
PCE AUC in blood,
mg-hr/L-d
19
33
(30-35)
Female and
male rats JISA
(1993)
MCL (Michaelis-
Menten)
BMDio
BMDLio
17
3.0
PCE AUC in blood,
mg-hr/L-d
33
58
(53-61)
Male rats
NTP (1986b)
Kidney tumors
BMDio
BMDLio
246
110
PCE AUC in blood,
mg-hr/L-d
0.90
1.6
(1.4-1.6)
Brain gliomas
BMDio
BMDLio
400
192
PCE AUC in blood,
mg-hr/L-d
0.62
1.1
(1.0-1.1)
Testicular interstitial cell
tumors
BMDio
BMDLio
31
14
PCE AUC in blood,
mg-hr/L-d
7.1
12
(11-13)
MCL
BMDio
BMDLio
28
15
PCE AUC in blood,
mg-hr/L-d
6.6
12
(10-12)
Total risk for any of
above four tumor types
BMDio
BMDLio
14
8.2
PCE AUC in blood,
mg-hr/L-d
12
21
(19-22)
Alternate Dose Metrics
Male mice
JISA (1993)
Hepatocellular aden-
omas or carcinomas
BMDio
BMDLio
97
69
TCA AUC in liver,
mg-hr/L-d
1.5
1.7
(1.3-1.8)
Female mice
JISA (1993)
Hepatocellular aden-
omas or carcinomas
BMDio
BMDLio
292
141
TCA AUC in liver,
mg-hr/L-d
0.72
0.85
(0.65-0.89)
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-100 DRAFT—DO NOT CITE OR QUOTE
-------
Male rats
Kidney tumors
BMD05
0.46
Total GSH metabolism,
243
120
NTP (1993)
bmdl05
0.21
mg/kg0.75-d
(0.045-140)
Table 5-20. Human equivalent slope factors, derived using primary dose
metrics and multistage model; tumor incidence data from JISA (1993) and
NTP (1993) (continued)
SF = slope factor; OSF = oral slope factor
PODs were estimated at the indicated BMRs in terms of extra risk; i.e., BMDL10 is the lower bound for the internal
dose metric on the level associated with 10% extra risk. Dose units are in the first column, which include cross-
species scaling to a human equivalent internal dose metric. See Appendix D for dose-response modeling details.
bSlope Factor = BMR/BMDLBmr in units of risk per dose metric unit.
The oral slope factor is given by the product of the slope factor in units of risk per dose metric unit and an oral dose
metric conversion factor (DMCFmg |:g_,kiy): Inhalation Unit Risk = BMR/BMDLBmr x DMCFmg |:g_d(iy. where the
DMCFmg/kg_day is derived from the PBPK model. The DMCFmg/kg_day for each dose metric is a constant factor
shown:
Dose metric
DMCFmg/kg.jay
Overall posterior mode
Range of posterior modes
Total liver oxidative metabolism
0.0438
0.0334-0.0459
Tetrachloroethylene blood AUC
1.74
1.58-1.82
TCA AUC in liver
1.18
0.903-1.24
Total GSH metabolism
0.512
0.00019-0.543
Values in bold correspond to using the overall posterior mode, and are carried forward for consideration as the
recommended cancer slope factor. The difference between the overall and alternative posterior modes is negligible
(relative to other uncertainties) except for the Total GSH metabolism dose metric.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-101 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
5.4.4.5. Uncertainties in Human Population Variability and Quantitative Adjustment for
Sensitive Populations (Age-Dependent Ajustment Factors)
The human variability in response to tetrachloroethylene is also poorly understood. The effect of
metabolic variation, including potential implications for differential toxicity, has not been well
studied. The extent of interindividual variability in tetrachloroethylene metabolism has not been
characterized. As noted above, several enzymes of the oxidative and GSH metabolism, notably
Cytochrome 2E1 (CYP2E1), CYP3A4, GSTZ, GST A, GSTM, and GSTT, show genetic
polymorphisms with the potential for variation in production of specific metabolites. Inducers of
CYP450 enzymes such as toluene, phenobarbital, and pregnenolone-16a-carbonitrile have been
shown to increase tetrachloroethylene metabolism, whereas CYP enzyme inhibitors such as SKF
525A, metyrapone, and carbon monoxide have been shown to decrease tetrachloroethylene
metabolism. Additionally, chronic exposure to tetrachloroethylene has been shown to cause
self-induction of metabolism. Human population variability has also been discussed in
Section 3.
Although a mutagenic MOA would indicate increased early-life susceptibility, there are
no data exploring whether there is differential sensitivity to tetrachloroethylene carcinogenicity
across life-stages. This lack of understanding about potential differences in metabolism and
susceptibility across exposed human populations thus represents a source of uncertainty.
Nevertheless, the existing data do support the possibility of a heterogeneous response that may
function additively to ongoing or background exposures, diseases, and biological processes. As
noted in Section 4.9.5, there is some evidence that certain subpopulations may be more
susceptible to exposure to tetrachloroethylene. These subpopulations include early and later
life-stages and groups defined by health and nutrition status, gender, race/ethnicity, genetics, and
multiple exposures and cumulative risk. These considerations strengthen the scientific support
for the choice of a linear nonthreshold extrapolation approach. However, because chemical-
specific life-stage susceptibility data are not available and the MOA for tetrachloroethylene has
not been established, the application of age-derived adjustment factors for early life exposures,
as discussed in Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to
Carcinogens (U.S. EPA. 2005b) is not recommended.
5.4.4.6. Concordance of Animal and Human Risk Estimates
Sufficient human health outcome data with quality exposure characterizations linked to
individual study subjects or epidemiologic studies with characterization of exposure-response
using a quantitative surrogate of tetrachloroethylene exposure are not available to derive cancer
risk values. Two recent analyses of epidemiologic studies provide some limited perspectives on
the human cancer risk values estimated from animal bioassays (Finkel 2010; van Wijngaarden
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-102 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
and Hertz-Picciotto. 2004). Each analysis assigns an exposure surrogate of average
tetrachloroethylene concentration to all exposed subjects, either based on information in the
published literature (van Wijngaarden and Hertz-Picciotto. 2004) or from monitoring data of
similar workplaces as those of study subjects (Finkel. 2010). EPA prefers that the
exposure-assessment approach of epidemiologic studies used for estimating lifetime cancer risk
represent not only the relevant conditions and exposures (e.g., through a job exposure matrix or
exposure model), but also subject-specific quantitative estimates of exposure. The
epidemiologic studies (Lynge et al.. 2006; Vaughan et al.. 1997) in the two analyses did not meet
these criteria; neither study assigned a unique exposure estimate to individual subjects nor did
they examine exposure-response using a quantitative exposure surrogate. Although not
sufficient to serve as a primary basis for dose-response assessment, these studies do provide
information without extrapolation from animals to human.
Finkel (2010) developed a crude estimate of the cumulative exposure of dry-cleaning
workers studied in Lynge et al. (2006) of 0.2 ppm based on data from other Nordic studies.
Using the strongest result from that study, a relative risk of bladder cancer of 1.44 (95% CI:
1.07-1.93), along with the estimated U.S. lifetime risk of bladder cancer of 2.39% from SEER
(Altekruse et al.. 2010) implies an inhalation unit risk estimate of 0.05 (95% CI: 0.008-0.1) per
ppm, or 8 x 10~6 (95% CI: 1 x 10 6— 16 x 10~6) per |ig/m3, the confidence interval of which
overlaps with the cancer risk estimates from combined male and female rat MCL tumors in JISA
(1993).
Van Wijngaarden and Hertz-Picciotto (2004) demonstrated a simple methodology using
epidemiologic data for four chemical exposures including tetrachloroethylene. For
tetrachloroethylene specifically, a linear dose-response model was fit to laryngeal cancer
observations in the upper airway cancer case-control study of Vaughan et al. (1997). Van
Wijngaarden and Hertz-Picciotto (2004) presented both an ED0i and LED0i (effective dose for a
1% additional lifetime risk over background and the lower confidence interval on this dose,
called the TD1 and LCL1 in their paper) for humans exposed for 45 years, 240 days/year, a
standard occupational exposure scenario. The ED0i was 228.40 mg/day and LED0i was
60.16 mg/day. In order to compare these results with those derived from the JISA (1973) study,
we assumed a continuous lifetime exposure (70 years, 365 days/year, and 20 m3/day breathing
rate), resulting in an equivalent ED0i of 4.8 mg/m3 and LED0i of 1.3 mg/m3. Using the
continuous lifetime equivalent LED0i as the POD and a low-dose linear approach, a unit risk
based upon Vaughan et al. (1997) is 0.01/1.3 x 103 |ig/m3 = 8 x 10~6 per |ig/m3 (0.05 per ppm).
A cancer risk estimate from human data using the ED0i as the POD is 0.01/4.8 x 103 |ig/m3 = 2
x 10~6 per |ig/m3 (0.01 per ppm). The higher of these two estimates is within 20% of the cancer
risk estimates from combined male and female rat MCL tumors in JISA (1993).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-103 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
These estimates are based on extrapolated exposure estimates, assume that bladder cancer
and laryngeal cancer, respectively, are the only carcinogenic hazard in humans, and may be
subject to other sources of bias. Thus, they should only be viewed as order of magnitude
estimates. Interestingly, however, they appear to be consistent both which each other and with
the cancer risk estimates from combined male and female rat MCL tumors in the JISA bioassay
(1993). Therefore, while estimates based on human data are not sufficient to serve as a primary
basis for dose-response assessment, they do suggest that the cancer risk estimates based on
rodent bioassays are plausible.
5.4.5. Summary of Uncertainties in Cancer Risk Values
A number of uncertainties underlie the cancer unit risk for tetrachloroethylene, as
discussed in the above sections. Table 5-21 summarizes the impact on the assessment of issues
such as the use of models and extrapolation approaches (particularly those underlying the
Guidelines for Carcinogen Risk Assessment, U.S. EPA. 2005a\ the effect of reasonable
alternatives, the decision concerning the preferred approach, and its justification.
The uncertainties presented in Table 5-21 have a varied impact on risk estimates. Some
suggest risks could be higher than was estimated, while others would decrease risk estimates or
have an impact of an uncertain direction. Several uncertainties are quantitatively characterized
for the significantly increased rodent tumors. These include the range of uncertainty in PBPK
modeling and dose metrics and the statistical uncertainty in the multistage modeling estimate.
Due to limitations in the data, particularly regarding the MOA and relative human sensitivity and
variability, the quantitative impact of other uncertainties of potentially equal or greater impact
has not been explored. As a result, an integrated quantitative analysis that considers all of these
factors was not undertaken.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-104 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-22. Summary of uncertainties in tetrachloroethylene cancer unit
risk estimate
Consideration/
approach
(section)
Impact on unit
risk
Decision
Justification
Bioassay (5.4.1)
i unit risk
threefold if NTP
study used
JISA study
JISA study used the lowest experimental
exposures (reduces extrapolation uncertainty)
and used three treated groups
PBPK modeling
and dose metrics
(5.4.3.1.5)
Alternatives could
t or I unit risk by
an unknown extent
Relied on total liver
oxidative metabolism and
tetrachloroethylene AUC,
in addition to
administered
concentration
Experimental evidence supports a role for
metabolism in toxicity, but actual responsible
metabolites are not clearly identified.
Cross-species
scaling
(5.4.3.2.2.3)
Alternatives could
i or | unit risk
(e.g., 3.5-fold i
[scaling by BW]
or | twofold
[scaling by
BW2'3])
(default approach) for
total oxidative or GSH
metabolism; direct animal
to human correspondence
when the dose-metric was
an AUC
There are no data to support alternatives. Use
of BW3'4 for metabolism rates and no scaling
for dose metrics expressed as AUCs are
consistent treatments of the available dose
metrics. While the true human
correspondence is unknown, this overall
approach is expected neither to over- or
underestimate human equivalent risks
Low-dose
extrapolation
procedure
(5.4.3.2.3)
Departure from
EPA's Guidelines
for Carcinogen
Risk Assessment
POD paradigm, if
justified, could j
or | unit risk an
unknown extent
Multistage model to
determine POD, linear
low-dose extrapolation
from POD (default
approach)
Available MOA data do not inform selection
of dose-response model but do not support
nonlinearity (mutagenicity is plausible
contributor and cannot be ruled out); linear
approach in absence of clear support for an
alternative is generally supported by scientific
deliberations supporting EPA's Guidelines for
Carcinogen Risk Assessment.
Model
uncertainty
Alternatives could
i or | unit risk
Multistage model for all
tumor sites except
Michaelis-Menten model
for MCLs from JISA
(1993) male rats, and
male and female rats
combined
No biologically based models available; no a
priori basis for selecting a model other than
multistage. Selected options tended to be
intermediate among the available alternatives.
See Appendix D.
Statistical
uncertainty at
POD (5.4.4.1.7)
i unit risk fivefold
if BMCio used
rather than
BMCLio
BMCL (default approach
for calculating plausible
upper bound)
Limited size of bioassay results in sampling
variability; lower bound is 95% confidence
interval on concentration.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-105 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-21. Summary of uncertainties in tetrachloroethylene cancer unit
risk estimate (continued)
Consideration/
approach
(section)
Impact on unit
risk
Decision
Justification
Species/gender
combination
(5.4.4.2,
Figure 5-14)
Human risk could
i or depending
on relative
sensitivity
Male and female rat MCL
MCL is the largest response, occurs in both
sexes and is reproducible across studies,
despite moderate background response rate.
There are no MOA data to guide extrapolation
approach for any choice. It was assumed that
humans are as sensitive as the most sensitive
rodent gender/species tested; true
correspondence is unknown. A carcinogenic
response occurs across test species, though
with differing tumor types. This supports the
general assumption that direct site
concordance is not necessary. Consistent with
this view, some human tumor types associated
with tetrachloroethylene are not found in
rodents (i.e., cervical, esophageal cancer
deaths).
Human
population
variability
sensitive
subpopulations
(5.4.4.5)
Low-dose risk f to
an unknown extent
Considered qualitatively
No data to support range of human
variability/sensitivity in metabolism or
response, including whether children are more
sensitive. Mutagenic MOA, which cannot be
ruled out, would indicate increased early-life
susceptibility.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 5-106 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF
HAZARD AND DOSE-RESPONSE
6 1 HUMAN HAZARD POTENTIAL
This section summarizes the human hazard potential for tetrachloroethylene. For
extensive discussions and references, see Section 2 for Exposure, Section 3 for toxicokinetics
and physiologically based pharmacokinetic (PBPK) modeling, and Sections 4.1-4.8 for the
epidemiologic and experimental studies of noncancer toxicity and carcinogenicity. Section 4.9
summarizes information on susceptibility, and Section 4.10 provides a more detailed summary of
noncancer toxicity and carcinogenicity.
6.1.1. Exposure (see Section 2)
Tetrachloroethylene is a volatile compound with relatively low water solubility. It is
widely used for dry cleaning of fabrics, for metal degreasing, and in manufacturing some
consumer products and other chemicals. Tetrachloroethylene has been detected in drinking,
ground, and surface water as well as in air, soil, food, and breast milk. The primary exposure
routes of concern are vapor inhalation and ingestion of contaminated water. Inhalation exposure
is the predominant route of exposure compared with ingestion, including from breast milk.
The highest environmental releases are to the air. Ambient tetrachloroethylene
concentrations vary from source to source and with proximity to the source. Outdoors, the high
volatility leads to increased ambient air concentrations near points of use (ATSDR. 1997a; U.S.
EPA. 1996b). The U.S. Environmental Protection Agency (EPA) has carried out modeling to
characterize the geographic distribution of tetrachloroethylene for its National-Scale Air Toxics
Assessment database (U.S. EPA. 1996b). Median census tract-based tetrachloroethylene
concentrations across the United States were estimated at about 0.3 |ig/m3 for urban areas and
0.1 |ig/m3 for rural areas (75% upper percentiles of 0.4 and 0.2 |ig/m3, respectively). Air
exposure may also occur from vapor intrusion, or during showering or bathing as dissolved
tetrachloroethylene in the warm tap water is volatilized.
Near points of use, such as dry cleaners or industrial facilities, indoor exposure to
tetrachloroethylene is more significant than outdoor exposure (U.S. EPA. 2001b). Adgate et al.
(2004a) measured tetrachloroethylene in outside and indoor air at school, indoor air at home, and
using personal samples on children, and demonstrated that levels are lower in homes with greater
ventilation (Adgate et al.. 2004a) and in homes in nonurban settings (Adgate et al.. 2004a;
Adgate et al.. 2004b). Indoor air concentrations in apartments above dry-cleaning shops have
been measured at up to 4.9 mg/m3 (Verberkand Scheffers. 1980) (see also Altmann etal.. 1995;
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-1 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Garetano and Gochfeld. 2000; McDermott et al.. 2005; Schreiber et al.. 1993; Schreiber et al..
2002). Measurements have also been made in a daycare center adjacent to a dry cleaner
(NYSDOH. 2005a. b, c), and in a classroom exposed to tetrachloroethylene from an air
—errssion from a small chemical factory" (Monster and Smolders. 1984a). Mean concentrations
inside dry-cleaning facilities were reported to be 454—1,390 mg/m3 in the United States and
164 mg/m3 in Nordic countries during the 1960s and 1970s. Overall levels declined from
95-210 mg/m3 in the 1980s to 20-70 mg/m3 over the next decades in these countries (Gold et
al.. 2008; Lynge et al.. 2006; Lynge et al.. 2011).
The off-gassing of garments that have recently been dry-cleaned may be of concern (see
also Thomas et al.. 1991; Tichenor et al.. 1990). Relatively high tetrachloroethylene air
concentrations have been measured in closets and automobiles containing freshly dry-cleaned
clothing. Using dry-cleaned clothes as a source, tetrachloroethylene levels inside a stationary
vehicle after 30 minutes reached 0.230 mg/m3 (Park et al.. 1998). A residential closet storing
newly dry-cleaned clothing had an air concentration of 2.9 mg/m3 after 1 day, which rapidly
declined to 0.5 mg/m3 and persisted for several days (Tichenor et al.. 1990). There is
one documented mortality case: a 2-year-old boy was found dead after being put to sleep in a
room with curtains that had been incorrectly dry cleaned (Gamier et al.. 1996).
Exposure to related compounds—including metabolites and other parent compounds that
produce similar metabolites—can alter or enhance tetrachloroethylene metabolism and toxicity
by generating higher internal metabolite concentrations than would result from
tetrachloroethylene exposure by itself.
6.1.2. Toxicokinetics and Physiologically Based Pharmacokinetic (PBPK) Modeling (see
Section 3)
Tetrachloroethylene is a lipophilic compound that readily crosses biological membranes.
Tetrachloroethylene is rapidly absorbed into the bloodstream following oral and inhalation
exposures. It can also be absorbed across the skin following dermal exposure to either pure or
diluted solvent or vapors (Nakai et al.. 1999; Poet et al.. 2002; Stewart and Dodd. 1964).
Additionally, tetrachloroethylene can be transferred transplacental^ and through breast milk
ingestion. See Section 3.1 for additional discussion of tetrachloroethylene absorption.
Once absorbed, tetrachloroethylene is distributed by first-order diffusion processes.
Animal studies provide clear evidence that tetrachloroethylene distributes widely to all tissues of
the body, readily crossing the blood:brain barrier and the placenta (Dallas et al.. 1994a; Ghantous
et al.. 1986; Savolainen et al.. 1977a; Schumann et al.. 1980). The highest tissue concentrations
were found in adipose tissue (60 or more times blood level) and in bngnrain and liver (4 and
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-2 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
5 times blood level, respectively). See Section 3.2 for additional discussion of
tetrachloroethylene distribution.
The metabolism of tetrachloroethylene is an important determinant of its toxicity.
Metabolites are generally thought to be responsible for toxicity—especially to the liver and
kidney. Tetrachloroethylene is metabolized in laboratory animals and in humans through at least
two distinct pathways: oxidative metabolism via the cytochrome P450 (CYP [also abbreviated as
P450 and CYP 450]) mixed-function oxidase system and glutathione (GSH) conjugation
followed by further biotransformation and processing, either through the cysteine conjugate
P-lyase pathway or by other enzymes including FM03 and CYP3 A (Anders et al.. 1988; Birner
et al.. 1996; Costa and Ivanetich. 1980; Daniel. 1963; Dekant et al.. 1987; Dekant et al.. 1989;
Filser and Bolt. 1979; IARC. 1995; Lash and Parker. 2001; Lash et al.. 1998; Pegg et al.. 1979;
U.S. EPA. 1985b. 1991b; Volkel et al.. 1998). The conjugative pathway is toxicologically
significant because it yields relatively potent toxic metabolites (Anders et al.. 1988; Dekant et al..
1986b; Dekant etal.. 1986c; Dekant et al.. 1989; Lash and Parker. 2001; Vamvakas et al.. 1987;
Vamvakas et al.. 1989b; Vamvakas et al.. 1989c; Werner et al.. 1996). Studies in both animals
and humans indicate that overall metabolism of tetrachloroethylene is relatively limited,
particularly at higher exposures (reviewed in Lash and Parker. 2001; U.S. EPA. 1985b. 1991b).
Although thought to be qualitatively similar, there are clear differences among species in the
quantitative aspects of tetrachloroethylene metabolism (Ikeda and Ohtsuii. 1972; Lash and
Parker. 2001; Schumann et al.. 1980; U.S. EPA. 1991b; Volkel et al.. 1998). See Section 3.3 for
additional discussion of tetrachloroethylene metabolism.
Tetrachloroethylene is excreted from the body by pulmonary excretion of the parent
compound and urinary excretion of metabolism products, with a small amount of pulmonary
excretion of metabolism products. Tetrachloroethylene that is not metabolized is exhaled
unchanged, and this process is the primary pathway of tetrachloroethylene excretion in humans
for all routes of administration (Guberan and Fernandez. 1974; Koppel et al.. 1985; Monster.
1979; Opdam and Smolders. 1986; Stewart and Dodd. 1964; 1974; 1977). Pulmonary excretion
of (unchanged) parent compound is also important in animals (Bogen et al.. 1992; Frantz and
Watanabe. 1983; Pegg et al.. 1979; Schumann et al.. 1980; Yllner. 1961). A very small amount
of tetrachloroethylene has been shown to be excreted through the skin (Bolanowska and
Golacka. 1972); however, it represents an insignificant percent of total tetrachloroethylene
disposition. See Section 3.4 for additional discussion of tetrachloroethylene excretion.
As part of this assessment, a PBPK model-based analysis of the population toxicokinetics
of tetrachloroethylene and its metabolites was developed in mice, rats, and humans (also reported
in Chiu and Ginsberg). This model was developed to address many of the limitations of the
existing models for tetrachloroethylene. Among the most important improvements are (1) the
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-3 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
utilization of all the available toxicokinetic data for tetrachloroethylene and its metabolites in
mice, rats, and humans; (2) the incorporation of available information on the internal
toxicokinetics of TCA derived from the most current PBPK modeling of trichloroethylene and
TCA; and (3) the separate estimation of oxidative and conjugation metabolism pathways. This
-harmonized" PBPK model used a limited Bayesian analysis implemented through Markov
chain Monte Carlo approach for parameter calibration. As expected, the major route of
elimination of absorbed tetrachloroethylene is predicted to be exhalation as parent compound,
with metabolism accounting for less than 20% of intake except in the case of mice exposed
orally, in which metabolism is predicted to be slightly over 50% at lower exposures. In all
three species, the concentration in blood, the extent of oxidation, and the amount of TCA
production is well-estimated, with residual uncertainties of-twofold. However, the resulting
range of estimates for the amount of GSH conjugation is quite wide in humans (~3,000-fold) and
mice (~60-fold). While even high-end estimates of GSH conjugation in mice are lower than
estimates of oxidation, in humans the estimated rates range from much lower to much higher
than rates for tetrachloroethylene oxidation. It is unclear to what extent this range reflects
uncertainty, variability, or a combination. Importantly, by separating total tetrachloroethylene
metabolism into separate oxidative and conjugative pathway, this analysis reconciles the
disparity between those previously published PBPK models that predicted either low or high
metabolism in humans. In essence, both conclusions are consistent with the data if augmented
with some additional qualifications: in humans, oxidative metabolism is low, while GSH
conjugation metabolism may be high or low, with uncertainty and/or interindividual variability
spanning three orders of magnitude. More direct data on the internal kinetics of
tetrachloroethylene and GSH conjugation, such as trichlorovinyl glutathione or trichlorovinyl
cysteine levels in blood and/or tissues, would be needed to better characterize the uncertainty and
variability in GSH conjugation in humans. Because of the substantial refinements from previous
PBPK models, this assessment utilizes the Chiu and Ginsberg (In Press) model to calculate
relevant dose-metrics that were then used in dose-response modeling. See Section 3.5 for
additional discussion of and details about PBPK modeling of tetrachloroethylene and
metabolites.
6.1.3. Noncancer Toxicity (see Section 4.10.1)
Noncancer effects of tetrachloroethylene identified in exposed humans and animals
include toxicity to the central nervous system, kidney, liver, immune and hematologic system,
and on development and reproduction. Neurotoxic effects have been characterized in human
controlled exposure, occupational and residential studies, as well as in experimental animal
studies, providing evidence of an association between tetrachloroethylene exposure and
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-4 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
neurological deficits. Tetrachloroethylene exposure primarily results in visual changes,
increased reaction time, and cognitive decrements in humans; in animal studies, effects on
vision, visual-spatial function, and reaction time, as well as brain weight changes were also seen.
Adverse effects on the kidney in the form of tubular toxicity, potentially mediated through
tetrachloroethylene GSH conjugation, have been reported in numerous well-conducted animal
studies. Although human studies have not systematically investigated nephrotoxicity, an
association between tetrachloroethylene exposure via inhalation and chronic kidney disease, as
measured by urinary excretion of renal proteins and end-stage renal disease, is supported. The
developmental and reproductive toxicity database for tetrachloroethylene includes a range of
data from appropriate well-conducted studies in several laboratory animal species plus limited
human data. Evidence of liver toxicity is primarily from several well-conducted rodent studies,
including chronic bioassays.
Other toxicity endpoints are less well characterized. The few published reports of
experimental studies examining immune or hematologic system toxicity are consistent with the
limited findings in the human occupational studies. These include a series of reports by Marth
(1987; 1985a; 1989) providing evidence of hemolytic anemia in young (2-week-old) female mice
exposed at low levels of tetrachloroethylene in drinking water (0.05 or 0.1 mg/kg-day for
7 weeks). The relative lack of additional data, including confirmatory reports of immunotoxic or
hematologic toxicity with low continuous exposures beginning in early lifestages, taken together
with evidence of immunotoxicity from structurally related solvents (Cooper et al.. 2009).
contributes to uncertainty in the database for tetrachloroethylene. No human studies identified
adverse effects on the respiratory tract, and no lung toxicities in rodents were reported in chronic
bioassays [National Toxicology Program [NTP], (1986b); National Cancer Institute [NCI],
(1977)1 or other published reports.
6.1.4. Neurological Effects (see Section 4.1)
The evidence for human neurotoxicity includes controlled experimental chamber
(Altmann et al.. 1990; Hake and Stewart. 1977) and epidemiologic studies that used standardized
neurobehavioral batteries (Altmann et al.. 1995; Echeverria et al.. 1995; Ferroni et al.. 1992;
Hake and Stewart. 1977; Seeber. 1989; Spinatonda et al.. 1997) or employed assessment of
visual function (Cavalleri et al.. 1994; Gobba et al.. 1998; NYSDOH. 2010; Schreiber et al..
2002; Storm et al.. In Press). Most of the studies evaluated neurological effects following an
occupational exposure to tetrachloroethylene (Cavalleri et al.. 1994; Echeverria et al.. 1995;
Ferroni et al.. 1992; Gobba et al.. 1998; Schreiber et al.. 2002; Seeber. 1989; Spinatonda et al..
1997). In addition, three studies examined neurological decrements from residential exposure
(Altmann et al.. 1995; NYSDOH. 2010; Schreiber et al.. 2002; Storm et al.. In Press). Two acute
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-5 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
experimental chamber studies (Altmann et al.. 1990; Hake and Stewart. 1977) have also been
reported. Together, the epidemiologic evidence indicates a broad range of cognitive, motor,
behavioral, and visual functional deficits following tetrachloroethylene exposure (U.S. EPA.
2004).
The research in animal models comprises acute and subchronic studies of the effects of
tetrachloroethylene on functional neurological endpoints (functional observation battery, motor
activity) (Kiellstrand etal.. 1985; Oshiro et al.. 2008). sensory system function as assessed by
evoked potential measurements (Boves et al.. 2009; Mattsson et al.. 1998; U.S. EPA. 1998b). or
pathological changes in the brain (Wang et al.. 1993). These studies, which support the
observations from human studies, reported notable effects on motor activity and motor function
following exposure to tetrachloroethylene in either the adult or the developmental period as well
as changes in evoked potentials following acute and subchronic exposures. In addition,
postmortem effects in animals were observed with pathological alterations in brain DNA, RNA,
or protein levels and brain weight changes.
In conclusion, the weight of evidence across the available studies of humans and animals
exposed to tetrachloroethylene indicates that chronic exposure to tetrachloroethylene can result
in decrements in color vision, visuospatial memory, and possibly other aspects of cognition and
neuropsychological function, including reaction time.
6.1.5. Summary of Other Noncancer Adverse Effects (see Sections 4.2, 4.3, 4.6, and 4.7)
In addition to evidence of toxicity to the central nervous system, tetrachloroethylene has
been shown to adversely affect the kidney, liver, immune and hematologic system, as well as
development and reproduction. The human and animal evidence for these effects is summarized
in the paragraphs below.
6.1.5.1. Kidney Toxicity (see Section 4.2)
The human evidence for kidney effects is limited because most available reports do not
include information on even a minimal core battery of tests for kidney function and only
one study reported on end-stage renal disease (ESRD). However, an association between
tetrachloroethylene exposure via inhalation and chronic kidney disease is supported by evidence
of urinary excretion of renal proteins (Mutti et al.. 1992; Verplanke et al.. 1999) and higher
ESRD, particularly hypertensive ESRD, with higher exposures (Calvert et al.. In Press). Mutti
et al. (1992) reported statistically significant increases in RBP, p2[j,-globulin, and albumin in
urine among dry cleaners as compared with matched controls. In addition, for seven different
urinary markers, the prevalence of individuals with abnormal values (>95th percentile of
controls) was four- to fivefold greater in the exposed group. Adverse effects on the kidney have
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-6 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
been observed in studies of animals exposed to high concentrations of tetrachloroethylene by
inhalation (JISA. 1993; 1986b), oral gavage (Ebrahim et al.. 1996; Green et al.. 1990; Jonker et
al.. 1996; NCI. 1977)(Ebrahim 2002)(Goldsworthy et al., 1988) and by intraperitoneal injection
of tetrachloroethylene metabolites (Elfarra and Krause. 2007). The nephrotoxic effects include
increased kidney-to-body weight ratios, hyaline droplet formation, glomerular nephrosis,"
karyomegaly (enlarged nuclei), cast formation, and other lesions or indicators of renal toxicity.
Overall, multiple lines of evidence support the conclusion that tetrachloroethylene causes
nephrotoxicity in the form of tubular toxicity, mediated potentially through GSH conjugation
products. Limitations to the database include the lack of human studies investigating drinking
water or other oral tetrachloroethylene exposures on kidney toxicity.
6.1.5.2. Liver Toxicity (see Section 4.3)
Two of four studies of occupationally exposed dry cleaners showed early indications of
liver toxicity, namely sonographic changes of the liver and altered serum concentrations of
one liver enzyme indicative of liver injury (Brodkin et al.. 1995; Gennari et al.. 1992). Frank
liver disease was not seen among these workers, nor were changes in other biomarkers indicative
of liver toxicity (e.g., serum transaminases), not unexpected given that subjects with signs of
liver disease were excluded in both studies. Liver toxicity was reported in multiple animal
species exposed to tetrachloroethylene via inhalation and oral routes of exposure. The effects
were characterized by increased liver weight, fatty changes, necrosis, inflammatory cell
infiltration, triglyceride increases and proliferation (Berman et al.. 1995; Buben and O'Flahertv.
1985; Ebrahim et al.. 1996; Goldsworthy and Popp. 1987; JISA. 1993; Jonker et al.. 1996;
Kiellstrand et al.. 1984; NTP. 1986b; Odum et al.. 1988b; Philip et al.. 2007; Schumann et al..
1980).
6.1.5.3. Immunologic and Hematopoetic Toxicity (see Section 4.6)
The strongest human study examining immunologic and hematologic effects of
tetrachloroethylene exposure in terms of sample size and use of an appropriately matched control
group is the study of 40 male dry-cleaning workers (mean exposure levels <140 ppm; mean
duration 7 years; mean blood tetrachoroethylene levels 1,685 |ig/L) by Emara et al. (2010).
Statistically significant decreases in red blood cell count and hemoglobin levels and increases in
total white cell counts and lymphocyte counts were seen in the exposed workers compared to
age- and smoking-matched controls. Similar effects were seen in mice (Ebrahim et al.. 2001). In
addition, increases in several other immunological parameters, including T lymphocyte and
natural killer cell subpopulations, IgE, and interleukin-4 levels were observed in
tetrachloroethylene-exposed dry-cleaning workers (Emara et al.. 2010). These immunologic
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-7 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
effects suggest an augmentation of Th2 responsiveness. The available data from experimental
studies assessing immunotoxic responses in animals are very limited (Aranyi et al.. 1986;
Germolec etal.. 1989; Hani oka et al.. 1995a). with one study (Aranyi et al.. 1986) suggesting
that short-term exposures may result in decreased immunological competence
(immunosuppression) in CD-I mice. The limited laboratory animal studies of hematological
toxicity demonstrated an effect on red blood cells [decreased RBC (Ebrahim et al.. 2001). or
decreased erythrocyte colony forming units (Seidel et al.. 1992)1. with reversible hemolytic
anemia observed in female mice exposed to low drinking water levels (0.05 mg/kg-bw day) of
tetrachloroethylene beginning at 2 weeks of age in one series of studies (Marth. 1987; Marth et
al.. 1985b; 1989). Ebrahim et al. (2001) also observed decreased hemoglobin, platelet counts
and packed cell volume, and increased WBC counts. The results of these studies, while limited,
support the human epidemiology studies. Additional data from inhalation, oral, and dermal
exposures of different durations are needed to assess the potential immunotoxicity of
tetrachloroethylene along multiple dimensions—including immunosuppression, autoimmunity,
and allergic sensitization. The relative lack of additional data, including confirmatory reports of
immunotoxic or hematologic toxicity with low continuous exposures beginning in early
lifestages, taken together with evidence of immunotoxicity from structurally related solvents
(Cooper et al.. 2009). contributes to uncertainty in the database for tetrachloroethylene.
6.1.5.4. Reproductive Toxicity (see Section 4.7)
The epidemiologic database is inconclusive concerning potential effects of
tetrachloroethylene exposure on spermatogenesis, menstruation, fertility or delayed conception
(Eskenazi etal.. 1991a; Eskenazi etal.. 1991b; Rachootin and Olsen. 1983; Sallmen et al.. 1998;
Sallmen et al.. 1995; Zielhuis et al.. 1989). One study of primarily unionized workers in the dry-
cleaning and laundry industries in California observed subtle deficits in sperm quality in relation
to increasing levels of three measures of exposure, including tetrachloroethylene in exhaled
breath (Eskenazi et al.. 1991b). This observation is supported by one report of abnormal sperm
in mice (Beliles et al.. 1980). Several studies of maternal occupational exposure to
tetrachloroethylene suggest an increased risk of spontaneous abortion, particularly at higher
levels (Doyle et al.. 1997; Kyyronen etal.. 1989; Lindbohm et al.. 1990; Olsen et al.. 1990;
Windham et al.. 1991). but other studies did not report an association with maternal (Ahlborg.
1990b; Olsen et al.. 1990) or paternal (Eskenazi et al.. 1991b; Lindbohm et al.. 1991; Taskinen et
al.. 1989) exposure. Some studies observed an increased odds ratio ranging from 1.4 to 4.7, but
risk estimates were statistically imprecise and some studies were limited in their ability to
evaluate potential confounding (Bosco et al.. 1987; Lindbohm et al.. 1990; Olsen et al.. 1990;
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-8 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Windham et al.. 1991). In general, the studies that used a more precise definition of exposure, or
categorized exposure into levels of increasing dose or intensity, observed higher risk estimates
(Doyle et al.. 1997; Kyyronen et al.. 1989; Lindbohm et al.. 1990; Olsen et al.. 1990). No
associations with incidence of spontaneous abortion were observed among two populations
exposed to tetrachloroethylene in drinking water, although the window of exposure used to
assess risk in both studies may not have had been precise enough to detect a small elevation in
risk (Aschengrau et al.. 2008; Aschengrau et al.. 2009a; Lagakos et al.. 1986). The finding of
spontaneous abortions in several human studies of dry cleaners is supported by the occurrence of
reduced birth weight and mortality in several animal studies (Carney et al.. 2006; Nelson et al..
1980; Schwetz et al.. 1975; Szakmary et al.. 1997; and in the F1 generation but not the F2
generation of Tinston. 1994).
6.1.5.5. Developmental Toxicity (see Section 4.7)
Stillbirths, congenital anomalies, or decreased birth weight were not associated with
maternal or paternal occupational exposure to tetrachloroethylene in several epidemiologic
studies (Bosco et al.. 1987; Kyyronen etal.. 1989; Lindbohm. 1995; Olsen et al.. 1990; Taskinen
et al.. 1989; Windham et al.. 1991). However, the studies analyzed congenital anomalies in a
combined category, and the number of exposed cases for specific types of anomalies was not
sufficient to evaluate risk with statistical precision. Some studies of tetrachloroethylene in
drinking water reported that exposure during pregnancy is associated with low birth weight
(Bove et al.. 1995; Lagakos et al.. 1986). eye/ear anomalies (Lagakos et al.. 1986). and oral clefts
(Aschengrau et al.. 2009b; Bove et al.. 1995; Lagakos et al.. 1986). No associations with
prenatal tetrachloroethylene exposure in drinking water were reported for small for gestational
age (Aschengrau et al.. 2008; Bove et al.. 1995). other classifications of congenital anomalies
(e.g.. musculoskeletal, cardiovascular; Lagakos et al.. 1986). or deficits in attention or
educational performance (Janulewicz et al.. 2008). Although a small increase in risk of small for
gestational age was reported for infants exposed prenatally to tetrachloroethylene at the Camp
Lejeune military base, the finding remains inconclusive until ATSDR completes its reanalysis
(Sonnenfeld et al.. 2001). Participants in some of the studies of drinking water contamination
were exposed to multiple pollutants (Bove et al.. 1995; Lagakos et al.. 1986). and it was not
possible to disentangle substance-specific risks. In animals, the developmental toxicity database
provides evidence of decreased prenatal survival, decreased fetal growth, delays in skeletal
ossification, and increased incidences of malformations following in utero exposure in rats, mice,
and/or rabbits (Carney et al.. 2006; Narotsky andKavlock. 1995; Schwetz et al.. 1975; Szakmary
et al.. 1997). The decreased survival and malformation findings in laboratory mammals were
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-9 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
supported by data from whole embryo culture (Saillenfait et al.. 1995) and Japanese medaka
assays (Saillenfait et al.. 1995); Spencer et al., 2001). Alterations in neurological function
following pre- and/or postnatal inhalation exposures to tetrachloroethylene were observed in rats
by Szakmary et al. (1997). Nelson et al. (1980). Fredriksson et al. (1993). and Tinston (1994).
These findings were supported by a study that found reductions in brain acetylcholine and
dopamine in rat offspring following gestational tetrachloroethylene exposures (Nelson et al..
1980). Limitations of the inhalation developmental toxicity studies include the lack of
dose-response information due to the use of a single treatment level in the prenatal
developmental toxicity assessment by Schwetz et al., (1975); the lack of either maternal or
developmental toxicity in Hardin et al., (1981); and absence of methodological details in study
reporting (Szakmary et al.. 1997).
6.1.6. Carcinogenicity (see Section 4.10.2)
Following EPA (2005a) Guidelines for Carcinogen Risk Assessment, tetrachloroethylene
is 4ikely to be carcinogenic in humans by all routes of exposure". This characterization is based
on suggestive evidence of carcinogenicity in epidemiologic studies and conclusive evidence that
the administration of tetrachloroethylene, either by ingestion or by inhalation to sexually mature
rats and mice, increases tumor incidence (JISA. 1993; NCI. 1977; NTP. 1986b).
Tetrachloroethylene increased the incidence of liver tumors (hepatocellular adenomas and
carcinomas) in male and female mice and of mononuclear cell leukemia (MCL) in both sexes of
rats. These findings were reproducible in multiple lifetime bioassays employing different rodent
strains and, in the case of mouse liver tumors, by inhalation and oral exposure routes. Additional
tumor findings in rats included significant increases in the NTP bioassay of testicular interstitial
cell tumors and kidney tumors in males, and brain gliomas in males and females. In mice,
hemangiosarcomas in liver, spleen, fat, and subcutaneous skin were reported in males in the
JISA study. The available epidemiologic studies provide a pattern of evidence associating
tetrachloroethylene exposure and several types of cancer, specifically bladder cancer,
non-Hodgkin lymphoma, and multiple myeloma. Associations and exposure-response
relationships for these cancers were reported in studies using higher quality (more precise)
exposure-assessment methodologies for tetrachloroethylene. Confounding by common lifestyle
factors such as smoking are unlikely explanations for the observed results. For other sites,
including esophageal, kidney, lung, liver, cervical, and breast cancer, more limited data are
available.
The specific active moiety(ies) and mode(s) of action involved in the carcinogenicity of
tetrachloroethylene and its metabolites are fully characterized. For rat kidney tumors, it is
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-10 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
generally believed that metabolites resulting from GSH conjugation of tetrachloroethylene are
involved. The hypothesized modes of action for this endpoint include mutagenicity, peroxisome
proliferation, a2[j,-globulin nephropathy, and cytotoxicity not associated with a2[j,-globulin
accumulation. For mouse liver tumors, it is generally believed that metabolites resulting from
P450-mediated oxidation of tetrachloroethylene are involved. The mode of action (MO A)
hypotheses for this endpoint concern mutagenicity, epigenetic effects (especially DNA
hypomethylation), oxidative stress, and receptor activation (focusing on a hypothesized PPARa
activation MO A). However, the available evidence is insufficient to support the conclusion that
either rat kidney or mouse liver tumors are mediated solely by one of these hypothesized modes
of action. In addition, no data are available concerning the metabolites or the mechanisms that
may contribute to the induction of other rodent tumors (including mononuclear cell leukemia,
brain gliomas, or testicular interstitial cell tumors in exposed rats and hemangiosarcomas in
exposed mice). Furthermore, no mechanistic hypotheses have been advanced for the human
cancers suggested to be increased with tetrachloroethylene exposure in epidemiologic studies,
including bladder cancer, non-Hodgkin lymphoma and multiple myeloma. Although
tetrachloroethylene is largely negative in genotoxicity assays—including in the Ames
mutagenicity test—tetrachloroethylene has been shown to induce modest genotoxic effects (e.g.,
micronuclei induction following in vitro or in vivo exposure, and DNA binding and single strand
breaks in tumor tissue) and mutagenic effects under certain metabolic activation conditions. In
addition, some tetrachloroethylene metabolites have been shown to be mutagenic. Thus, the
hypothesis that mutagenicity contributes to the tetrachloroethylene carcinogenesis cannot be
ruled out for one or more target organs, although the specific metabolic species or mechanistic
effects are not known.
6.1.7. Susceptibility (see Section 4.9)
There is some evidence that certain populations might be more susceptible to exposure to
tetrachloroethylene. Attributes that may increase susceptibility to tetrachloroethylene include
age, gender, race/ethnicity, genetics, preexisting disease, lifestyle factors, nutritional status,
socioeconomic status, and multiple exposures and cumulative risk. Although there is more
information on early life exposure to tetrachloroethylene than on other potentially susceptible
populations, there remain a number of uncertainties regarding childhood susceptibility.
Although inhalation of tetrachloroethylene is believed to be of most concern, pathways of
exposure for children are not well characterized. It is not clear to what extent tetrachloroethylene
may pass through the placenta in humans, as shown in rodent studies (Ghantous et al.. 1986;
Szakmary et al.. 1997); for some infants the primary route of exposure may be through breast
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-11 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
milk ingestion (see Sections 2.2.4 and 3.2), while for other infants the dose received through
ingestion of breast milk will become insignificant when compared with inhalation exposure and
subsequent dose (Schreiber. 1997); the amount of tetrachloroethylene ingested from food is not
well described; and it is not known to what extent tetrachloroethylene is absorbed by a child and
to which organs tetrachloroethylene and its metabolites may be distributed. The neurological
effects of tetrachloroethylene may constitute the most sensitive endpoints of concern for
noncancer effects, and limited data show that early life-stages may be more susceptible to visual
deficits than are adults (NYSDOH. 2005c. 2010; Schreiber et al.. 2002; Storm et al.. In Press),
yet developmental neurotoxic effects, particularly in the developing fetus, need further
evaluation using age-appropriate testing for assessment. There are a number of adverse health
effects observed uniquely in early lifestages, with no comparable observations in adults to
determine relative sensitivity (e.g., birth outcomes, autism, allergy); conversely, there are some
adverse outcomes that have been observed only in adults.
There is suggestive evidence that there may be greater susceptibility among the elderly,
but the available data is much more limited with related uncertainties. Improved PBPK
modeling that contains physiologic parameter information for infants and children (including, for
example, the effects of maternal inhalation exposure and the resulting concentration in breast
milk) and for older adults, and validation of these models, will aid in determining differences in
life stage toxicokinetics of tetrachloroethylene. The differences reported in the literature may
reflect a true difference in susceptibility by life stage, an incomplete assessment of these
outcomes in all life-stages, or latent outcomes associated with earlier exposure. More studies
specifically designed to evaluate effects in early and later life-stages are needed in order to more
fully characterize potential life stage-related tetrachloroethylene toxicity.
For other susceptibility factors, the data are more limited and based mainly on
nonchemical specific data that provides information on variation in physiology, exposure, and
toxicokinetics. Until quantitative conclusions can be made for each susceptibility factor, it will
be very hard to consider the impacts of changes in multiple susceptibility factors. In addition,
further evaluation of the effects of aggregate exposure to tetrachloroethylene from multiple
routes and pathways is needed. Similarly, the effects due to coexposures to other compounds
with similar or different MO As need to be evaluated.
6 2 DOSE-RESPONSE ASSESSMENT
This section summarizes the major conclusions of the dose-response analysis for
tetrachloroethylene noncancer effects and carcinogenicity, with more detailed discussions in
Section 5.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-12 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
6.2.1. Noncancer Effects (see Section 5.1)
6.2.2. Selection of Critical Effect and Principal Studies (see Section 5.1.1)
The database of human and animal studies on inhalation toxicity of tetrachloroethylene is
adequate to support derivation of inhalation and oral reference values. A number of targets of
toxicity from chronic exposure to tetrachloroethylene have been identified in published animal
and human studies. These targets include the central nervous system, kidney, liver, immune and
hematologic system, and development and reproduction. In general, neurological effects were
judged to be associated with lower tetrachloroethylene exposures.
The evidence for human neurotoxicity includes 12 well-conducted epidemiological
studies of tetrachloroethylene exposure. Of these, seven examined occupational exposure (i.e..
Cavalleri etal.. 1994; Echeverria et al.. 1995; Ferroni et al.. 1992; Gobba et al.. 1998; Schreiber
et al.. 2002; Seeber. 1989; Spinatonda et al.. 1997). three examined residential exposure (i.e.,
Altmann et al.. 1995; NYSDOH. 2010; Schreiber et al.. 2002; Storm et al.. In Press) and two
were acute-duration experimental chamber studies (i.e., Altmann et al.. 1990; Hake and Stewart.
1977). The animal database comprises acute-duration and subchronic-duration studies of the
effects of tetrachloroethylene on functional neurological endpoints (functional observation
battery, motor activity) (i.e., Ki ell strand et al.. 1985; Oshiro et al.. 2008). on sensory system
function as assessed by evoked potential (i.e.. Boves et al.. 2009; Mattsson et al.. 1998; U.S.
EPA. 1998b). or pathological changes in the brain (i.e.. Wang et al.. 1993).
Principal study selection from these candidate studies of central nervous system effects
involved evaluation of study characteristics as identified in Table 5-2. To summarize, human
studies are preferred to animal studies, as are studies of chronic duration and in residential
settings. Residential exposure is more likely to be continuous and of lower concentrations
compared with the more intermittent, higher concentration exposures experienced in work
settings. Three human studies were considered to be more methodologically sound based on
study quality attributes, including study population selection, exposure measurement methods,
and endpoint measurement methods. Thus, three studies—Seeber (1989). Cavalleri et al. (1994).
and Echeverria et al. (1995)—were judged to be principal studies for deriving a reference
concentration [RfC], none of which is a clearly superior candidate for identifying the point of
departure [POD], Endpoints selected for the RfC were reaction time measures (Echeverria et al..
1995). cognitive changes (Echeverria et al.. 1995; Seeber. 1989). and visual function changes
(Cavalleri et al.. 1994).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-13 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
6.2.3. Uncertainties and Application of Uncertainty Factors (UFs) (see Sections 5.1.3, 5.2.3,
and 5.3)
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, those 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 data set, of qualitative uncertainty.
Each of the candidate studies provided lowest-observed-adverse-effect levels (LOAELs) that
were selected as PODs. The adjusted LOAELs are as follows: 56 mg/m3 (for either visual
reproduction, pattern memory, and pattern recognition, or reaction time in pattern memory in
Echeverria et al.. 1995); 29 mg/m3 (for digit symbol cancellation, digit reproduction, and
perceptual speed in Seeber. 1989); and 15 mg/m3 (for color confusion in Cavalleri et al.. 1994).
No adjustment of the PODs was needed for animal-to-human extrapolation uncertainty.
Additionally, no adjustment was needed for subchronic-to-chronic uncertainty because the
principal studies involved chronic exposures. An overall uncertainty factor of 1,000 was applied
to each selected POD, comprised of the following uncertainty factors (UFs)
6.2.3.1. Human Variation
The UF of 10 was applied for human variation for all of the studies that were selected in
derivation of the RfC. These studies are from occupationally exposed subjects, who are
generally healthier than the overall population, and thus provide no data to determine the relative
effects of susceptible population including children, elderly, and/or people with compromised
health. Additionally, no information was presented in the human studies with which to examine
variation among subjects.
6.2.3.2. LOAEL-to-NOAEL Uncertainty
A UF of 10 is generally applied when the POD is a LOAEL due to a lack of a
no-observed-adverse-effect level [NOAEL], When NOAELs are used, a UF is not applied. For
all of the human studies and endpoints selected (Cavalleri etal.. 1994; Echeverria et al.. 1995;
Seeber. 1989). PODs were LOAELs and a UF of 10 was applied to these endpoints.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-14 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
6.2.3.3. Database Uncertainty
A database UF of 10 has been applied to address the lack of data to adequately
characterize the hazard and dose-response in the human population. A number of data gaps were
identified from both the human and animal literature, including the need for high quality
epidemiologic studies of residential exposures including children and the elderly, and
chronic-duration animal studies (including in developing animals) designed to define and
characterize the exposure-response relationships for the observed neurotoxicological effects,
particularly, reaction time, cognitive and visual function. Additionally, the available studies of
immunologic and hematologic toxicity studies (e.g.. Emara et al.. 2010; Marth. 1987) are limited,
but do raise concern for risk at exposures lower than those evaluated. The relative lack of data
taken together with the concern that other structurally related solvents have been associated with
immunotoxicity, particularly relating to autoimmune disease (Cooper et al.. 2009). contributes to
uncertainty in the database for tetrachloroethylene.
In addition, the available epidemiologic studies of residential exposures were judged to
be more limited for developing an RfC (Altmann et al.. 1995; NYSDOH. 2010; Schreiber et al..
2002; Storm et al.. In Press) based on consideration of selection bias, residual confounding
(population comparability) and/or selection of neurological methods. Yet the residential studies
yielded the most sensitive neurotoxic endpoint associated with tetrachloroethylene exposure,
decrement in visual contrast sensitivity (VCS). Because this specific endpoint was not evaluated
in any of the occupational studies, it cannot be concluded that similar or even greater VCS
changes would not occur at the higher exposures of the occupational studies. There were
impairments in Color Confusion Index for one set of occupationally exposed subjects (Cavalleri
et al.. 1994; Gobbaetal.. 1998). but this effect was not evaluated in other occupational studies.
There is also a lack of studies which evaluated the critical effects of reaction time, cognitive and
visual functional deficits in populations exposed to tetrachloroethylene at lower than the studied
occupational exposure levels, including at residential levels. These data gaps, and the lack of
developmental and immune functional assessment, therefore, represent significant uncertainty in
the tetrachloroethylene database.
6.2.4. Reference Concentration (see Section 5.1.3)
Based on the application of an overall uncertainty factor of 1,000 to each selected POD
from four different endpoints in the three principal studies (i.e., Cavalleri et al.. 1994; Echeverria
et al.. 1995; Seeber. 1989). candidate RfCs ranged from 0.015 to 0.056 mg/m3. A value of
0.04 mg/m3 is supported by these multiple studies, as a midpoint of the range of available values,
and is the recommended RfC for tetrachloroethylene.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-15 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
6.2.5. Reference Dose (see Section 5.2)
A reference dose [RfD] for tetrachloroethylene was developed through a route-to-route
extrapolation from the PODs in three neurotoxicological studies of occupational
tetrachloroethylene exposure (i.e.. Cavalleri et al.. 1994; Echeverria et al.. 1995; Seeber. 1989).
The harmonized PBPK model of Chiu and Ginsberg was used to derive the oral dose that would
result in the same tetrachloroethylene in blood area under the curve (AUC) as that following a
continuous inhalation exposure LOAELs from the three critical studies. Although it is not clear
if the noncancer effects observed in humans are the result of tetrachloroethylene itself and/or
one or more metabolites, tetrachloroethylene in the blood can safely be presumed to be a step in
the toxicity pathway. Moreover, the sensitivity to the choice of dose metric for route-to-route
extrapolation is low, with alternative dose metrics giving route-to-route conversions within
1.4-fold of the conversion based on tetrachloroethylene in blood. The resulting PODs were
2.7 mg/kg-day (Cavalleri etal.. 1994). 4.3 mg/kg-day (Echeverria et al.. 1995) and
4.6 mg/kg-day (Seeber. 1989). The same composite UF of 1,000 that was used for the RfC
derivation was applied for each of these PODs, yielding RfDs ranging from 2.6 x 10 3 to
9.7 x 10 3 mg/kg-day. From this, an RfD of 6 x 10~3 mg/kg-day is supported by these multiple
studies, as a midpoint of the range of available values rounded to one significant figure, and is
the recommended RfD for tetrachloroethylene. This RfD is equivalent to a drinking water
concentration of 0.21 mg/L, assuming a body weight of 70 kg and a daily water consumption of
2L.
6.2.6. Dose-Response Analyses for Noncancer Effects Other Than Critical Effect of
Neurotoxicity (see Sections 5.1.4 and 5.2.4)
Inhalation and oral dose-response analyses for noncancer effects other than the critical
effect of neurotoxicity were also conducted. The purpose of these analyses is twofold: (1) to
provide a quantitative characterization of the relative sensitivity of different organs/sy stems to
tetrachloroethylene, and (2), to provide information that may be useful for cumulative risk
assessment in which multiple chemicals have a common target organ/system other than the
central nervous system. The method of analysis is analogous to that described above for
neurotoxicity, using the NOAEL/LOAEL approach and the application of uncertainty factors to
studies of kidney, liver, immunologic and hematologic, and reproductive and developmental
toxicity. Specifically, human equivalent concentrations [HECs] and human equivalent doses
[HEDs] are derived using either (1) for inhalation exposure, the RfC methodology for a
category 3 gas, extrathoracic effects, adjusted for equivalent continuous exposure; (2) for oral
exposure, mg/kg-day dose adjusted for equivalent continuous exposure; or (3) for either route of
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-16 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
exposure, the PBPK model with an appropriate dose metric. The HECs and HEDs are then
treated as PODs to which uncertainty factors are applied.
The sample values for two outcomes domains—renal and hematologic toxicity—overlap
with the range of values based on the critical effect of neurotoxicity, thereby supporting the
selection of the critical effect. Specifically, for renal effects, the resulting values range from
0.03-0.2 mg/m3 for inhalation and 0.005-0.03 mg/kg-day for oral exposure, based on effects in
chronically exposed mice and rats (JISA. 1993) and occupationally exposed humans (Mutti et al..
1992). For hematologic toxicity, the resulting values were 0.04 mg/m3 for inhalation and
0.007 mg/kg-day for oral exposure, based on changes in hematological measures in
occupationally exposed humans (Emara et al.. 2010). These overlap with the ranges of
0.02-0.06 mg/m3 for inhalation and 0.003-0.01 mg/kg-day for oral exposure based on the
critical effect of neurotoxicity, and thereby providing additional support for the recommended
RfC and RfD. The sample values from the other outcome domains are less than 20-fold greater
than the RfC, and less than 10-fold greater than the RfD. This suggests that multiple effects may
occur at about the same exposure levels at which tetrachloroethylene begins to induce
neurotoxicity. These results also suggest that it is important to take into account effects from
tetrachloroethylene other than neurotoxicity when assessing the cumulative effects of multiple
exposures.
6.2.7. Cancer (see Section 5.2)
As summarized above, following EPA (2005a) Guidelines for Carcinogen Risk
Assessment, tetrachloroethylene is characterized as "Likely to be carcinogenic to humans" by all
routes of exposure based on some epidemiologic evidence and conclusive evidence in mice and
rats. No available epidemiologic studies of cancer were found to be suitable for dose-response
modeling assessment. Therefore, the following dose-response assessment is based on data from
rodent bioassays. Because the MO As for tetrachloroethylene carcinogenicity are not fully
characterized, the tumors reported in rodent bioassays are considered relevant to humans and a
low-dose linear extrapolation is used to estimate human cancer risk from rodent dose-response
data, in accordance with EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005a).
The several chronic studies in rats and mice include an oral gavage study in mice and rats
by NCI (1977) and two inhalation studies in mice and rats (JISA. 1993; NTP. 1986b). The NCI
(1977) rat and mouse oral gavage study had a number of limitations that made it less suitable for
dose-response modeling as compared to the other studies, including significantly higher early
noncancer morbidity and mortality in treated groups, a variable dosing schedule, and dosing for
less than the full duration of the bioassay. With respect to the other two bioassays, the JISA
(1993) bioassay included lower exposures of both mice and rats than the NTP (1986b) study, and
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-17 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
it included three exposure groups as compared to two exposure groups in the NTP (1986b) study.
Therefore, JISA (1993) provides a stronger basis for deriving dose-response relationships for risk
assessment purposes, insofar as all other aspects of these studies can be considered comparable.
Thus, for endpoints which were reported to be tetrachloroethylene-related in multiple
studies—i.e., liver tumors and mononuclear cell leukemias—the JISA (1993) study was used for
dose-response modeling. The JISA (1993) bioassay was also used for dose-response modeling
of the increased hemangiomas and hemangiosarcomas in male mice because it was the only
bioassay that reported this tumor type. The NTP (1986b) study was utilized for modeling the
increased incidence in renal cancers, brain cancers, and testicular tumors in male rats with
treatment, which were reported only in this bioassay. In male mice and male rats, multiple
treatment-related tumors were reported in the same study ((JISA. 1993). and (NTP. 1986b).
respectively); thus, the dose-response analyses of the combined risk of multiple tumors for those
experiments were also conducted.
The harmonized PBPK model of Chiu and Ginsberg (In Press) was used to perform the
interspecies extrapolation from rodents to humans, and for route-to-route extrapolation of the
inhalation bioassay results to oral exposures. The choice of the preferred dose-metric to use for
each endpoint was based on the strength of its association with the toxic moiety relevant to the
endpoint and an evaluation of uncertainties in the calculation of that dose-metric. For cancer,
total rate of oxidative metabolism in the liver was considered the most relevant dose metric for
tetrachloroethylene-induced liver tumors, and AUC of the parent compound in the blood was
considered the preferred dose metric for all other sites, including MCL. Alternative dose-metrics
were also used for the purposes of comparison. These include the AUC of TCA in the liver for
mouse liver tumors and the rate of GSH conjugation for rat kidney tumors.
6.2.8. Choice of Data Set for Use in Cancer Risk Estimation
The choice of data set for best representing an upper bound estimate of human
carcinogenic potency involves a number of factors, including the magnitude and robustness of
the response, the role of metabolism, the carcinogenic MO As, and the dose-response model fit
and resulting low-dose extrapolation predictions.
The highest magnitude and most robust responses for tetrachloroethylene carcinogenicity
in rodents are the increased incidences of liver tumors (hepatocellular adenomas and carcinomas)
in both sexes of mice and of MCL in both sexes of rats, with biologically and statistically
significant increases over background (see Section 5.4.2). These were also reported in multiple
bioassays. Other reported endpoints—including hemangiosarcomas in male mice, testicular
interstitial cell tumors and kidney tumors in male rats, and brain gliomas in male and female
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-18 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
rats—had lower incidences in terms of additional risk and were not reported in multiple studies.
Therefore, on the basis of this factor, mouse liver tumors and rat MCLs carry the greatest weight.
In terms of the role of metabolism, the specific toxic moieties have not been identified for
any endpoint. However, for mouse liver tumors and rat kidney tumors there are data that identify
the likely metabolic pathway involved—oxidation and GSH conjugation, respectively. A PBPK
model was developed to quantitatively account for species differences in these metabolic
pathways, but data were only adequate to address differences in oxidation. For GSH
conjugation, and for other endpoints for which the active metabolic pathway is unknown, the
AUC of tetrachloroethylene in blood was used as a dose metric, with the rationale that it is more
proximate to toxicity than administered dose.
In terms of MO A, only for rat kidney tumors and mouse liver tumors are there any
concrete hypotheses. However, the available evidence is insufficient to support the conclusion
that rat kidney or mouse liver tumors are mediated solely by one of these hypothesized modes of
action. In addition, no data are available concerning the mechanisms that may contribute to the
induction of other rodent tumors. Furthermore, no mechanistic hypotheses have been advanced
for the human cancers suggested to be increased with tetrachloroethylene exposure in
epidemiologic studies. Overall, the MO As involved in the carcinogenicity of tetrachloroethylene
and its metabolites are not fully characterized, and, thus, all the tumors observed in rodent
bioassays are considered relevant to humans, in accordance with EPA's Guidelines for
Carcinogen Risk Assessment (U.S. EPA. 2005a). Therefore, mechanistic data are not
informative as to the choice of data set that may best reflect human carcinogenic potency.
The final factor involves the dose-response model fit and resulting low-dose
extrapolation predictions. In terms of model fit, there is some uncertainty, particularly for
nonmonotonic and/or supralinear data sets such as MCLs and liver tumors; and, a number of
options were pursued to try to address this issue. Statistical parameter uncertainty at the POD is
quite modest at around twofold or less for all data sets except the combined male and female
MCL fits, which had statistical uncertainty at the POD of around fivefold. The dose-response
analysis of the combined male and female MCL data (JISA. 1993) resulted in the highest unit
risk estimated using a preferred dose metric. Some estimates based on alternative dose metrics
are higher than those based on the preferred dose metrics, reflecting uncertainty with respect to
the active moiety or metabolic pathway. Estimates of total tumor risk from these studies were at
most threefold lower than the most sensitive result based on male and female MCL data (JISA.
1993). Overall, due to its being the most sensitive endpoint and its having an acceptable degree
of dose-response modeling uncertainty, the combined male and female MCL data carry the
greatest weight from this perspective.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-19 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Given the significant gaps in the scientific knowledge regarding the metabolites and
mechanisms contributing to tetrachloroethylene-induced cancer, the factors given the strongest
consideration in selection among the available data set were the magnitude and robustness of the
response and the dose-response model fit and resulting low-dose extrapolation predictions.
Based on these factors, the dose-response analyses of the combined male and female rat MCL
from JISA (1993) were selected. This selection is supported by the strong and robust observed
response combined with the finding of the most sensitive unit risk estimate using a preferred
dose metric.
6.2.9. Inhalation Unit Risk Estimate (see Section 5.4.4.3)
The inhalation unit risk for tetrachloroethylene is defined as a plausible upper bound
lifetime extra risk of cancer from chronic inhalation of tetrachloroethylene per unit of air
concentration. The recommended inhalation unit risk value is 4 x 10~2 per ppm, or 6 x 10~6 per
jig/m3, based on the combined male and female rat MCL data, using preferred dose metrics and
modeling approaches. This estimate is based on the most sensitive endpoint modeled using
preferred dose metrics, with estimates from other tumor sites using preferred dose metrics (total
oxidative metabolites for hepatocellular tumors, tetrachloroethylene AUC in blood for all other
tumors) being lower by between three- and 30-fold. The recommended inhalation unit risk is
also within twofold of estimates of total tumor risk from multiple sites (brain, kidney, testes, and
MCL) in the NTP rat bioassay, using tetrachloroethylene AUC in blood as the preferred dose
metric. Estimates using alternative dose metrics (TCA AUC for hepatocellular tumors, GST
metabolism for kidney tumors) spanned a range from almost three orders of magnitude below to
more than twofold above the recommended inhalation unit risk.
Confidence in the recommended inhalation unit risk estimate is further increased by its
concordance with estimates reported by VanWinjngaarden and Hertz-Piccioto (2004) and Finkel,
(2010), based on two epidemiologic studies (Lynge et al., 2006; Vaughan et al., 1997), which
have central estimates ranging from 2 x 10 6 to 8 / 10 6 per |ig/m3 and upper bound estimates
ranging from 8 x 10 6 to l 6 / 10 6 per |ig/m3. The two such estimates available use average
tetrachloroethylene concentration as the exposure surrogate, either the time-weighted average or
average level from industrial monitoring studies, they assume that bladder cancer or laryngeal
cancer are the only carcinogenic hazard in humans, and they may be subject to some other
sources of bias, but provide information without extrapolation from animals to humans.
Therefore, although the studies lack estimates of tetrachloroethylene exposure intensity to
individual study subjects, precluding their use as a primary basis for dose-response assessment,
the estimates based on these human data support the plausibility of the cancer risk estimates
based on rodent bioassays.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-20 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
6.2.10. Oral Slope Factor Estimate (see Section 5.4.4.4)
The oral slope factor for tetrachloroethylene is defined as a plausible upper bound
lifetime extra risk of cancer from chronic ingestion of tetrachloroethylene per mg/kg-day oral
dose. Due to limitations in the oral bioassay data, summarized in Section 6.2.11, the oral slope
factor was developed from inhalation bioassay data. The recommended oral slope factor is
4 x 10~2 per mg/kg-day, rounding to one significant digit, based on route-to-route extrapolation
of the combined male and female rat MCL data, using preferred dose metrics and modeling
approaches. This estimate is based on the most sensitive endpoint modeled using preferred dose
metrics, with estimates from other tumor sites using preferred dose metrics being lower by
between three- and 30-fold. The recommended oral slope factor is less than twofold higher than
estimates of total tumor risk from multiple sites using preferred dose metrics (brain, kidney,
testes, and MCL in the NTP rat bioassay), thereby providing support for the recommended oral
slope factor. Estimates using alternative dose metrics spanned a range from almost three orders
of magnitude below to almost fourfold above the recommended oral slope factor. Confidence in
the recommended oral slope factor is further increased by the concordance of the recommended
inhalation unit risk estimate (from which the oral slope factor was derived) with estimates based
on the available human data, discussed above. Although estimates based on human data are not
sufficient to serve as a primary basis for dose-response assessment, they support the plausibility
of the cancer risk estimates based on rodent bioassays.
6.2.11. Uncertainties in Cancer Dose-Response Assessment
A number of uncertainties underlie the cancer unit risk for tetrachloroethylene, as
discussed in the above sections, with Table 5-13 in Section 5 summarizing the impact on the
assessment of issues such as the use of models and extrapolation approaches, the effect of
reasonable alternatives, the decision concerning the preferred approach, and its justification.
These uncertainties have a varied impact on risk estimates. Some suggest risks could be higher
than was estimated, while others would decrease risk estimates or have an impact of an uncertain
direction. Several uncertainties are quantitatively characterized for the significantly increased
rodent tumors. These include the range of uncertainty in PBPK modeling and dose metrics and
the statistical uncertainty in the multistage modeling estimate. Due to limitations in the data,
particularly regarding the MOA and relative human sensitivity and variability, the quantitative
impact of other uncertainties of potentially equal or greater impact has not been explored. As a
result, an integrated quantitative analysis that considers all of these factors was not undertaken.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-21 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
6 3 OVERALL CHARACTERIZATION OF TETRACHLOROETHYLENE HAZARD
AND DOSE-RESPONSE
There is substantial potential for human exposure to tetrachloroethylene because 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
tetrachloroethylene or which have common metabolites or targets of toxicity. Once exposed,
humans, as well as laboratory animal species, rapidly absorb tetrachloroethylene, which is then
distributed to tissues via systemic circulation, metabolized, and then excreted primarily in breath
as unchanged tetrachloroethylene or CO2, or in urine as metabolites.
Based on the available human epidemiologic data and experimental and mechanistic
studies, it is concluded that tetrachloroethylene poses a potential human health hazard for
noncancer toxicity to the central nervous system, kidney, liver, immune and hematologic system,
and on development and reproduction. Neurotoxicity is identified as a sensitive endpoint
following either oral or inhalation exposure to tetrachloroethylene. Neurotoxic effects have been
characterized in human controlled exposure, occupational and residential studies, as well as in
experimental animal studies, providing evidence that tetrachloroethylene exposure results in
visual changes, increased reaction time, and decrements in cognition. Following EPA (2005a)
Guidelines for Carcinogen Risk Assessment, tetrachloroethylene is "Likely to be Carcinogenic to
Humans" by all routes of exposure. This characterization is based on suggestive evidence of
carcinogenicity in epidemiologic studies and conclusive evidence that the administration of
tetrachloroethylene, either by ingestion or by inhalation to sexually mature rats and mice,
increases tumor incidence (JISA. 1993; NCI. 1977; NTP. 1986b). In the rodent bioassays,
tetrachloroethylene increased the incidence of liver tumors (hepatocellular adenomas and
carcinomas) in male and female mice and of MCL in both sexes of rats. These findings were
reproducible in multiple lifetime bioassays employing different rodent strains and, in the case of
mouse liver tumors, by the inhalation and oral exposure routes. Additional tumor findings in rats
included significant increases in the NTP bioassay of testicular interstitial cell tumors and kidney
tumors in males, and brain gliomas in males and females. In mice, hemangiosarcomas in liver,
spleen, fat, and subcutaneous skin were reported in males in the JISA study. The epidemiologic
evidence provides a pattern associating tetrachloroethylene exposure and several types of cancer,
including bladder cancer, non-Hodgkin lymphoma and multiple myeloma. Associations and
exposure-response relationships were reported by studies using more precise
exposure-assessments for tetrachloroethylene. For other sites, including esophageal, kidney,
lung, cervical and breast cancer, more limited data supporting a suggestive effect are available.
As tetrachloroethylene toxicity and carcinogenicity are generally associated with
tetrachloroethylene metabolism, susceptibility to tetrachloroethylene health effects may be
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-22 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
modulated by factors affecting toxicokinetics, including lifestage, gender, genetic
polymorphisms, race/ethnicity, preexisting health status, lifestyle, and nutrition status. These
and other factors (e.g., socioeconomic status and multiple exposures) may contribute to variation
in response to tetrachloroethylene or its metabolites, once produced. In addition, it is not known
how tetrachloroethylene interacts with known risk factors for human diseases.
Dose-response analyses of the noncancer database focused on the neurotoxicity data set
as a basis for derivation of inhalation and oral reference values via the LOAEL/NOAEL
approach. The identified principal studies demonstrated color vision changes (Cavalleri et al..
1994), cognitive and reaction time changes (Echeverria et al.. 1995). and neurobehavioral
changes in cognitive performance tasks (Seeber. 1989). An RfC estimate of 0.04 mg/m3 is
supported by these multiple studies, as a midpoint of the range of available values, and is the
recommended RfC for tetrachloroethylene. Similarly, the recommended RfD estimate for
noncancer effects of 6 x 10~3 mg/kg-day was derived through route-to-route extrapolation of the
above inhalation studies. The RfD is equivalent to a drinking water concentration of 0.21 mg/L,
assuming a body weight of 70 kg and a daily water consumption of 2 L. There is high
confidence in these recommended noncancer reference values because they are supported by
moderate- to high-confidence estimates from multiple human neurotoxicity studies.
Additionally, quantitative dose-response analyses of the findings in other toxicity domains (i.e.,
kidney, liver, immunologic and hematologic, and reproductive and developmental toxicity),
detailed in Section 5 and summarized in Sections 6.2.5 and 6.2.7, are considered to be supportive
of these values.
For cancer, the recommended inhalation unit risk, defined as a plausible upper-bound
excess lifetime cancer risk estimated to result from continuous exposure to tetrachloroethylene,
is 7 x 10~2 per ppm, or 1 x 10~5 per jig/m3. This estimate is based on analysis of the combined
male and female rat MCL data from JISA (1993). using PBPK model-derived dose metrics and
dose-response modeling. The recommended oral slope factor, developed by PBPK
model-derived route-to-route extrapolation from the same data, is 6 x 10~2 per mg/kg-day. The
recommended inhalation unit risk and oral slope factor are based on the most sensitive endpoint
modeled using preferred dose metrics, with best estimates from other data sets (using preferred
dose metrics) being lower by three- to 50-fold. The recommended inhalation unit risk and oral
slope factor are less than threefold higher than estimates of total tumor risk from multiple sites
using preferred dose metrics (brain, kidney, testes, and MCL in the NTP rat bioassay and
hepatocellular tumors), thereby providing support for the recommended oral slope factor.
Estimates using alternative dose metrics spanned a range from three orders of magnitude below
to fourfold above the recommended values. Although estimates based on human data are not
sufficient to serve as a primary basis for dose-response assessment, comparisons between
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-23 DRAFT—DO NOT CITE OR QUOTE
-------
1 estimates from two human studies and the recommended values suggest that the cancer risk
2 estimates based on rodent bioassays are plausible.
3
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 6-24 DRAFT—DO NOT CITE OR QUOTE
-------
7. REFERENCES
Aanderud, L.; Ursin, R.; Larsen, M. (1982). Effect of high pressure on EEG burst suppression dose of
thiopental in rats. Undersea Biomed Res, 9, 255-261.
http: //www .ncbi .nlm .nih. gov/pubmed/7135635
Abalan, F. and Lacoste, G. (1985). An unfamiliar hypnotic, chloral hydrate Un hypnotique
meconnu: L'hydrate de chloral. 107/31 (2945-2946).
Abbas, R.; Seckel, C. S.; MacMahon, K. L.; Fisher, J. W. (1997). Determination of kinetic rate constants
for chloral hydrate, trichloroethanol, trichloroacetic acid and dichloroacetic acid: A
physiologically based modeling approach. Toxicologist, 36, 32-33.
Abbas, R. R.; Seckel, C. S.; Kidney, J. K.; Fisher, J. W. (1996). Pharmacokinetic analysis of chloral
hydrate and its metabolism in B6C3F1 mice. Drug Metab Dispos, 24, 1340-1346.
http: //www .ncbi .nlm .nih. gov/pubmed/8971140
Abbate, C.; Giorgianni, C.; Munao, F.; Brecciaroli, R. (1993). Neurotoxicity induced by exposure to
toluene: an electrophysiologic study. Int Arch Occup Environ Health, 64, 389-392.
Abdullah, A.; Walzman, M.; Wade, A. (1993). Treatment of external genital warts comparing cryotherapy
(liquid nitrogen) and trichloroacetic acid. Sex Transm Dis, 20, 344-345.
http: //www .ncbi .nlm .nih. gov/pubmed/8108758
Abe, T. and Wakui, C. (1984). [Necessity for total trichloride compound measurement in screening tests
for workers exposed to trichloroethylene or 1,1,1,-trichloroethane]. Sangyo Igaku, 26, 492-499.
http: //www .ncbi .nlm .nih. gov/pubmed/6242169
Abecia, E.; Martinez-Jarreta, B.; Pinilla, I.; Larrosa, M.; Castellano, M.; Honrubia, M. (1996). Bilateral
optic neuritis in occupational exposure to trichloroethylene. Med Lav, 87, 432-436.
http://www.ncbi.nlm.nih.gov/pubmed/9045Q31
Abel, M. (1985). Medication-related risks of CT-procedures in neonates and young infants
Pramedikation und risiken bei computertomographien im neugeboren. 25/12 (599-601).
Abel, M. (1987). [Respiratory failure in a newborn infant following repeated sedation for computerized
tomography], Klin Padiatr, 199, 52-54. http://dx.doi.org/10.1055/s-2008-102676Q
Abemayor, E.; Kovachich, G. B.; Haugaard, N. (1978). Regulation of pyruvate dehydrogenase (PDH)
reaction in rat brain by dichloroacetate (DCA) and ATP. 20/3 (No. 199).
Abraham, A. G.; Cox, C.; West, S. (2010). The differential effect of ultraviolet light exposure on cataract
rate across regions of the lens. Invest Ophthalmol Vis Sci, 51, 3919-3923.
http://dx.doi.org/10.1167/iovs.09-4557
Abraham, D.; Patel, P.; Cooper, A. (1995). Isolation from rat kidney of a cytosolic high molecular weight
cysteine-S-conjugate beta-lyase with activity toward leukotriene E4. J Biol Chem, 270, 180-188.
http: //www .ncbi .nlm .nih. gov/pubmed/7814371
Abrahams, V. (1971). Spino-spinal mechanisms in the chloralose anaesthetized cat. J Physiol, 215, 755-
768. http://www.ncbi.nlm.nih.gov/pubmed/5090993
Abrahams, V. and Langworth, E. (1967). The contribution of background electrical activity to the form of
averaged evoked potentials in chloralose anesthetized cats. Exp Neurol, 18, 253-266.
http: //www .ncbi .nlm .nih. gov/pubmed/6028145
Abrahamson, S. and Valencia, R. (1980). Evaluation of substances of interest for genetic damage using
Drosophila melanogaster Final sex-linked recessive lethal test report to FDA on 13 compounds.
DC, 233-277-2119.
Absalom, A. and Adapa, R. (2007). Anxiolytics, sedatives and hypnotics
Obstetrics/Pharmacology. 8/8 (340-344).
ACGIH. (American Conference of Governmental Industrial Hygienists). (1996). Trichloroethylene
Documentation of the threshold limit values and biological exposure indices (Sixth ed., Vol. Ill,
pp. ST1-ST6). Cincinnati, OH: Author.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-1 DRAFT—DO NOT CITE OR QUOTE
-------
ACGIH. (American Conference of Governmental Industrial Hygienists). (2001a). Dichloromethane.
Cincinnati, OH: Author.
ACGIH. (American Conference of Governmental Industrial Hygienists). (2001b). Halothane.
ACGIH. (American Conference of Governmental Industrial Hygienists). (2001c). Trichloroacetic Acid..
ACGIH. (American Conference of Governmental Industrial Hygienists). (200 Id). Trichloroethylene.
ACGIH. (2007). Trichloroethylene: TLV chemical substances 7th edition: American Conference of
Governmental Industrial Hygienists.
Acharya, D. and Dearlove, O. R. (2001). Anaesthesia in pyruvate dehydrogenase deficiency. 56/8 (808-
809).
Acher, R.; Chauvet, J.; Chauvet, M. (1973). Phylogeny of the neurohypophysial hormones. The active
peptides of a primitive fish, the sturgeon (Acipenser sp.). Eur J Biochem, 40, 585-589.
http: //www .ncbi .nlm .nih. gov/pubmed/4781390
Adgate, J. L.; Church, T. R.; Ryan, A. D.; Ramachandran, G.; Fredrickson, A. L.; Stock, T. H., . . .
Sexton, K. (2004). Outdoor, indoor, and personal exposure to VOCs in children. Environ Health
Perspect, 112, 1386-1392. http://dx.doi.org/10.1289/ehp.7107
Aggazzotti, G.; Fantuzzi, G.; Predieri, G.; Righi, E.; Moscardelli, S. (1994a). Indoor exposure to
perchloroethylene (PCE) in individuals living with dry-cleaning workers. Sci Total Environ, 156,
133-137. http://dx.doi.org/10.1016/0048-9697(94)90349-2
Aggazzotti, G.; Fantuzzi, G.; Righi, E.; Predieri, G.; Gobba, F. M.; Paltrinieri, M.; Cavalleri, A. (1994b).
Occupational and environmental exposure to perchloroethylene (PCE) in dry cleaners and their
family members. Arch Environ Occup Health, 49, 487-493.
http: //www .ncbi .nlm .nih. gov/pubmed/7818292
Ahlborg, G. (1990a). Validity of exposure data obtained by questionnaire. Two examples from
occupational reproductive studies. Scand J Work Environ Health, 16, 284-288.
http: //www .ncbi .nlm .nih. gov/pubmed/23 89136
Ahlborg, G., Jr. (1990b). Pregnancy outcome among women working in laundries and dry-cleaning
shops using tetrachloroethylene. Am J Ind Med, 17, 567-575.
http://dx.doi.org/10.1002/aiim.47001705Q3
Ahlborg, G. and Bodin, L. (1991). Tobacco smoke exposure and pregnancy outcome among working
women. A prospective study at prenatal care centers in Orebro County, Sweden. Am J Epidemiol,
133, 338-347. http://www.ncbi.nlm.nih.gov/pubmed/1994696
Akiyama, T. E.; Nicol, C. J.; Fievet, C.; Staels, B.; Ward, J. M.; Auwerx, J., . . . Peters, J. M. (2001).
Peroxisome proliferator-activated receptor-alpha regulates lipid homeostasis, but is not associated
with obesity: studies with congenic mouse lines. J Biol Chem, 276, 39088-39093.
http://dx.doi.org/10.1074/ibc.M10707320Q
Albee, R. E.; Spencer, P. J.; Johnson, K. A.; Bradley, G. H. (1994). 13-week trichloroethylene (TCE)
vapor exposure of rats caused by ototoxicity but no effects on the trigeminal nerve. Toxicologist,
14, 351.
Alberati-Giani, D.; Malherbe, P.; Kohler, C.; Lang, G.; Kiefer, V.; Lahm, H. W.; Cesura, A. M. (1995).
Cloning and characterization of a soluble kynurenine aminotransferase from rat brain: Identity
with kidney cysteine conjugate [beta]-lyase. JNeurochem, 64, 1448-1455.
http://dx.doi.Org/10.1046/i.1471-4159.1995.64041448.x
Albers, J. and Berent, S. (2000). Controversies in neurotoxicology: current status. Neurol Clin, 18, 741 -
764. http://www ncbi nlm nih gov/pubmed/10873241
Albers, J. W.; Wald, J. J.; Werner, R. A.; Franzblau, A.; Berent, S. (1999). Absence of polyneuropathy
among workers previously diagnosed with solvent-induced toxic encephalopathy. J Occup
Environ Med, 41, 500-509. http://www ncbi nlm nih gov/pubmed/10390702
Albertini, S. (1990). Analysis of nine known or suspected spindle poisons for mitotic chromosome
malsegregation using Saccharomyces cerevisiae D61.M. Mutagenesis, 5, 453-459.
http://www.ncbi.nlm.nih.gov/pubmed/2263203
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-2 DRAFT—DO NOT CITE OR QUOTE
-------
Allen, B. C.; Crump, K. S.; Shipp, A. M. (1988). Correlation between carcinogenic potency of chemicals
in animals and humans. Risk Anal, 8, 531-544. http://dx.doi.Org/10.l 11 l/i.1539-
6924.1988.tb01193.x
Altmann, L.; Bottger, A.; Wiegand, H. (1990). Neurophysiological and psychophysical measurements
reveal effects of acute low-level organic solvent exposure in humans. Int Arch Occup Environ
Health, 62, 493-499. http://dx.doi.org/10.1007/BF0Q381179
Altmann, L.; Neuhann, H. F.; Kramer, U.; Witten, J.; Jermann, E. (1995). Neurobehavioral and
neurophysiological outcome of chronic low-level tetrachloroethene exposure measured in
neighborhoods of dry cleaning shops. Environ Res, 69, 83-89.
http://dx.doi.org/10.1006/enrs.1995.1028
Altmann, L.; Wiegand, H.; Bottger, A.; Elstermeier, F.; Winneke, G. (1992). Neurobehavioural and
neurophysiological outcomes of acute repeated perchloroethylene exposure. Appl Psychol, 41,
269-279. http://dx.doi.org/10.1111/i. 1464-0597.1992.tb00705,x
American Academy of Pediatrics Committee on Drugs and Committee on Environmental Health: Use of
chloral hydrate for sedation in children. (1993). Pediatrics, 92, 471-473.
http://www.ncbi.nlm.nih.gov/pubmed/8361811
Amet, Y.; Berthou, F.; Fournier, G.; Dreano, Y.; Bardou, L.; Cledes, J.; Menez, J. F. (1997). Cytochrome
P450 4A and 2E1 expression in human kidney microsomes. Biochem Pharmacol, 53, 765-771.
http://dx.doi.org/10.1016/S0006-2952(96')00821-0
Anders, M. W.; Lash, L.; Dekant, W.; Elfarra, A. A.; Dohn, D. R.; Reed, D. J. (1988). Biosynthesis and
biotransformation of glutathione S-conjugates to toxic metabolites. Crit Rev Toxicol, 18, 311-
341. http://dx.doi.org/10.3109/1040844880903747Q
Andersen, A.; Barlow, L.; Engeland, A.; Kjaerheim, K.; Lynge, E.; Pukkala, E. (1999a). Work-related
cancer in the Nordic countries. Scand J Work Environ Health, 25, 1-116.
http: //www .ncbi .nlm .nih. gov/pubmed/10507118
Andersen, H.; Larsen, S.; Spliid, H.; Christensen, N. D. (1999b). Multivariate statistical analysis of organ
weights in toxicity studies. Toxicology, 136, 67-77. http://dx.doi.org/10.1016/S030Q-
483X(99)00056-6
Anderson, B. E.; Zeiger, E.; Shelby, M. D.; Resnick, M. A.; Gulati, D. K.; Ivett, J. L.; Loveday, K. S.
(1990). Chromosome aberration and sister chromatid exchange test results with 42 chemicals.
Environ Mol Mutagen, 16, 55-137. http://dx.doi.org/10.1002/em.28501605Q5
Anderson, S.; Dunn, C.; Cattley, R.; Corton, J. (2001). Hepatocellular proliferation in response to a
peroxisome proliferator does not require TNFalpha signaling. Carcinogenesis, 22, 1843-1851.
http: //www .ncbi .nlm .nih. gov/pubmed/11698348
Anderson, W. B.; Board, P. G.; Gargano, B.; Anders, M. W. (1999). Inactivation of glutathione
transferase zeta by dichloroacetic acid and other fluorine-lacking [alpha]-haloalkanoic acids.
Chem Res Toxicol, 12, 1144-1149. http://dx.doi.org/10.1021/tx9900851
Andrews, J.; Nichols, H.; Schmid, J.; Mole, L.; Hunter, E.; Klinefelter, G. (2004). Developmental toxicity
of mixtures: the water disinfection by-products dichloro-, dibromo- and bromochloro acetic acid
in rat embryo culture. Reprod Toxicol, 19, 111-116.
http://dx.doi.Org/10.1016/i.reprotox.2004.06.005
Andrys, C.; Hanovcova, I.; Chylkova, V.; Tejral, J.; Eminger, S.; Prochazkova, J. (1997). Immunological
monitoring of dry-cleaning shop workers - exposure to tetrachloroethylene. Cent Eur J Public
Health, 5, 136-142. http://www.ncbi.nlm.nih.gov/pubmed/9386901
Anger, W. K.; Liang, Y. X.; Nell, V.; Kang, S. K.; Cole, D.; Bazylewicz-Walczak, B., . . . Sizemore, O. J.
(2000). Lessons learned - 15 years of the WHO-NCTB: A review. Neurotoxicology, 21, 837-846.
http://www.ncbi.nlm.nih.gOv/pubmed/l 1130289
Anthony, M. (1983). Drugs in migraine. 24/8 (89-113).
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.
http: //www .ncbi .nlm .nih. gov/pubmed/7552463
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-3 DRAFT—DO NOT CITE OR QUOTE
-------
Anttila, S.; Hirvonen, A.; Husgafvel-Pursiainen, K.; Kaijalainen, A.; Nurminen, T.; Vainio, H. (1994).
Combined effect of CYP1A1 inducibility and GSTM1 polymorphism on histological type of lung
cancer. Carcinogenesis, 15, 1133-1135. http://dx.doi.Org/10.1093/carcin/15.6.1133
Appendix B~Reference method for the determination of suspended particulates in the atmosphere (high-
volume method), 40, C.F.R. § 50 (1977).
Aranyi, C.; O'Shea, W. J.; Graham, J. A.; Miller, F. J. (1986). The effects of inhalation of organic
chemical air contaminants on murine lung host defenses. Fundam Appl Toxicol, 6, 713-720.
http://dx.doi.org/10.1016/0272-0590(86)90184-3
Aronson, K.; Siemiatycki, J.; Dewar, R.; Gerin, M. (1996). Occupational risk factors for prostate cancer:
Results from a case-control study in Montreal, Quebec, Canada. Am J Epidemiol, 143, 363-373.
http://www.ncbi.nlm.nih.gov/pubmed/8633620
Aryal, B. K.; Khuder, S. A.; Schaub, E. A. (2001). Meta-analysis of systemic sclerosis and exposure to
solvents. Am J Ind Med, 40, 271-274. http://dx.doi.org/10.10Q2/aiim.1098
Asal, N. R.; Geyer, J. R.; Risser, D. R.; Lee, E. T.; Kadamani, S.; Cherng, N. (1988). Risk factors in renal
cell carcinoma. II. Medical history, occupation, multivariate analysis, and conclusions. Cancer
Detect Prev, 13, 263-279. http://www.ncbi.nlm.nih.gov/pubmed/3266567
Aschengrau, A. and 3rd, S. G. (2003). Essentials of Epidemiology in Public Health. Sudbury, MA: Jones
& Bartlett Publishers.
Aschengrau, A.; Ozonoff, D.; Paulu, C.; Coogan, P.; Vezina, R.; Heeren, T.; Zhang, Y. (1993). Cancer
risk and tetrachloroethylene-contaminated drinking water in Massachusetts. Arch Environ Health,
48, 284-292. http://www.ncbi.nlm.nih.gov/pubmed/8215591
Aschengrau, A.; Paulu, C.; Ozonoff, D. (1998). Tetrachloroethylene-contaminated drinking water and the
risk of breast cancer. Environ Health Perspect, 106, 947-953.
http://www.ncbi.nlm.nih.gov/pubmed/9703477
Aschengrau, A.; Rogers, S.; Ozonoff, D. (2003). Perchloroethylene-contaminated drinking water and the
risk of breast cancer: Additional results from Cape Cod, Massachusetts, USA. Environ Health
Perspect, 111, 167-173. http://www.ncbi.nlm.nih.gov/pubmed/12573900
Aschengrau, A.; Weinberg, J.; Rogers, S.; Gallagher, L.; Winter, M.; Vieira, V., . . . Ozonoff, D. (2008).
Prenatal exposure to tetrachloroethylene-contaminated drinking water and the risk of adverse
birth outcomes. Environ Health Perspect, 116, 814-820. http://dx.doi.org/10.1289/ehp.10414
Aschengrau, A.; Weinberg, J. M.; Gallagher, L. G.; Winter, M. R.; Vieira, V. M.; Webster, T. F.;
Ozonoff, D. M. (2009a). Exposure to Tetrachloroethylene-Contaminated Drinking Water and the
Risk of Pregnancy Loss. 1, 23-34. http://dx.doi.org/10.1007/sl2403-009-00Q3-x
Aschengrau, A.; Weinberg, J. M.; Janulewicz, P. A.; Gallagher, L. G.; Winter, M. R.; Vieira, V. M., . . .
Ozonoff, D. M. (2009b). Prenatal exposure to tetrachloroethylene-contaminated drinking water
and the risk of congenital anomalies: A retrospective cohort study. Environ Health, 8, 44.
http://dx.doi.org/10.1186/1476-069X-8-44
Ashby, J.; Brady, A.; Elcombe, C. R.; Elliott, B. M.; Ishmael, J.; Odum, J., . . . Purchase, I. F. (1994).
Mechanistically-based human hazard assessment of peroxisome proliferator-induced
hepatocarcinogenesis. Hum Exp Toxicol, 13, S1-S117.
http://dx.doi.org/10.1177/09603271940130Q201
Atkinson, A.; Meeks, R. G.; Roy, D. (1993). Increased oxidative stress in the liver of mice treated with
trichloroethylene. 31, 297-304.
ATSDR. (Agency for Toxic Substances and Disease Registry). (1997a). Toxicological profile for
tetrachloroethylene. Atlanta, GA: U.S. Department of Health and Humans Services. Retrieved
from http://www.atsdr.cdc.gov/toxprofiles/tpl8.html.
ATSDR. (Agency for Toxic Substances and Disease Registry). (1997b). Toxicological profile for
trichloroethylene (update). (PB98-101165). Atlanta, GA: U.S. Department of Health and Human
Services, Public Health Service, Agency for Toxic Substances and Disease Registry. Retrieved
from http://www.atsdr.cdc.gov/toxprofiles/tp 19-p.pdf.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-4 DRAFT—DO NOT CITE OR QUOTE
-------
ATSDR. (Agency for Toxic Substances and Disease Registry). (1998a). Biomarkers of kidney function
for environmental health field studies. (PB98-142128; ATSDRHS9896). Atlanta, GA: Author.
Retrieved from http://www.ntis.gov/search/product.aspx?ABBR=PB98142128.
ATSDR. (Agency for Toxic Substances and Disease Registry). (1998b). Volatile organic compounds in
drinking water and adverse pregnancy outcomes. United States Marine Corps Base, Camp
Lejeune, NC. (PB98-156540; ATSDRHS9897). Atlanta, GA: Author. Retrieved from
http: //www .atsdr. cdc. gov/hs/lei eune/.
ATSDR. (Agency for Toxic Substances and Disease Registry). (2003). 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. Atlanta, GA: Author. Retrieved
from http://www.atsdr.cdc.gov/sites/leieune/survev full.html.
ATSDR. (Agency for Toxic Substances and Disease Registry). (2006). Toxicological profile for 1,1,1
trichloroethane. Atlanta, GA: Author. Retrieved from http://www.atsdr.cdc.gov/toxpro2.html.
ATSDR. (Agency for Toxic Substances and Disease Registry). (2010). Toxicological profiles:
Information about contaminants found at hazardous waste sites, from
http: //www .atsdr. cdc. gov/toxprofile s/index.asp
Auperin, A.; Benhamou, S.; Ory-Paoletti, C.; Flamant, R. (1994). Occupational risk factors for renal cell
carcinoma: A case-control study. Occup Environ Med, 51, 426-428.
http://dx.doi.org/10.1136/oem.51.6.426
Austin, H.; Delzell, E.; Grufferman, S.; Levine, R.; Morrison, A. S.; Stolley, P. D.; Cole, P. (1987). Case-
control study of hepatocellular carcinoma, occupation, and chemical exposures. J Occup Med, 29,
665-669. http://www.ncbi.nlm.nih.gov/pubmed/2821204
Bachir, S.; Ambrosio, M.; Federici, V.; Barnier, H. (1998). Hydrothermal oxidation of organochlorines.
Analusis, 26, 237-244.
Baelum, J. (1999). Acute symptoms during non-inhalation exposure to combinations of toluene,
trichloroethylene, and n-hexane. Int Arch Occup Environ Health, 72, 408-410.
http://www.ncbi.nlm.nih.gov/pubmed/10473841
Bagnell, P. C. and Ellenberger, H. A. (1977). Obstructive jaundice due to a chlorinated hydrocarbon in
breast milk. Can Med Assoc J, 117, 1047-1048.
Bakke, B.; Stewart, P.; Waters, M. (2007). Uses of and exposure to trichloroethylene in U.S. industry: a
systematic literature review. J Occup Environ Hyg, 4, 375-390.
http://dx.doi.org/10.1080/154596207013Q1763
Bale, A. S.; Adams, T. L.; Bushnell, P. J.; Shafer, T. J.; Boyes, W. K. (2005). Role of NMDA, nicotinic,
and GABA receptors in the steady-state visual-evoked potential in rats. Pharmacol Biochem
Behav, 82, 635-645. http://dx.doi.Org/10.1016/i.pbb.2005.ll.003
Band, P.; Le, N.; Fang, R.; Threlfall, W.; Gallagher, R. (1999). Identification of occupational cancer risks
in British Columbia. Part II: A population-based case-control study of 1516 Prostatic cancer
cases. J Occup Environ Med, 41, 233-247. http://www.ncbi.nlm.nih.gov/pubmed/10224589
Band, P. R.; Le, N. D.; Fang, R.; Deschamps, M.; Gallagher, R. P.; Yang, P. (2000). Identification of
occupational cancer risks in British Columbia: A population-based case-control study of 995
incident breast cancer cases by menopausal status, controlling for confounding factors. J Occup
Environ Med, 42, 284-310. http://www.ncbi.nlm.nih.gov/pubmed/10738708
Bartels, M. J. (1994). Quantitation of the tetrachloroethylene metabolite N-acetyl-S-
(trichlorovinyl)cysteine in rat urine via negative ion chemical ionization gas
chromatography/tandem mass spectrometry. Biological Mass Spectrometry, 23, 689-694.
http://dx.doi.org/10.1002/bms. 1200231107
Bartonicek, V. and Teisinger, J. (1962). Effect of tetraethyl thiuram disulphide (disulfiram) on
metabolism of trichloroethylene in man. Occup Environ Med, 19, 216-221.
http://dx.doi.Org/10.l 136/oem. 19.3.216
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-5 DRAFT—DO NOT CITE OR QUOTE
-------
Bartsch, H.; Malaveille, C.; Barbin, A.; Planche, G. (1979). Mutagenic and alkylating metabolites of halo-
ethylenes, chlorobutadienes and dichlorobutenes produced by rodent or human liver tissues:
Evidence for oxirane formation by P450-linked microsomal mono-oxygenases. Arch Toxicol, 41,
249-277. http://dx.doi.org/10.1007/BF0Q296896
Basch, E.; Foppa, I.; Liebowitz, R.; Nelson, J.; Smith, M.; Sollars, D.; Ulbricht, C. (2004). Lavender
(Lavandula angustifolia Miller). 4/2 (63-78).
Baylin, S. B.; Herman, J. G.; Graff, J. R.; Vertino, P. M.; Issa, J. P. (1998). Alterations in DNA
methylation: a fundamental aspect of neoplasia. Adv Cancer Res, 72, 141-196.
http://dx.doi.org/10.1016/S0065-230X(08)60702-2
Beaudreuil, S.; Lasfargues, G.; Laueriere, L.; El Ghoul, Z.; Fourquet, F.; Longuet, C., . . . Buchler, M.
(2005). Occupational exposure in ANCA-positive patients: A case-control study. Kidney Int, 67,
1961-1966. http://dx.doi.Org/10.llll/i.1523-1755.2005.00295.x
Beckstead, M. J.; Weiner, J. L.; Eger, E. I.; Gong, D. H.; Mihic, S. J. (2000). Glycine and [gamma]-
aminobutyric acid(A) receptor function is enhanced by inhaled drugs of abuse. Mol Pharmacol,
57, 1199-1205. http://www.ncbi.nlm.nih.gov/pubmed/10825391
Beliles, R. P. (2002). Concordance across species in the reproductive and developmental toxicity of
tetrachloroethylene. Toxicol Ind Health, 18, 91-106.
http://dx.doi.org/10.1191/0748233702thl37oa
Bellinger, D.; Hu, H.; Titlebaum, L.; Needleman, H. L. (1994). Attentional correlates of dentin and bone
lead levels in adolescents. Arch Environ Health, 49, 98-105.
http: //www .ncbi .nlm .nih. gov/pubmed/8161248
Benane, S.; Blackman, C.; House, D. (1996). Effect of perchloroethylene and its metabolites on
intercellular communication in clone 9 rat liver cells. J Toxicol Environ Health, 48, 427-437.
http: //www .ncbi .nlm .nih. gov/pubmed/8751833
Bergamaschi, E.; Mutti, A.; Bocchi, M. C.; Alinovi, R.; Olivetti, G.; Ghiggeri, G. M.; Franchini, I.
(1992). Rat model of perchloroethylene-induced renal dysfunctions. Environ Res, 59, 427-439.
http://dx.doi.org/10.1016/S0013-9351(05)80046-5
Berger, T. and Horner, C. M. (2003). In vivo exposure of female rats to toxicants may affect oocyte
quality. Reprod Toxicol, 17, 273-281. http://dx.doi.org/10.1016/S0890-6238(03)00009-l
Berman, E.; Schlicht, M.; Moser, V. C.; MacPhail, R. C. (1995). A multidisciplinary approach to
toxicological screening: I. Systemic toxicity. J Toxicol Environ Health, 45, 127-143.
http://dx.doi.org/10.1080/15287399509531986
Bernard, A. and Lauwerys, R. (1995). Low-molecular-weight proteins as markers of organ toxicity with
special reference to Clara cell protein. Toxicol Lett, 77, 145-151. http://dx.doi.org/10.1016/Q378-
4274(95)03284-3
Bernstine, M. L. (1954). Cardiac arrest occurring under trichloroethylene analgesia: report of a case with
recovery. 68, 262-266.
Beyer, R. P.; Fry, R. C.; Lasarev, M. R.; McConnachie, L. A.; Meira, L. B.; Palmer, V. S., . . .
Consortium, M. o. t. T. R. (2007). Multicenter study of acetaminophen hepatotoxicity reveals the
importance of biological endpoints in genomic analyses. Toxicol Sci, 99, 326-337.
http: //dx. doi. org/10.1093/toxsci/kfm 150
Bhargava, K.; Srivastava, V.; Barthwal, J.; Sinha, J. (1976). Significance of spinal autonomic loci in post
coronary ligation arrhythmia. Neuropharmacology, 15, 625-633.
http: //www .ncbi .nlm .nih. gov/pubmed/995232
Bhattacharya, R. K. and Schultze, M. O. (1971). Properties of DNA isolated from tissues of calves treated
with S-(l,2-dichlorovinyl)-L-cysteine: II. Primer-template activity for bacterial DNA
polymerases. Arch Biochem Biophys, 145, 575-582. http://dx.doi.org/10.1016/S00Q3-
9861(71)80017-6
Bhattacharya, R. K. and Schultze, M. O. (1972). Properties of DNA treated with S-(l,2-dichlorovinyl)-L-
cysteine and a lyase. Arch Biochem Biophys, 153, 105-115. http://dx.doi.org/10.1016/00Q3-
9861(72)90426-2
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-6 DRAFT—DO NOT CITE OR QUOTE
-------
Bhunya, S. P. and Behera, B. C. (1987). Relative genotoxicity of trichloroacetic acid (TCA) as revealed
by different cytogenetic assays: Bone marrow chromosome aberration, micronucleus and sperm-
head abnormality in the mouse. Mutat Res Genet Toxicol Environ Mutagen, 188, 215-221.
http://dx.doi.org/10.1016/0165-1218(87)90092-9
Bhunya, S. P. and Jena, G. B. (1996). The evaluation of clastogenic potential of trichloroacetic acid
(TCA) in chick in vivo test system. Mutat Res Genet Toxicol Environ Mutagen, 367, 253-259.
http://dx.doi.org/10.1016/S0165-1218(96)90085-3
Bichsel, P.; Oliver, J.; Coulter, D.; Brown, J. (1988). Recording of visual-evoked potentials in dogs with
scalp electrodes. J Vet Intern Med, 2, 145-149. http://www.ncbi.nlm.nih.gov/pubmed/3225809
BIP Study Group. (Bezafibrate Infarction Prevention Study Group). (2000). Secondary prevention by
raising HDL cholesterol and reducing triglycerides in patients with coronary artery disease: The
Bezafibrate Infarction Prevention (BIP) study. Circulation, 102, 21-27.
http: //www .ncbi .nlm .nih. gov/pubmed/10880410
Birner, G.; Richling, C.; Henschler, D.; Anders, M. W.; Dekant, W. (1994). Metabolism of
tetrachloroethene in rats: Identification ofN[epsilon]-(dichloroacetyl)-L-lysine and N[epsilon]-
(trichloroacetyl)-L-lysine as protein adducts. Chem Res Toxicol, 7, 724-732.
http://dx.doi.org/10.1021/txQ0042a003
Birner, G.; Rutkowska, A.; Dekant, W. (1996). N-acetyl-S-(l,2,2-trichlorovinyl)-L-cysteine and 2,2,2-
trichloroethanol: Two novel metabolites of tetrachloroethene in humans after occupational
exposure. Drug Metab Dispos, 24, 41-48.
Bishop, B. and Laverty, R. (1989). Dose-dependent reduction by Ro 15-4513 in mice of the effects of
ethanol and some other general depressant drugs. Eur J Pharmacol, 162, 265-271.
http: //www .ncbi .nlm .nih. gov/pubmed/2498108
Bitter, I.; Dossenbach, M. R.; Brook, S.; Feldman, P. D.; Metcalfe, S.; Gagiano, C. A., . . . Group, O. H.
S. (2004). Olanzapine versus clozapine in treatment-resistant or treatment-intolerant
schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry, 28, 173-180.
http://dx.doi.Org/10.1016/i.pnpbp.2003.09.033
Blair, A.; Decoufle, P.; Grauman, D. (1979). Causes of death among laundry and dry cleaning workers.
Am J Public Health, 69, 508-511. http://www.ncbi.nlm.nih.gov/pubmed/434285
Blair, A.; Hartge, P.; Stewart, P. A.; 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.
http://dx.doi.Org/10.l 136/oem. 55.3.161
Blair, A.; Linos, A.; Stewart, P. A.; Burmeister, L. F.; Gibson, R.; Everett, G., . . . Cantor, K. P. (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. http://dx.doi.org/10.1002/aiim.47002302Q7
Blair, A.; Petralia, S. A.; Stewart, P. A. (2003). Extended mortality follow-up of a cohort of dry cleaners.
Ann Epidemiol, 13, 50-56. http://dx.doi.org/10.1016/S1047-2797(02)00250-8
Blair, A.; Stewart, P. A.; Tolbert, P. E.; Grauman, D.; Moran, F. X.; Vaught, J.; Rayner, J. (1990). Cancer
and other causes of death among a cohort of dry cleaners. Br J Ind Med, 47, 162-168.
http://dx.doi.Org/10.l 136/oem.47.3.162
Blot, W. J. and McLaughlin, J. K. (1999). The changing epidemiology of esophageal cancer. Semin
Oncol, 26, 2-8. http://www ncbi nlm nih gov/pubmed/10566604
Bliimcke, S.; Schwartzkopff, W.; Lobeck, H.; Edmondson, N. A.; Prentice, D. E.; Blane, G. F. (1983).
Influence of fenofibrate on cellular and subcellular liver structure in hyperlipidemic patients.
Atherosclerosis, 46, 105-116. http://dx.doi.org/10.1016/0021-9150(83)90169-7
Blunden, S. (1981). Drugs and the soldier. J R Army Med Corps, 127, 72-77.
http: //www .ncbi .nlm .nih. gov/pubmed/7252908
Board, P. G.; Baker, R. T.; Chelvanayagam, G.; Jermiin, L. S. (1997). Zeta, a novel class of glutathione
transferases in a range of species from plants to humans. Biochem J, 328, 929-935.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-7 DRAFT—DO NOT CITE OR QUOTE
-------
Boettner, E. A. and Muranko, H. J. (1969). Animal breath data for estimating the exposure of humans to
chlorinated hydrocarbons. Am Ind Hyg Assoc J, 30, 437-442.
http: //www .ncbi .nlm .nih. gov/pubmed/5 823423
Bogen, K. T.; BW Jr, C.; Machicao, L. K. (1992). Dermal absorption of dilute aqueous chloroform,
trichloroethylene, and tetrachloroethylene in hairless guinea pigs. Toxicol Sci, 18, 30-39.
http://dx.doi.org/10.1016/0272-0590(92)90192-K
Boice, J.; Marano, D.; Fryzek, J.; Sadler, C.; McLaughlin, J. (1999). Mortality among aircraft
manufacturing workers. Occup Environ Med, 56, 581-597.
http: //www .ncbi .nlm .nih. gov/pubmed/10615290
Bois, F. Y.; Gelman, A.; Jiang, J.; Maszle, D. R.; Zeise, L.; Alexeef, G. (1996). Population toxicokinetics
of tetrachloroethylene. Arch Toxicol, 70, 347-355. http://dx.doi.org/10.1007/s00204005Q284
Bois, F. Y.; Zeise, L.; Tozer, T. N. (1990). Precision and sensitivity of pharmacokinetic models for cancer
risk assessment: Tetrachloroethylene in mice, rats, and humans. Toxicol Appl Pharmacol, 102,
300-315. http://dx.doi.org/10.1016/0041-008X(90')90029-T
Bolanowska, W. and Golacka, J. (1972). [Inhalation and excretion of tetrachloroethylene in men in
experimental conditions]. Med Pr, 23, 109-119.
Bolt, H. M.; Laib, R. J.; Filser, J. G. (1982). Reactive metabolites and carcinogenicity of halogenated
ethylenes. Biochem Pharmacol, 31, 1-4. http://dx.doi.org/10.1016/0006-2952(82)90227-1
Bonatti, S.; Cavalieri, Z.; Viaggi, S.; Abbondandolo, A. (1992). The analysis of 10 potential spindle
poisons for their ability to induce CREST-positive micronuclei in human diploid fibroblasts.
Mutagenesis, 7, 111-114. http ://www.ncbi .nlm .nih.gov/pubmed/15 79065
Bond, G.; McLaren, E.; Cartmill, J.; Wymer, K.; Sobel, W.; Lipps, T.; Cook, R. (1987). Cause-specific
mortality among male chemical workers. Am J Ind Med, 12, 353-383.
http: //www .ncbi .nlm .nih. gov/pubmed/3 674026
Bond, G.; McLaren, E.; Sabel, F.; Bodner, K.; Lipps, T.; Cook, R. (1990). Liver and biliary tract cancer
among chemical workers. Am J Ind Med, 18, 19-24.
http: //www .ncbi .nlm .nih. gov/pubmed/2378367
Bonse, G. and Henschler, D. (1976). Chemical reactivity, biotransformation, and toxicity of
polychlorinated aliphatic compounds. Crit Rev Toxicol, 4, 395-409.
http: //www .ncbi .nlm .nih. gov/pubmed/791581
Bonse, G.; Urban, T.; Reichert, D.; Henschler, D. (1975). Chemical reactivity, metabolic oxirane
formation and biological reactivity of chlorinated ethylenes in the isolated perfused rat liver
preparation. Biochem Pharmacol, 24, 1829-1834. http://dx.doi.org/10.1016/0006-2952(75)90468-
2
Borzio, M.; Fargion, S.; Borzio, F.; Fracanzani, A. L.; Croce, A. M.; Stroffolini, T., . . . Roncalli, M.
(2003). Impact of large regenerative, low grade and high grade dysplastic nodules in
hepatocellular carcinoma development. J Hepatol, 39, 208-214.
http: //www .ncbi .nlm .nih. gov/pubmed/12873817
Bosco, M. G.; Figa-Talamanca, I.; Salerno, S. (1987). Health and reproductive status of female workers in
dry cleaning shops. Int Arch Occup Environ Health, 59, 295-301.
http://dx.doi.org/10.1007/BF0Q377741
Bosgra, S.; Mennes, W.; Seinen, W. (2005). Proceedings in uncovering the mechanism behind
peroxisome proliferator-induced hepatocarcinogenesis. Toxicology, 206, 309-323.
http://dx.doi.Org/10.1016/i.tox.2004.07.015
Boulet, L. P. (1988). Increases in airway responsiveness following acute exposure to respiratory irritants:
Reactive airway dysfunction syndrome or occupational asthma? Chest, 94, 476-481.
http://dx.doi.Org/10.1378/chest.94.3.476
Bove, F.; Shim, Y.; Zeitz, P. (2002). Drinking water contaminants and adverse pregnancy outcomes: A
review. Environ Health Perspect, 110, 61-74. http://ehpnet 1 .niehs.nih.gov/docs/2002/suppl-1/61-
74bove/abstract.html
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-8 DRAFT—DO NOT CITE OR QUOTE
-------
Bove, F. J.; Fulcomer, M. C.; Klotz, J. B.; Esmart, J.; Dufficy, E. M.; Savrin, J. E. (1995). Public drinking
water contamination and birth outcomes. Am J Epidemiol, 141, 850-862.
Bowen, S. E.; Batis, J. C.; Paez-Martinez, N.; Cruz, S. L. (2006). The last decade of solvent research in
animal models of abuse: Mechanistic and behavioral studies. Neurotoxicol Teratol, 28, 636-647.
http://dx.doi.Org/10.1016/i.ntt.2006.09.005
Bowler, R. M.; Mergler, D.; Huel, G.; Harrison, R.; Cone, J. (1991). Neuropsychological impairment
among former microelectronics workers. Neurotoxicology, 12, 87-103.
http://www.ncbi.nlm.nih.gov/pubmed/2014Q71
Bowman, K. J. (1982). A method for quantitative scoring of the Farnsworth Panel D-15. Acta
Ophthalmol, 60, 907-916. http://dx.doi.Org/10.llll/i.1755-3768.1982.tb00621.x
Box, G. E. P.; Hunter, W. G.; Hunter, J. S. (1978). Statistics for experimenters: an introduction to design,
data analysis, and model building. New York, NY: Wiley.
Boyes, W. K.; Bercegeay, M.; Oshiro, W. M.; Krantz, Q. T.; Kenyon, E. M.; Bushnell, P. J.; Benignus, V.
A. (2009). Acute perchloroethylene exposure alters rat visual-evoked potentials in relation to
brain concentrations. Toxicol Sci, 108, 159-172. http://dx.doi.org/10.1093/toxsci/kfh265
Bozzelli, J. W. and Kebbekus, B. B. (1983). Volatile organic compounds in the ambient atmosphere of
the New Jersey, New York area.
Bradley, W. (1984). Treatable peripheral neuropathies. 147, 14-21.
http: //www .ncbi .nlm .nih. gov/pubmed/6328776
Braissant, O. and Wahli, W. (1998). Differential expression of peroxisome proliferator-activated receptor-
alpha, -beta, and -gamma during rat embryonic development. Endocrinology, 139, 2748-2754.
http: //www .ncbi .nlm .nih. gov/pubmed/9607781
Brambilla, G. and Martelli, A. (2004). Failure of the standard battery of short-term tests in detecting some
rodent and human genotoxic carcinogens. Toxicology, 196, 1-19.
http://dx.doi.Org/10.1016/i.tox.2003.l 1.003
Branda, M. and Wands, J. R. (2006). Signal transduction cascades and hepatitis B and C related
hepatocellular carcinoma. Hepatology, 43, 891-902. http://dx.doi.org/10.1002/hep.21196
Breckenridge, A. (1971). Pathophysiological factors influencing drug kinetics. Acta Pharmacol Toxicol,
29 Suppl 3, 225-232. http://www.ncbi.nlm.nih.gov/pubmed/5316404
Bredeloux, P.; Dubuc, I.; Costentin, J. (2007). Comparisons between bupropion and dexamphetamine in a
range of in vivo tests exploring dopaminergic transmission. Br J Pharmacol, 150, 711-719.
http://dx.doi.org/10.1038/si.bip.0707151
Brentnall, T. A.; Crispin, D. A.; Rabinovitch, P. S.; Haggitt, R. C.; Rubin, C. E.; Stevens, A. C.; Burmer,
G. C. (1994). Mutations in the p53 gene: an early marker of neoplastic progression in ulcerative
colitis. Gastroenterology, 107, 369-378. http://www.ncbi.nlm.nih.gov/pubmed/8039614
Breslow, N. E. and Day, N. E. (1994). Statistical Method in Cancer Research. Volume 2 - The Design and
Analysis of Cohort Studies (IARC Scientific Publications No. 82 ed. Vol. 2). New York, NY:
Oxford University Press.
Bringmann, G.; God, R.; Feineis, D.; Wesemann, W.; Riederer, P.; Rausch, W., . . . Sontag, K. (1995).
The TaClo concept: l-trichloromethyl-l,2,3,4-tetrahydro-beta-carboline (TaClo), a new toxin for
dopaminergic neurons. J Neural Transm Suppl, 46, 235-244.
http: //www .ncbi .nlm .nih. gov/pubmed/8 821060
Briving, C.; Jacobson, I.; Hamberger, A.; Kjellstrand, P.; Haglid, K. G.; Rosengren, L. E. (1986). Chronic
effects of perchloroethylene and trichloroethylene on the gerbil brain amino acids and
glutathione. Neurotoxicology, 7, 101-108. http://www.ncbi.nlm.nih.gov/pubmed/2872637
Broadwell, D. K.; Darcey, D. J.; Hudnell, H. K.; Otto, D. A.; Boyes, W. K. (1995). Work-site clinical and
neurobehavioral assessment of solvent-exposed microelectronics workers. Am J Ind Med, 27,
677-698. http://www.ncbi.nlm.nih.gov/pubmed/7611305
Brodkin, C. A.; Daniell, W.; Checkoway, H.; Echeverria, D.; Johnson, J.; Wang, K., . . . Gretch, D.
(1995). Hepatic ultrasonic changes in workers exposed to perchloroethylene. Occup Environ
Med, 52, 679-685.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-9 DRAFT—DO NOT CITE OR QUOTE
-------
Bronzetti, G.; Bauer, C.; Corsi, C.; Del Carratore, R.; Galli, A.; Nieri, R.; Paolini, M. (1983). Genetic and
biochemical studies on perchloroethylene 'in vitro' and 'in vivo'. Mutat Res Genet Toxicol
Environ Mutagen, 116, 323-331. http://dx.doi.org/10.1016/0165-1218(83)90070-8
Bronzetti, G.; Galli, A.; Corsi, C.; Cundari, E.; Del Carratore, R.; Nieri, R.; Paolini, M. (1984). Genetic
and biochemical investigation on chloral hydrate in vitro and in vivo. Mutat Res, 141, 19-22.
http://www.ncbi.nlm.nih.gov/pubmed/6384771
Brown, L. M.; Hoover, R.; Silverman, D.; Baris, D.; Hayes, R.; Swanson, G. M., . . . Fraumeni JF, J.
(2001). Excess incidence of squamous cell esophageal cancer among US Black men: role of
social class and other risk factors. Am J Epidemiol, 153, 114-122.
Brown Dzubow, R.; Makris, S.; Siegel Scott, C.; Barone, S. J. (2010). Early lifestage exposure and
potential developmental susceptibility to tetrachloroethylene. Birth Defects Res B Dev Reprod
Toxicol, 89, 50-65. http://dx.doi.org/10.1002/bdrb.20222
Brownson, R. C.; Alavanja, M. C.; Chang, J. C. (1993). Occupational risk factors for lung cancer among
nonsmoking women: a case-control study in Missouri (United States). Cancer Causes Control, 4,
449-454. http://www.ncbi.nlm.nih.gov/pubmed/8218877
Buben, J. A. and O'Flaherty, E. J. (1985). Delineation of the role of metabolism in the hepatotoxicity of
trichloroethylene and perchloroethylene: A dose-effect study. Toxicol Appl Pharmacol, 78, 105-
122. http://www.ncbi.nlm.nih.gov/pubmed/2994252
Buelke-Sam, J.; Kimmel, C. A.; Adams, J.; Nelson, C. J.; Vorhees, C. V.; Wright, D. C., . . . Geyer, M. A.
(1985). Collaborative Behavioral Teratology Study: results. Neurobehav Toxicol Teratol, 7, 591-
624.
Bull, R. J.; Orner, G. A.; Cheng, R. S.; Stillwell, L.; Stauber, A. J.; Sasser, L. B., . . . Thrall, B. D. (2002).
Contribution of dichloroacetate and trichloroacetate to liver tumor induction in mice by
trichloroethylene. Toxicol Appl Pharmacol, 182, 55-65. http://dx.doi.org/10.1006/taap.20Q2.9427
Bull, R. J.; Sanchez, I. M.; Nelson, M. A.; Larson, J. L.; Lansing, A. J. (1990). Liver tumor induction in
B6C3F1 mice by dichloroacetate and trichloroacetate. Toxicology, 63, 341-359.
http://dx.doi.org/10.1016/0300-483X(90)90195-M
Bull, R. J.; Sasser, L. B.; Lei, X. C. (2004). Interactions in the tumor-promoting activity of carbon
tetrachloride, trichloroacetate, and dichloroacetate in the liver of male B6C3F1 mice. Toxicology,
199, 169-183. http://dx.doi.Org/10.1016/i.tox.2004.02.018
Burns, P. B. and Swanson, G. M. (1991). Risk of urinary bladder cancer among blacks and whites: The
role of cigarette use and occupation. Cancer Causes Control, 2, 371-379.
http://dx.doi.org/10.1007/BF00Q54297
Bushnell, P. J.; Shafer, T. J.; Bale, A. S.; Boyes, W. K.; Simmons, J. E.; Eklund, C.; Jackson, T. L.
(2005). Developing an exposure-dose-response model for the acute neurotoxicity of organic
solvents: overview and progress on in vitro models and dosimetry. Environ Toxicol Pharmacol,
19, 607-614. http://dx.doi.org/10.1016/i.etap.2004.12.026
Byczkowski, J. Z. and Fisher, J. W. (1994). Lactational transfer of tetrachloroethylene in rats. Risk Anal,
14, 339-349.
Byczkowski, J. Z. and Fisher, J. W. (1995). A computer program linking physiologically based
pharmacokinetic model with cancer risk assessment for breast-fed infants. 46, 155-163.
Byczkowski, J. Z.; Kinkead, E. R.; Leahy, H. F.; Randall, G. M.; Fisher, J. W. (1994). Computer
simulation of the lactational transfer of tetrachloroethylene in rats using a physiologically based
model. Toxicol Appl Pharmacol, 125, 228-236.
Byers, V. S.; Levin, A. S.; Ozonoff, D. M.; Baldwin, R. W. (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. http://www.ncbi.nlm.nih.gov/pubmed/3260823
Cai, P.; Konig, R.; Khan, M. F.; Kaphalia, B. S.; Ansari, G. A. (2007). Differential immune responses to
albumin adducts of reactive intermediates of trichloroethene in MRL+/+ mice. Toxicol Appl
Pharmacol, 220, 278-283. http://dx.doi.Org/10.1016/i.taap.2007.01.020
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-10 DRAFT—DO NOT CITE OR QUOTE
-------
Cai, S.-X.; Huang, M.-Y.; Chen, Z.; Liu, Y.-T.; Jin, C.; Watanabe, T., . . . Ikeda, M. (1991). Subjective
symptom increase among dry-cleaning workers exposed to tetrachloroethylene vapor. Ind Health,
29, 111-121. http://dx.doi.Org/10.2486/indhealth.29.l 11
Cal/EPA. (California Environmental Protection Agency). (2001). Public health goal for
tetrachloroethylene in drinking water. Sacramento, CA: Office of Environmental Health Hazard
Assessment. Retrieved from http://www.oehha.org/water/phg/pdf/PCEAug2001 .pdf.
Calabrese, E. J. (1983). Principles of animal extrapolation. New York: Wiley.
Calaresu, F. and Thomas, M. (1971). The function of the paramedian reticular nucleus in the control of
heart rate in the cat. J Physiol, 216, 143-158. http://www.ncbi.nlm.nih.gov/pubmed/5558346
Caldwell, D. J. (1999). Review of mononuclear cell leukemia in F-344 rat bioassays and its significance
to human cancer risk: A case study using alkyl phthalates. Regul Toxicol Pharmacol, 30, 45-53.
http://www.ncbi.nlm.nih.gov/pubmed/10464046
Callen, D. F.; Wolf, C. R.; Philpot, R. M. (1980). Cytochrome P-450 mediated genetic activity and
cytotoxicity of seven halogenated aliphatic hydrocarbons in Saccharomyces cerevisiae. Mutat Res
Genet Toxicol Environ Mutagen, 77, 55-63. http://dx.doi.org/10.1016/0165-1218(80)90120-2
Calvert, G. M.; Ruder, A. M.; Petersen, M. R. (In Press). Mortality and end-stage renal disease incidence
among dry cleaning workers. Occup Environ Med. http://dx.doi.Org/10.l 136/oem.2010.060665
Calvert, G. M.; Ruder, A. M.; Petersen, M. R. (2010). Mortality and end-stage renal disease incidence
among dry cleaning workers. Occup Environ Med. http://dx.doi.Org/10.l 136/oem.2010.060665
Calvert, J. G. (1976). Hydrocarbon involvement in photochemical smog formation in Los Angeles
atmosphere. Environ Sci Technol, 10, 256-262. http://dx.doi.org/10.1021/es60114a003
Campagna, D.; Gobba, F.; Mergler, D.; Moreau, T.; Galassi, C.; Cavalleri, A.; Huel, G. (1996). Color
vision loss among styrene-exposed workers neurotoxicological threshold assessment.
Neurotoxicology, 17, 367-374. http://www.ncbi.nlm.nih.gov/pubmed/8856733
Campagna, D.; Mergler, D.; Huel, G.; Belanger, S.; Truchon, G.; Ostiguy, C.; Drolet, D. (1995). Visual
dysfunction among styrene-exposed workers. Scand J Work Environ Health, 21, 382-390.
http: //www .ncbi .nlm .nih. gov/pubmed/8571095
Campbell, J. A.; Corrigall, A. V.; Guy, A.; Kirsch, R. E. (1991). Immunohistologic localization of alpha,
mu, and pi class glutathione S-transferases in human tissues. Cancer, 67, 1608-1613.
http://dx.doi.org/10.1002/1097-0142(19910315)67:6<1608: :AID-CNCR2820670623>3.0.CQ:2-S
Canada, E. and Health Canada. (1993). Tetrachloroethylene. Ottawa, Canada: Minister of Supply and
Services Canada. Retrieved from http://www.hc-sc.gc.ca/ewh-semt/pubs/contaminants/psll-
lsp 1/trichloroethvlene/index-eng .php.
Canner, P. L.; Berge, K. G.; Wenger, N. K.; Stamler, J.; Friedman, L.; Prineas, R. J.; Friedewald, W.
(1986). Fifteen year mortality in Coronary Drug Project patients: long-term benefit with niacin. J
Am Coll Cardiol, 8, 1245-1255. http ://www. ncbi .nlm .nih.gov/pubmed/3 782631
Cano, M. I. and Pollan, M. (2001). Non-Hodgkin's lymphomas and occupation in Sweden. Int Arch
Occup Environ Health, 74, 443-449. http://dx.doi.org/10.1007/s00420010Q248
Capo, M.; Frejo, M.; Sevil, B. (1997). The use of the glial fibrillary acidic protein (GFAP) biomarker as
an early detection model of chemically induced cancer. J Environ Pathol Toxicol Oncol, 16, 33-
39. http ://www.ncbi .nlm.nih. gov/pubmed/925 693 0
Carlson, G. P. (1974). Enhancement of the hepatotoxicity of trichloroethylene by inducers of drug
metabolism. Res Commun Mol Pathol Pharmacol, 7, 637-640.
http: //www .ncbi .nlm .nih. gov/pubmed/4363 3 5 8
Carney, E. W.; Thorsrud, B. A.; Dugard, P. H.; Zablotny, C. L. (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. http://dx.doi.org/10.1002/bdrb.20Q91
Carpenter, C. P. (1937). The chronic toxicity of tetrachlorethylene. J Ind Hyg Toxicol, 19, 323-336.
Carpenter, S. P.; Lasker, J. M.; Raucy, J. L. (1996). Expression, induction, and catalytic activity of the
ethanol-inducible cytochrome P450 (CYP2E1) in human fetal liver and hepatocytes. Mol
Pharmacol, 49, 260-268. http://www ncbi nlm nih gov/pubmed/8632758
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-11 DRAFT—DO NOT CITE OR QUOTE
-------
Caselli, P. (1978). [Depression, cancer of the rectum, suicide by trichloroethylene ingestion. Suicide
modality in Perugia from 1928 to 1977], 38, 17-24. http://www ncbi nlm niIvgov/pubmcd/642682
Cavalleri, A.; Gobba, F.; Paltrinieri, M.; Fantuzzi, G.; Righi, E.; Aggazzotti, G. (1994). Perchloroethylene
exposure can induce colour vision loss. Neurosci Lett, 179, 162-166.
http://dx.doi.org/10.1016/0304-3940(94)90959-8
Ceaurriz, J. D.; Bonnet, P.; Certin, C.; Muller, J.; Guenier, J. P. (1981). Chemicals as central nervous
system depressa.
Cederberg, H.; Henriksson, J.; Binderup, M. L. (2010). DNA damage detected by the alkaline comet
assay in the liver of mice after oral administration of tetrachloroethylene. Mutagenesis, 25, 133-
138. http://dx.doi.org/10.1093/mutage/gep051
Chakrabarti, S.; Malick, M.; Denniel, C.; Greselin, E. (1992). Species differences in the nephrotoxic
response to S-(l,2-dichlorovinyl)glutathione. Toxicol Lett, 60, 343-351.
http: //www .ncbi .nlm .nih. gov/pubmed/1595093
Chaloupka, Z. and Myslivecek, J. (1963). [Pharmacology and physical modification of so-called
premature acquiring of rhythem of evokeed cortical potentials]. 5, 184-185.
http://www.ncbi.nlm.nih.gov/pubmed/14070642
Chang, L. W.; Daniel, F. B.; DeAngelo, A. B. (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, 20, 277-288.
http://dx.doi.org/10.1002/em.28502004Q6
Chang, L. W.; DeAngelo, A. B.; Daniel, F. B. (1989). Evaluation of DNA strand breaks (SB) by
chloroacetic acids (CAs) and chloroacetaldehydes (CADs) in vivo and in vitro. Proc Ann Mtg
Am Assoc Cane Res, 30, 146.
Chang, Y.; Tai, C.; Yang, S.; Chen, C.; Shih, T.; Lin, R.; Liou, S. (2003). A cohort mortality study of
workers exposed to chlorinated organic solvents in Taiwan. Ann Epidemiol, 13, 652-660.
http://dx.doi.org/10.1016/S 1047-2797(03)00038-3
Chang, Y.; Tai, C.; Yang, S.; Lin, R.; Sung, F.; Shih, T.; Liou, S. (2005). Cancer incidence among
workers potentially exposed to chlorinated solvents in an electronics factory. J Occup Health, 47,
171-180. http://www.ncbi.nlm.nih.gov/pubmed/15824483
Chapin, R. E.; Norton, R. M.; Popp, J. A.; Bus, J. S. (1982). The effects of 2,5-hexanedione on
reproductive hormones and testicular enzyme activities in the F-344 rat. Toxicol Appl Pharmacol,
62, 262-272. http://www.ncbi.nlm.nih.gov/pubmed/6120587
Charbonneau, M.; Short, B. G.; Lock, E. A.; Swenberg, J. A. (1987). Mechanism of petroleum induced
sex-specific protein droplet nephropathy and renal cell proliferation in Fischer-344 rats:
Relevance to humans Trace Substances in Environmental Health, 21st Annual Conference, St
Louis, MO (pp. 263-273). Columbia, MO: University of Missouri.
Chatterjee, N.; Hartge, P.; Cerhan, J.; Cozen, W.; Davis, S.; Ishibe, N., . . . Severson, R. (2004). Risk of
non-Hodgkin's lymphoma and family history of lymphatic, hematologic, and other cancers.
Cancer Epidemiol Biomarkers Prev, 13, 1415-1421.
http: //www .ncbi .nlm .nih. gov/pubmed/15342441
Chemical profile: Trichloroethylene. (1997). Chemical Marketing Reporter.
Chen, C. and Blancato, J. (1987). Role of pharmacokinetic modeling in risk assessment:
Perchloroethylene as an example Pharmacokinetics in Risk Assessment: Drinking Water and
Health (Vol. 8, pp. 367-388). Washington, DC: National Academy Press.
Chen, H.-H.; Chan, M.-H.; Fu, S.-H. (2002a). Behavioural effects of tetrachloroethylene exposure in rats:
Acute and subchronic studies. Toxicology, 170, 201-209. http://dx.doi.org/10.1016/S030Q-
483X(01)00544-3
Chen, S. J.; Wang, J. L.; Chen, J. H.; Huang, R. N. (2002b). 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, 464-472. http://dx.doi.org/10.1016/SQ891-
5849(02)00817-1
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-12 DRAFT—DO NOT CITE OR QUOTE
-------
Chia, S.; Lee, J.; Chia, K.; Chan, O. (2004). Low birth weight in relation to parental occupations-a
population-based registry in Singapore (1994-1998). Neurotoxicol Teratol, 26, 285-290.
http://dx.doi.Org/10.1016/i.ntt.2003.10.009
Chien, Y. C. (1997). The influences of exposure pattern and duration on elimination kinetics and
exposure assessment of tetrachloroethylene in humans [PhD]. New Brunswick, NJ, Rutgers
University.
Chiu, W. A. and Bois, F. Y. (2006). Revisiting the population toxicokinetics of tetrachloroethylene. Arch
Toxicol, 80, 382-385. http://dx.doi.org/10.1007/s00204-006-0Q61-9
Chiu, W. A.; Micallef, S.; Monster, A. C.; 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. http://dx.doi.org/10.1093/toxsci/kfll29
Chiu, W. A.; Okino, M. S.; Evans, M. V. (2009). Characterizing uncertainty and population variability in
the toxicokinetics of trichloroethylene and metabolites in mice, rats, and humans using an
updated database, physiologically based pharmacokinetic (PBPK) model, and Bayesian approach.
Toxicol Appl Pharmacol, 241, 36-60. http://dx.doi.Org/10.1016/i.taap.2009.07.032
Chloral hydrate: a risky old psychotropic drug. (1998). Prescrire Int, 7, 88-89.
http://www.ncbi.nlm.nih.gov/pubmed/10342927
Chmielewska, B. (1984). Pharmacological properties of new cyclic derivatives of phenylsuccinimide and
their influence on noradrenaline, dopamine, serotonin, 5-hydroxyindoleacetic acid, gamma-
aminobutyric acid concentrations and monoaminooxidase activity in animal brains. Pharmazie,
39, 259-262. http://www.ncbi.nlm.nih.gov/pubmed/6204346
Chong, S. and Remington, G. (2000). Clozapine augmentation: safety and efficacy. Schizophr Bull, 26,
421-440. http://www ncbi nlm nih gov/pubmed/10885641
Chow, W. H.; McLaughlin, J. K.; Malker, H. S.; Linet, M. S.; Weiner, J. A.; Stone, B. J. (1995).
Esophageal cancer and occupation in a cohort of Swedish men. Am J Ind Med, 27, 749-757.
http://dx.doi.org/10.1002/aiim.47002705Q9
Christensen, J. and Rasmussen, K. (1990). [Exposure of Danish workers to trichloroethylene during the
period 1947-1987], Ugeskr Laeger, 152, 464-466. http://www.ncbi.nlm.nih.gov/pubmed/2309349
Chynoweth, R. (1983). Psychotherapeutic drugs. 24/10 (115-126).
Claude, J. C.; Frentzel-Beyme, R. R.; Kunze, E. (1988). Occupation and risk of cancer of the lower
urinary tract among men: A case-control study. Int J Cancer, 41, 371-379.
http://dx.doi.org/10.1002/iic.29104103Q9
Clavel, J.; Mandereau, L.; Conso, F.; Limasset, J. C.; Pourmir, I.; Flandrin, G.; Hemon, D. (1998).
Occupational exposure to solvents and hairy cell leukaemia. Occup Environ Med, 55, 59-64.
http://dx.doi.Org/10.l 136/oem. 55.1.59
Clewell, H. J. III. (2005). Use of mode of action in risk assessment: Past, present, and future. Regul
Toxicol Pharmacol, 42, 3-14. http://dx.doi.Org/10.1016/i.vrtph.2005.01.008
Clewell, H. J., Ill and Andersen, M. E. (1994). Physiologically-based pharmacokinetic modeling and
bioactivation of xenobiotics. Toxicol Ind Health, 10, 1-24.
Clewell, H. J., III.; Gentry, P. R.; Covington, T. R.; Gearhart, J. M. (2000). Development of a
physiologically based pharmacokinetic model of trichloroethylene and its metabolites for use in
risk assessment. Environ Health Perspect, 108, 283-305.
http: //www .ncbi .nlm .nih. gov/pubmed/10807559
Clewell, H. J., Ill; Gentry, P. R.; Covington, T. R.; Sarangapani, R.; Teeguarden, J. G. (2004). Evaluation
of the potential impact of age- and gender-specific pharmacokinetic differences on tissue
dosimetry. Toxicol Sci, 79, 381-383. http://dx.doi.org/10.1093/toxsci/kfh 109
Clewell, H. J., Ill; Gentry, P. R.; Kester, J. E.; Andersen, M. E. (2005). Evaluation of physiologically
based pharmacokinetic models in risk assessment: An example with perchloroethylene. Crit Rev
Toxicol, 35, 413-433. http://dx.doi.org/10.1080/10408440590931994
Clofibrate and niacin in coronary heart disease. (1975). JAMA, 231, 360-381.
http://www.ncbi.nlm.nih.gov/pubmed/1088963
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-13 DRAFT—DO NOT CITE OR QUOTE
-------
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, 556-561.
http: //www .ncbi .nlm .nih. gov/pubmed/9679115
Cojocel, C.; Beuter, W.; Miiller, W.; Mayer, D. (1989). Lipid peroxidation: A possible mechanism of
trichloroethylene-induced nephrotoxicity. Toxicology, 55, 131-141.
http://dx.doi.org/10.1016/0300-483X(89)90180-7
Coler, H. R. and Rossmiller, H. R. (1953). Tetrachlorethylene exposure in a small industry. AMA Arch
Ind Hyg Occup Med, 8, 227-238. http://www.ncbi.nlm.nih.gov/pubmed/13079324
Coleridge, H. M.; Coleridge, J. C. G.; Luck, J. C.; Norman, J. (1968). The effect of four volatile
anaesthetic agents on the impulse activity of two types of pulmonary receptor. Br J Anaesth, 40,
484-492.
Collier, J. M.; Cover, C. C.; Johnson, P. D.; Selmin, O.; Runyon, R. B. (1998). Molecular analysis of
cardio-teratogenicity of trichloroethylene in embryonic rats.
Collingwood, T.; Urnov, F.; Chatteijee, V.; Wolffe, A. (2001). Chromatin remodeling by the thyroid
hormone receptor in regulation of the thyroid-stimulating hormone alpha-subunit promoter. J Biol
Chem, 276, 34227-34234. http://dx.doi.org/10.1074/ibc.M10517220Q
Collins, J.; Kawahara, M.; Homma, E.; Kitahata, L. (1983). Alpha-chloralose suppression of neuronal
activity. Life Sci, 32, 2995-2999. http://www.ncbi.nlm.nih.gov/pubmed/6865645
Colt, J.; Baris, D.; Stewart, P.; Schned, A.; Heaney, J.; Mott, L., . . . Karagas, M. (2004). Occupation and
bladder cancer risk in a population-based case-control study in New Hampshire. Cancer Causes
Control, 15, 759-769. http://dx.doi.Org/10.1023/B:CACO.0000043426.28741.a2
Colt, J. S.; Karagas, M. R.; Schwenn, M.; Baris, D.; Johnson, A.; Stewart, P., . . . Silverman, D. T. (2011).
Occupation and bladder cancer in a population-based case-control study in Northern New
England. Occup Environ Med, 68, 239-249. http://dx.doi.Org/10.l 136/oem.2009.052571
Columbano, A.; Ledda-Columbano, G. M.; Ennas, M. G.; Curto, M.; Chelo, A.; Pani, P. (1990). Cell
proliferation and promotion of rat liver carcinogenesis: Different effect of hepatic regeneration
and mitogen induced hyperplasia on the development of enzyme-altered foci. Carcinogenesis, 11,
771-776. http://dx.doi.Org/10.1093/carcin/ll.5.771
Committee of Principal Investigators. (1978). A co-operative trial in the primary prevention of ischaemic
heart disease using clofibrate. British Heart Journal, 40, 1069-1118.
http: //www .ncbi .nlm .nih. gov/pubmed/3 61054
Connor, T. H.; Theiss, J. C.; Hanna, H. A.; Monteith, D. K.; Matney, T. S. (1985). Genotoxicity of
organic chemicals frequently found in the air of mobile homes. Toxicol Lett, 25, 33-40.
http://dx.doi.org/10.1016/0378-4274(85)90097-9
Consonni, D.; De Matteis, S.; Lubin, J. H.; Wacholder, S.; Tucker, M.; Pesatori, A. C., . . . Landi, M. T.
(2010). Lung cancer and occupation in a population-based case-control study. Am J Epidemiol,
171, 323-333. http://dx.doi.org/10.1093/aie/kwp391
Cooper, A. J. (1994). Enzymology of cysteine S-conjugate beta-lyases. In M. W. Anders & W. Dekant
(Eds.), Conjugation-dependent carcinogenicity and toxicity of foreign compounds (Vol. 27, pp.
71-113). San Diego, CA: Academic Press.
Cooper, G. S.; Parks, C. G.; Treadwell, E. L.; St Clair, E. W.; Gilkeson, G. S.; Dooley, M. A. (2004).
Occupational risk factors for the development of systemic lupus erythematosus. J Rheumatol, 31,
1928-1933. http://www ncbi nlm nih gov/pubmed/15468355
Corbett, T. H.; Hamilton, G. C.; Yoon, M. K.; Endres, J. L. (1973). Occupational exposure of operating
room personnel to trichlorethylene. Can J Anaesth, 20, 675-678.
http://www.ncbi.nlm.nih.gov/pubmed/4780181
Corbin, M.; McLean, D.; Mannetje, A.; Dryson, E.; Walls, C.; McKenzie, F., . . . Pearce, N. (2011). Lung
cancer and occupation: A New Zealand cancer registry-based case-control study. Am J Ind Med,
54, 89-101. http://dx.doi.org/10.1002/aiim.209Q6
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-14 DRAFT—DO NOT CITE OR QUOTE
-------
Cordier, S.; Clavel, J.; Limasset, J. C.; Boccon-Gibod, L.; Le Moual, N.; Mandereau, L.; Hemon, D.
(1993). Occupational risks of bladder cancer in France: A multicentre case-control study. Int J
Epidemiol, 22, 403-411. http://dx.doi.Org/10.1093/iie/22.3.403
Corrections. (2008). Biol Reprod, 79, 787-787. http://dx.doi.org/10.1095/biolreprod.108.Q72272
Corton, J.; Lapinskas, P.; Gonzalez, F. (2000). Central role of PPARalpha in the mechanism of action of
hepatocarcinogenic peroxisome proliferators. Mutat Res, 448, 139-151.
http://www.ncbi.nlm.nih.gov/pubmed/10725468
Costa, A. K. and Ivanetich, K. M. (1980). Tetrachloroethylene metabolism by the hepatic microsomal
cytochrome P-450 system. Biochem Pharmacol, 29, 2863-2869. http://dx.doi.org/10.1016/00Q6-
2952(80)90023-4
Costa, A. K. and Ivanetich, K. M. (1984). Chlorinated ethylenes: their metabolism and effect on DNA
repair in rat hepatocytes. Carcinogenesis, 5, 1629-1636.
http://www.ncbi.nlm.nih.gov/pubmed/6499114
Costa, C.; Barbara, M.; Catania, S.; Silvari, V.; Germano, M. P. (2004). Cytotoxicity evaluation after
coexposure to perchloroethylene and selected peroxidant drugs in rat hepatocytes. Toxicol In
Vitro, 18, 37-44. http://dx.doi.org/10.1016/S0887-2333(03)00133-4
Costantini, A.; Benvenuti, A.; Vineis, P.; Kriebel, D.; Tumino, R.; Ramazzotti, V., . . . Miligi, L. (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.
http://dx.doi.org/10.1002/aiim.2Q592
Costantini, A. S.; Miligi, L.; Kriebel, D.; Ramazzotti, V.; Rodella, S.; Scarpi, E., . . . Vineis, P. (2001). A
multicenter case-control study in Italy on hematolymphopoietic neoplasms and occupation.
Epidemiology, 12, 78-87. http://www.ncbi.nlm.nih.gOv/pubmed/l 1138825
Costas, K.; Knorr, R. S.; Condon, S. K. (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. http://dx.doi.org/10.1016/50048-9697(02)00169-9
Covington, T. R.; Gentry, P. R.; Van Landingham, C. B.; Andersen, M. E.; Kester, J. E.; Clewell, H. J.
(2007). The use of Markov chain Monte Carlo uncertainty analysis to support a public health goal
for perchloroethylene. Regul Toxicol Pharmacol, 47, 1-18.
http://dx.doi.Org/10.1016/i.vrtph.2006.06.008
Crebelli, R.; Conti, G.; Conti, L.; Carere, A. (1985). Mutagenicity of trichloroethylene, trichloroethanol
and chloral hydrate in Aspergillus nidulans. Mutat Res Genet Toxicol Environ Mutagen, 155,
105-111. http://dx.doi.org/10.1016/0165-1218(85)90126-0
Cummings, B. S.; Lasker, J. M.; Lash, L. H. (2000a). Expression of glutathione-dependent enzymes and
cytochrome P450s in freshly isolated and primary cultures of proximal tubular cells from human
kidney. J Pharmacol Exp Ther, 293, 677-685. http://www.ncbi.nlm.nih.gov/pubmed/10773044
Cummings, B. S.; Parker, J. C.; Lash, L. H. (2000b). Role of cytochrome P450 and glutathione S-
transferase alpha in the metabolism and cytotoxicity of trichloroethylene in rat kidney. Biochem
Pharmacol, 59, 531-543. http://dx.doi.org/10.1016/S0006-2952(99)00374-3
Cummings, B. S.; Zangar, R. C.; Novak, R. F.; Lash, L. H. (1999). Cellular distribution of cytochromes
P-450 in the rat kidney. Drug Metab Dispos, 27, 542-548.
D'Souza, R.; Bruckner, J.; Feldman, S. (1985). Oral and intravenous trichloroethylene pharmacokinetics
in the rat. J Toxicol Environ Health, 15, 587-601. http://www ncbi nlm nih gov/pubmed/4046066
Dai, X.; De Souza, A. T.; Dai, H.; Lewis, D. L.; Lee, C. K.; Spencer, A. G., . . . He, Y. D. (2007).
PPARalpha siRNA-treated expression profiles uncover the causal sufficiency network for
compound-induced liver hypertrophy. 3, e30. http://dx.doi.org/10.1371/iournal.pcbi.003003Q
Dallas, C.; Gallo, J.; Ramanathan, R.; Muralidhara, S.; Bruckner, J. (1991). Physiological
pharmacokinetic modeling of inhaled trichloroethylene in rats. Toxicol Appl Pharmacol, 110,
303-314. http: //www. ncbi .nlm. nih. go v/pubme d/1891776
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-15 DRAFT—DO NOT CITE OR QUOTE
-------
Dallas, C. E.; Chen, X. M.; Muralidhara, S.; Varkonyi, P.; Tackett, R. L.; Bruckner, J. V. (1994a). Use of
tissue disposition data from rats and dogs to determine species differences in input parameters for
a physiological model for perchloroethylene. Environ Res, 67, 54-67.
http://dx.doi.org/10.1006/enrs. 1994.1064
Dallas, C. E.; Chen, X. M.; Muralidhara, S.; Varkonyi, P.; Tackett, R. L.; Bruckner, J. V. (1995).
Physiologically based pharmacokinetic model useful in prediction of the influence of species,
dose, and exposure route on perchloroethylene pharmacokinetics. J Toxicol Environ Health, 44,
301-317. http://dx.doi.org/10.10Q6/taap.1994.1179
Dallas, C. E.; Chen, X. M.; O'Barr, K.; Muralidhara, S.; Varkonyi, P.; Bruckner, J. V. (1994b).
Development of a physiologically based pharmacokinetic model for perchloroethylene using
tissue concentration-time data. Toxicol Appl Pharmacol, 128, 50-59.
http: //dx. doi. org/10.1006/taap .1994.1179
Daniel, F. B.; DeAngelo, A. B.; Stober, J. A.; Olson, G. R.; Page, N. P. (1992). Hepatocarcinogenicity of
chloral hydrate, 2-chloroacetaldehyde, and dichloroacetic acid in the male B6C3F1 mouse.
Fundam Appl Toxicol, 19, 159-168. http://dx.doi.org/10.1016/0272-0590(92)90147-A
Daniel, F. B.; Meier, J. R.; DeAngelo, A. B. (1993). Advances in research on carcinogenic and genotoxic
by-products of chlorine disinfection: Chlorinated hydroxyfuranones and chlorinated acetic acids.
Ann 1st Super Sanita, 29, 279-291. http://www.ncbi.nlm.nih.gov/pubmed/8279719
Daniel, J. W. (1963). The metabolism of 36Cl-labelled trichloroethylene and tetrachloroethylene in the
rat. Biochem Pharmacol, 12, 795-802. http://dx.doi.org/10.1016/0006-2952(63)90109-6
Daniell, W. E.; Claypoole, K. H.; Checkoway, H.; Smith-Weller, T.; Dager, S. R.; Townes, B. D.;
Rosenstock, L. (1999). Neuropsychological function in retired workers with previous long-term
occupational exposure to solvents. Occup Environ Med, 56, 93-105.
http: //www .ncbi .nlm .nih. gov/pubmed/10448313
Danni, O.; Brossa, O.; Burdino, E.; Milillo, P.; Ugazio, G. (1981). Toxicity of halogenated hydrocarbons
in pretreated rats - an experimental model for the study of integrated permissible limits of
environmental poisons. Int Arch Occup Environ Health, 49, 165-176.
Dantzler, W.; Evans, K.; Wright, S. (1995). Kinetics of interactions of para-aminohippurate, probenecid,
cysteine conjugates and N-acetyl cysteine conjugates with basolateral organic anion transporter in
isolated rabbit proximal renal tubules. J Pharmacol Exp Ther, 272, 663-672.
http: //www .ncbi .nlm .nih. gov/pubmed/785 3180
Dardaine, V.; Legras, A.; Lanotte, R.; Brasset, N.; Furet, Y. (1992). Unrecognised chloralose poisoning.
Intensive Care Med, 18, 497. http://www.ncbi.nlm.nih.gov/pubmed/1289380
Darnerud, P. and Olsson, L. (1990). Transplacental passage and fetal kidney binding of 14C-dichlorovinyl
cysteine (DCVC) in mice. Toxicol Lett, 52, 63-72.
http: //www .ncbi .nlm .nih. gov/pubmed/235 65 72
De Ceaurriz, J.; Desiles, J. P.; Bonnet, P.; Marignac, B.; Muller, J.; Guenier, J. P. (1983). Concentration-
dependent behavioral changes in mice following short-term inhalation exposure to various
industrial solvents. Toxicol Appl Pharmacol, 67, 383-389. http://dx.doi.org/10.1016/0Q41-
008X(83)90322-8
de Faire, U.; Ericsson, C. G.; Hamsten, A.; Nilsson, J. (1995). Design features of a five-year Bezafibrate
Coronary Atherosclerosis Intervention Trial (BECAIT). Drugs under Research, 21, 105-124.
http: //www .ncbi .nlm .nih. gov/pubmed/7555614
de Jong, P. and Brenner, B. (2004). From secondary to primary prevention of progressive renal disease:
the case for screening for albuminuria. Kidney Int, 66, 2109-2118.
http://dx.doi.Org/10.llll/i.1523-1755.2004.66001.x
De Jonghe, W. R. A. and Adams, F. C. (1986). Biogeochemical cycling of organic lead compounds.
De La Iglesia, F. A.; Lewis, J. E.; Buchanan, R. A.; Marcus, E. L.; McMahon, G. (1982). Light and
electron microscopy of liver in hyperlipoproteinemic patients under long-term gemfibrozil
treatment. Atherosclerosis, 43, 19-37. http://www.ncbi.nlm.nih.gov/pubmed/6807326
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-16 DRAFT—DO NOT CITE OR QUOTE
-------
De Roos, A.; Olshan, A.; Teschke, K.; Poole, C.; Savitz, D.; Blatt, J., . . . Pollock, B. (2001). Parental
occupational exposures to chemicals and incidence of neuroblastoma in offspring. Am J
Epidemiol, 154, 106-114. http://dx.doi.Org/10.1093/aie/154.2.106
De Smet, K.; Briining, T.; Blaszkewicz, M.; Bolt, H.; Vercruysse, A.; Rogiers, V. (2000).
Biotransformation of trichloroethylene in collagen gel sandwich cultures of rat hepatocytes. Arch
Toxicol, 74, 587-592. http://www ncbi nlm nih gov/pubmed/11201665
DeAngelo, A. B.; Daniel, F. B.; Most, B. M.; Olson, G. R. (1996). The carcinogenicity of dichloroacetic
acid in the male fischer 344 rat. Toxicology, 114, 207-221. http://dx.doi.org/10.1016/S030Q-
483X(96)03510-X
DeAngelo, A. B.; Daniel, F. B.; Most, B. M.; Olson, G. R. (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. http://dx.doi.org/10.1080/00984109708984Q74
DeAngelo, A. B.; Daniel, F. B.; Stober, J. A.; Olson, G. R. (1991). The carcinogenicity of dichloroacetic
acid in the male B6C3F1 mouse. Fundam Appl Toxicol, 16, 337-347.
http://dx.doi.org/10.1016/0272-0590(91)90118-N
DeAngelo, A. B.; Daniel, F. B.; Wong, D. M.; George, M. H. (2008). The induction of hepatocellular
neoplasia by trichloroacetic acid administered in the drinking water of the male B6C3F1 mouse. J
Toxicol Environ Health A, 71, 1056-1068. http://dx.doi.org/10.1080/15287390802111952
DeAngelo, A. B.; George, M. H.; House, D. E. (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 A, 58, 485-507.
http: //www .ncbi .nlm .nih. gov/pubmed/10632141
Deerberg, F. and Miiller-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. http: //www. ncbi .nlm. nih. go v/pubme d/5 5 08977
Dees, C. and Travis, C. (1994). Trichloroacetate stimulation of liver DNA synthesis in male and female
mice. Toxicol Lett, 70, 343-355. http://dx.doi.org/10.1016/0378-4274(94)90129-5
Degrassi, F. and Tanzarella, C. (1988). Immunofluorescent staining of kinetochores in micronuclei: A
new assay for the detection of aneuploidy. Mutat Res Genet Toxicol Environ Mutagen, 203, 339-
345. http://dx.doi.org/10.1016/0165-l 161(88)90030-1
Dekant, W.; Berthold, K.; Vamvakas, S.; Henschler, D.; Anders, M. W. (1988). Thioacylating
intermediates as metabolites of S-(l,2-dichlorovinyl)-L-cysteine and S-(l,2,2-trichlorovinyl)-L-
cysteine formed by cysteine conjugate beta-lyase. Chem Res Toxicol, 1, 175-178.
http://dx.doi.org/10.1021/tx00003a0Q8
Dekant, W.; Birner, G.; Werner, M.; Parker, J. (1998). Glutathione conjugation of perchloroethene in
subcellular fractions from rodent and human liver and kidney. Chem Biol Interact, 116, 31-43.
http://dx.doi.org/10.1080/20018091Q94565
Dekant, W. and Henschler, D. (1999). Organ-specific carcinogenicity of haloalkenes mediated by
glutathione conjugation. J Cancer Res Clin Oncol, 125, 174-181.
http: //www .ncbi .nlm .nih. gov/pubmed/1023 5471
Dekant, W.; Martens, G.; Vamvakas, S.; Metzler, M.; Henschler, D. (1987). Bioactivation of
tetrachloroethylene: role of glutathione S-transferase-catalyzed conjugation versus Cytochrome
P-450-dependent phospholipid alkylation. Drug Metab Dispos, 15, 702-709.
http: //www .ncbi .nlm .nih. gov/pubmed/2891489
Dekant, W.; Metzler, M.; Henschler, D. (1986a). Identification of S-l,2-dichlorovinyl-N-acetyl-cysteine
as a urinary metabolite of trichloroethylene: a possible explanation for its nephrocarcinogenicity
in male rats. Biochem Pharmacol, 35, 2455-2458. http://dx.doi.org/10.1016/0006-2952(86)90039-
0
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-17 DRAFT—DO NOT CITE OR QUOTE
-------
Dekant, W.; Metzler, M.; Henschler, D. (1986b). Identification of S-l,2,2-trichlorovinyl-N-acetylcysteine
as a urinary metabolite of tetrachloroethylene: Bioactivation through glutathione conjugation as a
possible explanation of its nephrocarcinogenicity. J Biochem Toxicol, 1, 57-72.
http://dx.doi.org/10.1002/ibt.25700102Q6
Dekant, W.; Vamakas, S.; Berthold, K.; Schmitt, S.; Henschler, D. (1986c). Cysteine conjugate beta-lyase
mediated cleavage and mutagenicity of the tetrachloroethylene metabolite s-l,l,2-trichlorovinyl-
cysteine. Toxicol Lett, 31, 209.
Dekant, W.; Vamvakas, S.; Anders, M. W. (1989). Bioactivation of nephrotoxic haloalkenes by
glutathione conjugation: formation of toxic and mutagenic intermediates by cysteine conjugate
beta-lyase. Drug Metab Rev, 20, 43-83. http://dx.doi.org/10.3109/036025389Q8994144
Dekant, W.; Vamvakas, S.; Berthold, K.; Schmidt, S.; Wild, D.; Henschler, D. (1986d). Bacterial beta-
lyase mediated cleavage and mutagenicity of cysteine conjugates derived from the
nephrocarcinogenic alkenes trichloroethylene, tetrachloroethylene and hexachlorobutadiene.
Chem Biol Interact, 60, 31-45. htto://dx.doi.org/10.1016/0009-2797(86)90015-3
Delahunt, B.; Bethwaite, P. B.; Nacey, J. N. (1995). Occupational risk for renal cell carcinoma. A case-
control study based on the New Zealand Cancer Registry. Br J Urol, 75, 578-582.
http://dx.doi.Org/10.llll/i.1464-410X.1995.tb07410.x
Delfino, R. J.; Gone, H.; Linn, W. S.; Pellizzari, E. D.; Hu, Y. (2003a). Asthma symptoms in Hispanic
children and daily ambient exposures to toxic and criteria air pollutants. Environ Health Perspect,
111, 647-656. http://dx.doi.org/10.1289/ehp.5992
Delfino, R. J.; Gong, H.; Linn, W. S.; Hu, Y.; Pellizzari, E. D. (2003b). Respiratory symptoms and peak
expiratory flow in children with asthma in relation to volatile organic compounds in exhaled
breath and ambient air. J Expo Anal Environ Epidemiol, 13, 348-363.
http://dx.doi.org/10.1038/si.iea.7500287
Dell'oste, V.; Azzimonti, B.; Mondini, M.; De Andrea, M.; Borgogna, C.; Mesturini, R., . . . Gariglio, M.
(2008). Altered expression of UVB-induced cytokines in human papillomavirus-immortalized
epithelial cells. J Gen Virol, 89, 2461-2466. http://dx.doi.Org/10.1099/vir.0.83586-0
DeMarini, D. M.; Perry, E.; Shelton, M. L. (1994). Dichloroacetic acid and related compounds: Induction
of prophage in E. coli and mutagenicity and mutation spectra in Salmonella TA100. Mutagenesis,
9, 429-437. http://dx.doi.Org/10.1093/mutage/9.5.429
Deonandan, R.; Campbell, K.; Ostbye, T.; Tummon, I.; Robertson, J. (2000). A comparison of methods
for measuring socio-economic status by occupation or postal area. Chronic Dis Can, 21, 114-118.
http://www.ncbi.nlm.nih.gov/pubmed/11082347
Desi, I.; Farkas, I.; Kemeny, T. (1968). The early effects of low DDT doses on the nervous system in
animal experiments. Experientia, 24, 51-52. http://www.ncbi.nlm.nih.gov/pubmed/5637617
Devanandan, M.; Eccles, R.; Lewis, D.; Stenhouse, D. (1969). Responses of extensor alpha-motoneurones
in cats anaesthetised with chloralose. Exp Brain Res, 8, 177-189.
http: //www .ncbi .nlm .nih. gov/pubmed/5 808757
Dietert, R. R. (2008). Developmental immunotoxicology (DIT): windows of vulnerability, immune
dysfunction and safety assessment. J Immunotoxicol, 5, 401-412.
http://dx.doi.org/10.1080/154769108Q2483324
Dietz, A. C. and Schnoor, J. L. (2001). Phytotoxicity of chlorinated aliphatics to hybrid poplar (Populus
deltoides x nigra DN34). Environ Toxicol Chem, 20, 389-393.
Ding, R.; Shen, T.; Zhu, Q. X. (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 ZaZhi, 24, 662-665. http://www.ncbi.nlm.nih.gov/pubmed/17181943
Diodovich, C.; Ferrario, D.; Casati, B.; Malerba, I.; Marafante, E.; Parent-Massin, D.; Gribaldo, L.
(2005). Sensitivity of human cord blood cells to tetrachloroethylene: Cellular and molecular
endpoints. Arch Toxicol, 79, 508-514. http://dx.doi.org/10.1007/s00204-005-Q662-8
DiRenzo, A. B.; Gandolfi, A. J.; Sipes, I. G. (1982). Microsomal bioactivation and covalent binding of
aliphatic halides to DNA. Toxicol Lett, 11, 243-252.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-18 DRAFT—DO NOT CITE OR QUOTE
-------
Dmitrieva, N. V. (1967). [On the metabolism of tetrachloroethylene]. Gig Tr Prof Zabol, 11, 54-56.
http://www.ncbi.nlm.nih.gov/pubmed/5629546
Dobkin, A.; Byles, P.; Neville, J. (1966). Neuroendocrine and metabolic effects of general anaesthesia
during spontaneous breathing, controlled breathing, milk hypoxia, and mild hypercarbia. 13, 130-
171. http://www.ncbi.nlm.nih.gov/pubmed/4381209
Dobrev, I. D.; Andersen, M. E.; Yang, R. S. (2001). Assessing interaction thresholds for trichloroethylene
in combination with tetrachloroethylene and 1,1,1-trichloroethane using gas uptake studies and
PBPK modeling. Arch Toxicol, 75, 134-144. http://dx.doi.org/10.1007/s00204010Q216
Dobrev, I. D.; Andersen, M. E.; Yang, R. S. (2002). In silico toxicology: Simulating interaction
thresholds for human exposure to mixtures of trichloroethylene, tetrachloroethylene, and 1,1,1-
trichloroethane. Environ Health Perspect, 110, 1031-1039. http://dx.doi.org/12361929
Doherty, A. T.; Ellard, S.; Parry, E. M.; Parry, J. M. (1996). An investigation into the activation and
deactivation of chlorinated hydrocarbons to genotoxins in metabolically competent human cells.
Mutagenesis, 11, 247-274. http://dx.doi.Org/10.1093/mutage/ll.3.247
Dohn, D. R. and Anders, M. W. (1982). Assay of cysteine conjugate beta-lyase activity with S-(2-
benzothiazolyl)cysteine as the substrate. Anal Biochem, 120, 379-386.
http://dx.doi.org/10.1016/0003-2697(82)90361-X
Dombrowski, F.; Klingmiiller, D.; Pfeifer, U. (1998). Insulinomas derived from hyperplastic intra-hepatic
islet transplants. Am J Pathol, 152, 1025-1038. http://www.ncbi.nlm.nih.gov/pubmed/9546363
Donoghue, A. M.; Dryson, E. W.; Wynn-Williams, G. (1995). Contrast sensitivity in organic-solvent-
induced chronic toxic encephalopathy. J Occup Environ Med, 37, 1357-1363.
http://www.ncbi.nlm.nih.gov/pubmed/8749741
Dosemeci, M.; Cocco, P.; Chow, W. H. (1999). Gender differences in risk of renal cell carcinoma and
occupational exposures to chlorinated aliphatic hydrocarbons. Am J Ind Med, 36, 54-59.
http: //www .ncbi .nlm .nih. gov/pubmed/10361587
Dosemeci, M.; Cocco, P.; Gomez, M.; Stewart, P. A.; Heineman, E. F. (1994). Effects of three features of
a job-exposure matrix on risk estimates. Epidemiology, 5, 124-127.
http://www.ncbi.nlm.nih.g0v/pubmed/8117771
Dow, J. and Green, T. (2000). Trichloroethylene induced vitamin B(12) and folate deficiency leads to
increased formic acid excretion in the rat. Toxicology, 146, 123-136.
http://www.ncbi.nlm.nih.gov/pubmed/10814845
Doyle, P.; Roman, E.; Beral, V.; Brookes, M. (1997). Spontaneous abortion in dry cleaning workers
potentially exposed to perchloroethylene. Occup Environ Med, 54, 848-853.
http://dx.doi.Org/10.l 136/oem.54.12.848
Dreessen, B.; Westphal, G.; Biinger, J.; Hallier, E.; Miiller, M. (2003). Mutagenicity of the glutathione
and cysteine S-conjugates of the haloalkenes l,l,2-trichloro-3,3,3-trifluoro-l-propene and
trichlorofluoroethene in the Ames test in comparison with the tetrachloroethene-analogues. Mutat
Res Genet Toxicol Environ Mutagen, 539, 157-166. http://dx.doi.org/10.1016/S1383-
5718(03)00160-8
Droz, P. O. and Fernandez, J. G. (1978). Trichloroethylene exposure Biological monitoring by breath and
urine analyses. Occup Environ Med, 35, 35-42.
Droz, P. O. and Guillemin, M. P. (1986). Occupational exposure monitoring using breath analysis. J
Occup Med, 28, 593-602. http://www.ncbi.nlm.nih.gov/pubmed/3746479
Droz, P. O.; Wu, M. M.; Cumberland, W. G.; Berode, M. (1989). Variability in biological monitoring of
solvent exposure. I. Development of a population physiological model. Br J Ind Med, 46, 447-
460. http://www.ncbi.nlm.nih.gov/pubmed/2765418
Dryson, E.; 't Mannetje, A.; Walls, C.; McLean, D.; McKenzie, F.; Maule, M., . . . Pearce, N. (2008).
Case-control study of high risk occupations for bladder cancer in New Zealand. Int J Cancer, 122,
1340-1346. http://dx.doi.org/10.1002/iic.23194
Dubrow, R. and Gute, D. (1987). Cause-specific mortality among Rhode Island jewelry workers. Am J
Ind Med, 12, 579-593. http://www.ncbi.nlm.nih.gov/pubmed/2961258
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-19 DRAFT—DO NOT CITE OR QUOTE
-------
Duffel, M. W. and Jakoby, W. B. (1982). Cysteine S-conjugate N-acetyltransferase from rat kidney
microsomes. Mol Pharmacol, 21, 444-448. htto://www ncbi nlm nib gov/pubmed/6892478
Dugas, M. (1985). Gilles de la Tourette's syndrome. Present status of tic disease Le syndrome
de gilles de la tourette. Etat actuel de la maladie des tics. 14/10 (589-593).
Diindar, Y.; Boland, A.; Strobl, J.; Dodd, S.; Haycox, A.; Bagust, A., . . . Walley, T. (2004). Newer
hypnotic drugs for the short-term management of insomnia: a systematic review and economic
evaluation. Health Technol Assess, 8, iii-x, 1-125.
htto: //www .ncbi .nlm .nih. gov/pubmed/15193209
Dunnett, S.; Torres, E.; Richards, H.; Barker, R. (1998). Effects of surgical anaesthesia on the viability of
nigral grafts in the rat striatum. Cell Transplant, 7, 567-572.
htto: //www .ncbi .nlm .nih. gov/pubmed/9853585
Dunson, D.; Holloman, C.; Calder, C.; Gunn, L. (2004). Bayesian modeling of multiple lesion onset and
growth from interval-censored data. Biometrics, 60, 676-683. http://dx.doi.org/10. Ill 1/i .0006-
341X.2004.00217.X
Duplus, E. and Forest, C. (2002). Is there a single mechanism for fatty acid regulation of gene
transcription. Biochem Pharmacol, 64, 893-901. htto://www ncbi nlm nih gov/pubmed/12213584
Eastmond, D. A.; Hartwig, A.; Anderson, D.; Anwar, W. A.; Cimino, M. C.; Dobrev, I., . . . Vickers, C.
(2009). Mutagenicity testing for chemical risk assessment: Update of the WHO/IPCS
Harmonized Scheme. Mutagenesis, 24, 341-349. http://dx.doi.org/10.1093/mutage/gep014
Eastmond, D. A. and Tucker, J. D. (1989). Kinetochore localization in micronucleated cytokinesis-
blocked Chinese hamster ovary cells: a new and rapid assay for identifying aneuploidy-inducing
agents. Mutat Res, 224, 517-525. htto://www.ncbi.nlm.nih.gov/pubmed/2685592
Ebrahim, A. S.; Babakrishnan, K.; Sakthisekaran, D. (1996). Perchloroethylene-induced alterations in
glucose metabolism and their prevention by 2-deoxy-D-glucose and vitamin E in mice. J Appl
Toxicol, 16, 339-348. http://dx.doi.org/10.1002/(SICni099-1263(199607)16:4<339::AID-
JAT352>3.0.CO:2-3
Ebrahim, A. S.; Babu, E.; Thirunavukkarasu, C.; Sakthisekaran, D. (2001). Protective role of vitamin E,
2-deoxy-D-glucose, and taurine on perchloroethylene induced alterations in ATPases. Drug Chem
Toxicol, 24, 429-437. http://dx.doi.org/10.1081/DCT-100106267
Echeverria, D.; Heyer, N.; Checkoway, H.; Brodkin, C. A.; Bittner, A. J.; Toutonghi, G.; Ronhovde, N.
(1994). A behavioral investigation of occupational exposures to solvents: Perchloroethylene
among dry cleaners, and styrene among reinforced fiberglass laminators. (BSRC-100/94/040).
Seattle, WA: Battelle Centers for Public Health Research and Evaluation.
Echeverria, D.; White, R. F.; Sampaio, C. (1995). A behavioral evaluation of PCE exposure in patients
and dry cleaners: A possible relationship between clinical and preclinical effects. J Occup
Environ Med, 37, 667-680. http://www.ncbi.nlm.nih.gov/pubmed/7670913
Effect of fenofibrate on progression of coronary-artery disease in type 2 diabetes: the Diabetes
Atherosclerosis Intervention Study, a randomised study. (2001). Lancet, 357, 905-910.
htto: //www .ncbi .nlm .nih. gov/pubmed/11289345
Egelhoff, J.; Ball, W.; Koch, B.; Parks, T. (1997). Safety and efficacy of sedation in children using a
structured sedation program. AJR Am J Roentgenol, 168, 1259-1262.
htto: //www .ncbi .nlm .nih. gov/pubmed/9129423
Eger EI, 1.1. (1963). A mathematical model of uptake and distribution. In E. M. Papper & R. J. Kitz
(Eds.), Uptake and distribution of anesthetic agents (pp. 72-87). New York, NY: McGraw-Hill
Book Company, Inc.
Elfarra, A. A. and Krause, R. J. (2007). S-(l,2,2-trichlorovinyl)-L-cysteine sulfoxide, a reactive
metabolite of S-(l,2,2-Trichlorovinyl)-L-cysteine formed in rat liver and kidney microsomes, is a
potent nephrotoxicant. J Pharmacol Exp Ther, 321, 1095-1101.
http://dx.doi.org/10.1124/jpet. 107.120444
Elton, M. (1986). Alcohol withdrawal: clinical symptoms and management of the syndrome. Acta
Psychiatr Scand Suppl, 327, 80-90. http://www.ncbi.nlm.nih.gov/pubmed/2875614
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-20 DRAFT—DO NOT CITE OR QUOTE
-------
Emara, A. M.; Abo El-Noor, M. M.; Hassan, N. A.; Wagih, A. A. (2010). Immunotoxicity and
hematotoxicity induced by tetrachloroethylene in egyptian dry cleaning workers. Inhal Toxicol,
22, 117-124. http://dx.doi.org/10.3109/08958370902934894
Emmert, B.; Biinger, J.; Keuch, K.; Miiller, M.; Emmert, S.; Hallier, E.; Westphal, G. A. (2006).
Mutagenicity of cytochrome P450 2E1 substrates in the Ames test with the metabolic competent
S. typhimurium strain YG7108pin3ERb5. Toxicology, 228, 66-76.
http://dx.doi.Org/10.1016/i.tox.2006.08.013
Entz, R. C. and Diachenko, G. W. (1988). Residues of volatile halocarbons in margarines. Food Addit
Contam, 5, 267-276. http://www.ncbi.nlm.nih.gov/pubmed/3396733
Epstein, D. L.; Nolen, G. A.; Randall, J. L.; Christ, S. A.; Read, E. J.; Stober, J. A.; Smith, M. K. (1992).
Cardiopathic effects of dichloroacetate in the fetal Long-Evans rat. Teratology, 46, 225-235.
http://dx.doi.org/10.1002/tera.14204603Q6
Erbsloh, J. (1981). [Immobilization of horses with drugs]. 9, 221-226.
http: //www .ncbi .nlm .nih. gov/pubmed/7348931
Erickson, C. K.; Tyler, T. D.; Harris, R. A. (1978). Ethanol: Modification of acute intoxication by
divalent-cations. Science, 199, 1219 - 1221.
Eskenazi, B.; Bracken, M.; Holford, T.; Grady, J. (1988). Exposure to organic solvents and hypertensive
disorders of pregnancy. Am J Ind Med, 14, 177-188.
http: //www .ncbi .nlm .nih. gov/pubmed/3207103
Eskenazi, B.; Fenster, L.; Hudes, M.; Wyrobek, A. J.; Katz, D. F.; Gerson, J.; Rempel, D. M. (1991a). A
study of the effect of perchloroethylene exposure on the reproductive outcomes of wives of dry-
cleaning workers. Am J Ind Med, 20, 593-600. http://dx.doi.org/10.1002/aiim.47002005Q3
Eskenazi, B.; Wyrobek, A. J.; Fenster, L.; Katz, D. F.; Sadler, M.; Lee, J., . . . Rempel, D. M. (1991b). A
study of the effect of perchloroethylene exposure on semen quality in dry cleaning workers. Am J
Ind Med, 20, 575-591. http://dx.doi.org/10.1002/aiim.47002005Q2
Essing, H.-G.; Schacke, G.; Schaller, K.-H.; Valentin, H. (1973). Occupational-medical studies on the
dynamics of tetrachloroethylene in the organism. 24, 242-244.
Esteller, M. (2003). Relevance of DNA methylation in the management of cancer. Lancet Oncol, 4, 351-
358. http://www.ncbi.nlm.nih.gov/pubmed/12788407
Fabbro-Peray, P.; Daures, J. P.; Rossi, J. F. (2001). Environmental risk factors for non-Hodgkin's
lymphoma: A population-based case-control study in Languedoc-Roussillon, France. Cancer
Causes Control, 12, 201-212. http://dx.doi.org/10.1023/A: 1011274922701
Fang, H. L.; Strom, S. C.; Cai, H.; Falany, C. N.; Kocarek, T. A.; Runge-Morris, M. (2005). Regulation of
human hepatic hydroxysteroid sulfotransferase gene expression by the peroxisome proliferator-
activated receptor alpha transcription factor. Mol Pharmacol, 67, 1257-1267.
http://dx.doi.Org/10.l 124/mol. 104.005389
FDA. (U.S. Food and Drug Administration). (2003). Food and Drug Administration total diet study:
summary of residues found, ordered by pesticide. Center for Food Safety and Nutrition.
Feighner, J. (1985). A comparative trial of fluoxetine and amitriptyline in patients with major depressive
disorder. J Clin Psychiatry, 46, 369-372. http://www.ncbi.nlm.nih.gov/pubmed/3897204
Fender, H. (1993). Chromosomenanalytische Undersuchungen bei Textilreinigern. In D. Arndt & G. Obe
(Eds.), Cytogenetische Methoden im Rahmen des Populationsmonitoring (pp. 71-76). Miinchen:
MMV Verlag.
Ferguson, R. and Vernon, R. (1970). Trichloroethylene in combination with CNS drugs. Effects on
visual-motor tests. Arch Environ Health, 20, 462-467.
http://www.ncbi.nlm.nih.gov/pubmed/4393402
Fernandez, J.; Guberan, E.; Caperos, J. (1976). Experimental human exposures to tetrachloroethylene
vapor and elimination in breath after inhalation. Am Ind Hyg Assoc J, 36, 143-150.
http: //www .ncbi .nlm .nih. gov/pubmed/1266733
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-21 DRAFT—DO NOT CITE OR QUOTE
-------
Ferreira-Gonzalez, A.; DeAngelo, A. B.; Nasim, S.; Garrett, C. T. (1995). Ras oncogene activation during
hepatocarcinogenesis in B6C3F1 male mice by dichloroacetic and trichloroacetic acids.
Carcinogenesis, 16, 495-500. http://dx.doi.Org/10.1093/carcin/16.3.495
Ferroni, C.; Selis, L.; Mutti, A.; Folli, D.; Bergamaschi, E.; Franchini, I. (1992). Neurobehavioral and
neuroendocrine effects of occupational exposure to perchloroethylene. Neurotoxicology, 13, 243-
247. http://www.ncbi.nlm.nih.gov/pubmed/1508425
Filser, J. G. and Bolt, H. M. (1979). Pharmacokinetics of halogenated ethylenes in rats. Arch Toxicol, 42,
123-136. http://dx.doi.org/10.1007/BF00316492
Finckh, A.; Cooper, G. S.; Chibnik, L. B.; Costenbader, K. H.; Watts, J.; Pankey, H.,. . . Karlson, E. W.
(2006). Occupational silica and solvent exposures and risk of systemic lupus erythematosus in
urban women. Arthritis Rheum, 54, 3648-3654. http://dx.doi.org/10.1002/art.22210
Finucane, T. (1984). Flurazepam and other benzodiazepines. Ann Intern Med, 101, 403-404.
http://www.ncbi.nlm.nih.gov/pubmed/6147111
Fisher, D.; O'Keeffe, C.; Stanski, D.; Cronnelly, R.; Miller, R.; Gregory, G. (1982). Pharmacokinetics and
pharmacodynamics of d-tubocurarine in infants, children, and adults. Anesthesiology, 57, 203-
208. http://www.ncbi.nlm.nih.gov/pubmed/7114542
Fisher, J.; Lumpkin, M.; Boyd, J.; Mahle, D.; Bruckner, J. D.; El-Masri, H. A. (2004). PBPK modeling of
the metabolic interactions of carbon tetrachloride and tetrachloroethylene in B6C3F1 mice.
Environ Toxicol Pharmacol, 16, 93-15.
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, 425-431.
http://dx.doi.org/10.1080/15428119791Q12667
Flavell, D. M.; Ireland, H.; Stephens, J. W.; Hawe, E.; Acharya, J.; Mather, H., . . . Humphries, S. E.
(2005). Peroxisome proliferator-activated receptor alpha gene variation influences age of onset
and progression of type 2 diabetes. Diabetes, 54, 582-586.
http: //www .ncbi .nlm .nih. gov/pubmed/15677519
Flavell, D. M.; Jamshidi, Y.; Hawe, E.; Pineda Torra, I.; Taskinen, M. R.; Frick, M. H., . . . Syvanne, M.
(2002). Peroxisome proliferator-activated receptor alpha gene variants influence progression of
coronary atherosclerosis and risk of coronary artery disease. Circulation, 105, 1440-1445.
http: //www .ncbi .nlm .nih. gov/pubmed/11914252
Fleming-Jones, M. E. and Smith, R. E. (2003). Volatile organic compounds in foods: a five year study. J
Agric Food Chem, 51, 8120-8127.
Flood, T. J. (1997). Case-referent study of childhood leukemia in Maricopa County, Arizona: 1965-1990.
Phoenix, AZ: Arizona Dept. of Health Services, Bureau of Epidemiology and Disease Control
Services, Office of Chronic Disease Epidemiology, Office of Environmental Health.
Forkert, P.; 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, 306-311.
http: //www .ncbi .nlm .nih. gov/pubmed/12584157
Franchini, I.; Cavatorta, A.; Falzoi, M.; Lucertini, S.; Mutti, A. (1983). Early indicators of renal damage
in workers exposed to organic solvents. Int Arch Occup Environ Health, 52, 1-9.
Frankel, D. M.; Johnson, C. E.; Pitt, H. M. (1957). Preparation and properties of tetrachloroethylene
oxide. J Org Chem, 22, 1119-1120.
Frantz, S. W. and Watanabe, P. G. (1983). Tetrachloroethylene: Balance and tissue distribution in male
Sprague-Dawley rats by drinking-water administration. Toxicol Appl Pharmacol, 69, 66-72.
http://dx.doi.org/10.1016/0041-008X(83')90120-5
Fredriksson, A.; Danielsson, B. R. G.; Eriksson, P. (1993). Altered behaviour in adult mice orally
exposed to tri- and tetrachloroethylene as neonates. Toxicol Lett, 66, 13-19.
http://dx.doi.org/10.1016/0378-4274(93)90074-8
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-22 DRAFT—DO NOT CITE OR QUOTE
-------
Fredriksson, M.; Bengtsson, N. O.; Hardell, L.; Axelson, O. (1989). Colon cancer, physical activity, and
occupational exposures. A case-control study. Cancer, 63, 1838-1842.
http: //www .ncbi .nlm .nih. gov/pubmed/27025 92
Freeman, S. R.; Drake, A. L.; Heilig, L. F.; Graber, M.; McNealy, K.; Schilling, L. M.; Dellavalle, R. P.
(2006). Statins, fibrates, and melanoma risk: a systematic review and meta-analysis. 98, 1538-
1546. http://dx.doi.org/10.1093/inci/dii412
Frenette, B.; Mergler, D.; Bowler, R. (1991). Contrast-sensitivity loss in a group of former
microelectronics workers with normal visual acuity. 68, 556-560.
Frick, M. H.; Syvanne, M.; Nieminen, M. S.; Kauma, H.; Majahalme, S.; Virtanen, V., . . . Taskinen, M.
R. (1997). Prevention of the angiographic progression of coronary and vein-graft atherosclerosis
by gemfibrozil after coronary bypass surgery in men with low levels of HDL cholesterol. Lopid
Coronary Angiography Trial (LOCAT) Study Group. Circulation, 96, 2137-2143.
http://www.ncbi.nlm .nih.gov/pubmed/9337181
Furnus, C. C.; Ulrich, M. A.; Terreros, M. C.; Dulout, F. N. (1990). The induction of aneuploidy in
cultured Chinese hamster cells by propionaldehyde and chloral hydrate. Mutagenesis, 5, 323-326.
http: //www .ncbi .nlm .nih. gov/pubmed/2398816
Fuscoe, J. C.; Afshari, A. J.; George, M. H.; DeAngelo, A. B.; Tice, R. R.; Salman, T.; Allen, J. W.
(1996). In vivo genotoxicity of dichloroacetic acid: Evaluation with the mouse peripheral blood
micronucleus assay and the single cell gel assay. Environ Mol Mutagen, 27, 1-9.
http://dx.doi.org/10.1002/(SICD1098-2280a996)27:l3.0.CQ:2-L
Gaertner, R. R. W.; Trpeski, L.; Johnson, K. C. (2004). A case-control study of occupational risk factors
for bladder cancer in Canada. Cancer Causes Control, 15, 1007-1019.
http://dx.doi.org/10.1007/slQ552-004-1448-7
Gaillard, Y.; Billault, F.; Pepin, G. (1995). Tetrachloroethylene fatality: Case report and simple gas
chromatographic determination in blood and tissues. Forensic Sci Int, 76, 161-168.
http://dx.doi.org/10.1016/0379-0738(95)01813-1
Gallbladder disease as a side effect of drugs influencing lipid metabolism. Experience in the Coronary
Drug Project. (1977). N Engl J Med, 296, 1185-1190.
http://dx.doi.org/10.1056/NEJM1977Q5262962101
Galloway, S. M.; Armstrong, M. J.; Reuben, C.; Colman, S.; Brown, B.; Cannon, C., . . . Zeiger, E.
(1987). Chromosome aberrations and sister chromatid exchanges in Chinese hamster ovary cells:
Evaluations of 108 chemicals. Environ Mol Mutagen, 10, 1-175.
http://dx.doi.org/10.1002/em.28501005Q2
Gama, C. and Meira, J. B. (1978). Occupational acro-osteolysis. J Bone Joint Surg Am, 60, 86-90.
Ganning, A.; Brunk, U.; Dallner, G. (1984). Phthalate esters and their effect on the liver. Hepatology, 4,
541-547. http: //www. ncbi .nlm. nih. go v/pubme d/6373551
Ganning, A. E.; Brunk, U.; Edlund, C.; Elhammer, A.; Dallner, G. (1987). Effects of prolonged
administration of phthalate ester on the liver. Environ Health Perspect, 73, 251-258.
http: //www .ncbi .nlm .nih. gov/pubmed/3 665868
Garabrant, D. H.; Lacey JV, J.; Laing, T. J.; Gillespie, B. W.; Mayes, M. D.; Cooper, B. C.; Schottenfeld,
D. (2003). Scleroderma and solvent exposure among women. Am J Epidemiol, 157, 493-500.
http://dx.doi.org/10.1093/aie/kwf223
Garetano, G. and Gochfeld, M. (2000). Factors influencing tetrachloroethylene concentrations in
residences above dry-cleaning establishments. Arch Environ Health, 55, 59-68.
http://dx.doi.org/10.1080/000398900096Q3387
Gariot, P.; Barrat, E.; Drouin, P.; Genton, P.; Pointel, J. P.; Foliguet, B., . . . Debry, G. (1987).
Morphometric study of human hepatic cell modifications induced by fenofibrate. Metabolism, 36,
203-210. http: //www. ncbi .nlm. nih. go v/pubme d/3 821501
Gamier, R.; Bedouin, J.; Pepin, G.; Gaillard, Y. (1996). Coin-operated dry cleaning machines may be
responsible for acute tetrachloroethylene poisoning: Report of 26 cases including one death. Clin
Toxicol, 34, 191-197. http://www.ncbi.n1m nih gov/pubmed/8618253
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-23 DRAFT—DO NOT CITE OR QUOTE
-------
Ge, R.; Wang, W.; Kramer, P. M.; Yang, S.; Tao, L.; Pereira, M. A. (2001a). Wy-14,643-induced
hypomethylation of the c-myc gene in mouse liver. Toxicol Sci, 62, 28-35.
http://www.ncbi.nlm.nih.gov/pubmed/11399790
Ge, R.; Yang, S.; Kramer, P. M.; Tao, L.; Pereira, M. A. (2001b). The effect of dichloroacetic acid and
trichloroacetic acid on DNA methylation and cell proliferation in B6C3F1 mice. J Biochem Mol
Toxicol, 15, 100-106. htto://dx.doi.org/10.1002/ibt.5
Gearhart, J. M.; Mahle, D. A.; Greene, R. J.; Seckel, C. S.; Flemming, C. D.; Fisher, J. W.; 3rd, C. H.
(1993). Variability of physiologically based pharmacokinetic (PBPK) model parameters and their
effects on PBPK model predictions in a risk assessment for perchloroethylene (PCE). Toxicol
Lett, 68, 131-144. http://dx.doi.org/10.1016/0378-4274(93)90126-1
Geller, A. M. (2001). A table of color distance scores for quantitative scoring of the Lanthony Desaturate
color vision test. Neurotoxicol Teratol, 23, 265-267.
Gennari, P.; Naldi, M.; Motta, R.; Nucci, M. C.; Giacomini, C.; Violante, F. S.; Raffi, G. B. (1992).
gamma-Glutamyltransferase isoenzyme pattern in workers exposed to tetrachloroethylene. Am J
Ind Med, 21, 661-671. htto://dx.doi.org/10.1002/aiim.4700210506
Gentry, P. R.; Covington, T. R.; Clewell, H. J., III. (2003). Evaluation of the potential impact of
pharmacokinetic differences on tissue dosimetry in offspring during pregnancy and lactation.
Regul Toxicol Pharmacol, 38, 1-16. http://dx.doi.org/10.1016/S0273-2300(03)00047-3
George, J. D.; Price, C. J.; Navarro, H. A.; Marr, M. C.; Myers, C. B.; Hunter ES Schwetz, B. A.; Shelby,
M. D. (1995). Developmental toxicity ofthiophenol (THIO) in rats and rabbits. Toxicologist, 15,
160.
Germolec, D. R.; Yang, R. S. H.; Ackermann, M. F.; Rosenthal, G. J.; Boorman, G. A.; Blair, P.; Luster,
M. I. (1989). Toxicology studies of a chemical mixture of 25 groundwater contaminants: II.
Immunosuppression in B6C3F1 mice. Fundam Appl Toxicol, 13, 377-387.
http://dx.doi.org/10.1016/0272-0590(89)90275-3
Ghantous, H.; Danielsson, B. R. G.; Dencker, L.; Gorczak, J.; Vesterberg, O. (1986). Trichloroacetic acid
accumulates in murine amniotic fluid after tri- and tetrachloroethylene inhalation. Basic Clin
Pharmacol Toxicol, 58, 105-114. http://dx.doi.Org/10.l 111/i. 1600-0773.1986.tb00078.x
Ghose, K. (1985). Psychopharmacological agents in geriatric medicine. 21/4 (187-197).
Giller, S.; Le Curieux, F.; Erb, F.; Marzin, D. (1997). Comparative genotoxicity of halogenated acetic
acids found in drinking water. Mutagenesis, 12, 321-328.
http://www.ncbi.nlm.nih.gov/pubmed/9379909
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, 1243-1249. http://dx.doi.org/10.1289/ehp.7568
Ginsberg, G.; Smolenski, S.; Hattis, D.; Guyton, K. Z.; Johns, D. O.; B, S. (2009). Genetic polymorphism
in glutathione transferases (GST): Population distribution of GSTM1, Tl, and PI conjugating
activity. J Toxicol Environ Health B Crit Rev, 12, 389-439.
http://dx.doi.org/10.1080/109374009Q3158375
Gobba, F.; Righi, E.; Fantuzzi, G.; Predieri, G.; Cavazzuti, L.; Aggazzotti, G. (1998). Two-year evolution
of perchloroethylene-induced color-vision loss. Arch Environ Health, 53, 196-198.
http://www.ncbi.nlm.nih.gov/pubmed/9814715
Gold, L.; Milliken, K.; Stewart, P.; Purdue, M.; Severson, R.; Seixas, N., . . . De Roos, A. (2010a).
Occupation and multiple myeloma: an occupation and industry analysis. Am J Ind Med, 53, 768-
779. http://dx.doi.org/10.1002/aiim.20857
Gold, L.; Stewart, P.; Milliken, K.; Purdue, M.; Severson, R.; Seixas, N., . . . De Roos, A. (2010b). The
relationship between multiple myeloma and occupational exposure to six chlorinated solvents.
Occup Environ Med. http://dx.doi.Org/10.l 136/oem.2009.054809
Gold, L. S.; De Roos, A. J.; Waters, M.; Stewart, P. (2008). Systematic literature review of uses and
levels of occupational exposure to tetrachloroethylene. J Occup Environ Hyg, 5, 807-839.
http://dx.doi.org/10.1080/15459620802510866
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-24 DRAFT—DO NOT CITE OR QUOTE
-------
Gold, L. S.; Manley, N. B.; Slone, T. H.; Rohrbach, L.; Garfinkel, G. B. (2005). Supplement to the
Carcinogenic Potency Database (CPDB): results of animal bioassays published in the general
literature through 1997 and by the National Toxicology Program in 1997-1998. Toxicol Sci, 85,
747-808. http://dx.doi.org/10.1093/toxsci/kfil61
Goldman, J. A. (1996). Connective tissue disease in people exposed to organic chemical solvents:
Systemic sclerosis (scleroderma) in dry cleaning plant and aircraft industry workers. J Clin
Rheumatol, 2, 185-190. http://www nchi nlm nih gov/pubmed/19078063
Goldstein, A.; Aronow, L.; Kalman, S. M. (1969). Principles of drug action. New York, NY: Harper and
Row.
Goldsworthy, T. L.; Lyght, O.; Burnett, V. L.; Popp, J. A. (1988). Potential role of [alpha]-2[mu]-
globulin, protein droplet accumulation, and cell replication in the renal carcinogenicity of rats
exposed to trichloroethylene, perchloroethylene, and pentachloroethane. Toxicol Appl Pharmacol,
96, 367-379. http://dx.doi.org/10.1016/0041-008X(88)90095-6
Goldsworthy, T. L. and Popp, J. A. (1987). Chlorinated hydrocarbon-induced peroxisomal enzyme
activity in relation to species and organ carcinogenicity. Toxicol Appl Pharmacol, 88, 225-233.
htto: //dx .doi.org/10.1016/0041 -008X(87)90008-1
Goodman, D. G.; Ward, J. M.; Squire, R. A.; Chu, K. C.; Linhart, M. S. (1979). Neoplastic and
nonneoplastic lesions in aging F344 rats. Toxicol Appl Pharmacol, 48, 237-248.
http://dx.doi.org/10.1016/004 l-008X(79')90029-2
Governa, M.; Calisti, R.; Coppa, G.; Tagliavento, G.; Colombi, A.; Troni, W. (2003). Urinary excretion
of 2,5-hexanedione and peripheral polyneuropathies in workers exposed to hexane. J Toxicol
Environ Health, 20, 219-228.
Gralewicz, S. and Dyzma, M. (2005). Organic solvents and the dopaminergic system. Int J Occup Med
Environ Health, 18, 103 - 113.
Grawe, J.; Niisse, M.; Adler, I. (1997). Quantitative and qualitative studies of micronucleus induction in
mouse erythrocytes using flow cytometry. I. Measurement of micronucleus induction in
peripheral blood polychromatic erythrocytes by chemicals with known and suspected
genotoxicity. Mutagenesis, 12, 1-8. http://www.ncbi.nlm.nih.gov/pubmed/9025090
Gray, T. J.; Lake, B. G.; Beamand, J. A.; Foster, J. R.; Gangolli, S. D. (1983). Peroxisomal effects of
phthalate esters in primary cultures of rat hepatocytes. Toxicology, 28, 167-179.
http://dx.doi.org/10.1016/0300-483X(83)90115-4
Green, T.; Dow, J.; Ellis, M. K.; Foster, J. R.; Odum, J. (1997). The role of glutathione conjugation in the
development of kidney tumours in rats exposed to trichloroethylene. Chem Biol Interact, 105, 99-
117. http://www.ncbi.nlm.nih.gov/pubmed/9251723
Green, T.; Odum, J.; Nash, J. A.; Foster, J. R. (1990). Perchloroethylene-induced rat kidney tumors: An
investigation of the mechanisms involved and their relevance to humans. Toxicol Appl
Pharmacol, 103, 77-89. http://dx.doi.org/10.1016/0041-008X(90)90264-U
Greenberg, S.; Faerber, E.; Aspinall, C.; Adams, R. (1993). High-dose chloral hydrate sedation for
children undergoing MR imaging: safety and efficacy in relation to age. AJR Am J Roentgenol,
161, 639-641. http://www nchi nlm nih gov/pubmed/8352124
Greim, H.; Bonse, G.; Radwan, Z.; Reichert, D.; Henschler, D. (1975). Mutagenicity in vitro and potential
carcinogenicity of chlorinated ethylenes as a function of metabolic oxirane formation. Biochem
Pharmacol, 24, 2013-2017. http://dx.doi.org/10.1016/0006-2952(75)90396-2
Grivennikov, S. I.; Greten, F. R.; Karin, M. (2010). Immunity, inflammation, and cancer. Cell, 140, 883-
899. http://dx.doi.Org/10.1016/i.cell.2010.01.025
Gu, Z. W.; Sele, B.; Chmara, D.; Jalbert, P.; Vincent, M.; Vincent, F., . . . Faure, J. (1981a). [Effects of
trichlorethylene and its metabolites on the rate of sister chromatid exchange. In vivo and in vitro
study on the human lymphocytes]. Ann Genet, 24, 105-106.
http: //www .ncbi .nlm .nih. gov/pubmed/6977287
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-25 DRAFT—DO NOT CITE OR QUOTE
-------
Gu, Z. W.; Sele, B.; Jalbert, P.; Vincent, M.; Vin. (1981b). Induction d'echanges entre les chromatides
soeurs (SCE) par le trichloroethylene et ses metabolites [Induction of sister chromatid exchange
by trichloroethylene and its metabolites]. Toxicol Eur Res, 3, 63-67.
Gualandi, G. (1987). Use of alpha- and beta-tubulin mutants for the study of spontaneous and induced
chromosomal mis-distribution in Aspergillus nidulans. Mutat Res, 178, 33-41.
Guberan, E. and Fernandez, J. (1974). Control of industrial exposure to tetrachloroethylene by measuring
alveolar concentrations: theoretical approach using a mathematical model. Occup Environ Med,
31, 159-167.
Guidelines for estimating exposures, 51 34042-34054 (1986).
Guidelines for reproductive toxicity risk assessment, 61 56274-56322 (1996).
Gulyas, H. and Hemmerling, L. (1990). Tetrachloroethene air pollution originating from coin-operated
dry cleaning establishments. Environ Res, 53, 90-99.
htto: //www .ncbi .nlm .nih. gov/pubmed/22263 80
Guyton, K. Z.; Liu, Y.; Gorospe, M.; Xu, Q.; Holbrook, N. J. (1996). Activation of mitogen-activated
protein kinase by H202. Role in cell survival following oxidant injury. J Biol Chem, 271, 4138-
4142. htto://www ncbi nlm nih gov/pubmed/8626753
Hake, C. L. and Stewart, R. D. (1977). Human exposure to tetrachloroethylene: Inhalation and skin
contact. Environ Health Perspect, 21, 231-238. htto://www.ncbi.nlm.nih. gov/pubmed/612448
Hakkola, J.; Pasanen, M.; Hukkanen, J.; Pelkonen, O.; Maenpaa, J.; Edwards, R. J., . . . Raunio, H.
(1996a). Expression of xenobiotic-metabolizing cytochrome P450 forms in human full-term
placenta. Biochem Pharmacol, 51, 403-411. http://dx.doi.org/10.1016/0006-2952(95)02184-1
Hakkola, J.; Pelkonen, O.; Pasanen, M.; Raunio, H. (1998). Xenobiotic-metabolizing cytochrome P450
enzymes in the human feto-placental unit: Role in intrauterine toxicity. Crit Rev Toxicol, 28, 35-
72. http://dx.doi.org/10.1080/104Q8449891344173
Hakkola, J.; Raunio, H.; Purkunen, R.; Pelkonen, O.; Saarikoski, S.; Cresteil, T.; Pasanen, M. (1996b).
Detection of cytochrome P450 gene expression in human placenta in first trimester of pregnancy.
Biochem Pharmacol, 52, 379-383. http://dx.doi.org/10.1016/0006-2952(96)00216-X
Hallstrom, C. (1983). Which hypnotic - if any? , 30/3 (188-192).
Hanefeld, M.; Kemmer, C.; Kadner, E. (1983). Relationship between morphological changes and lipid-
lowering action of p-chlorphenoxyisobutyric acid (CPIB) on hepatic mitochondria and
peroxisomes in man. Atherosclerosis, 46, 239-246. http://dx.doi.org/10.1016/0Q21-
9150(83)90115-6
Hanefeld, M.; Kemmer, C.; Leonhardt, W.; Kunze, K. D.; Jaross, W.; Haller, H. (1980). Effects of p-
chlorophenoxyisobutyric acid (CPIB) on the human liver. Atherosclerosis, 36, 159-172.
http://www.ncbi.nlm.nih.gov/pubmed/7406947
Hanioka, N.; Jinno, H.; Takahashi, A.; Nakano, K.; Yoda, R.; Nishimura, T.; Ando, M. (1995a).
Interaction of tetrachloroethylene with rat hepatic microsomal P450-dependent monooxygenases.
Xenobiotica, 25, 151-165. http://dx.doi.org/10.3109/00498259509Q61841
Hanioka, N.; Jinno, H.; Toyo'oka, T.; Nishimura, T.; Ando, M. (1995b). Induction of rat liver drug-
metabolizing enzymes by tetrachloroethylene. Arch Environ Contam Toxicol, 28, 273-280.
http://dx.doi.org/10.1007/BF002131Q2
Hardell, L.; Eriksson, M.; Lenner, P.; Lundgren, E. (1981). Malignant lymphoma and exposure to
chemicals, especially organic solvents, chlorophenols and phenoxy acids: A case-control study.
Br J Cancer, 43, 169-176. http://www ncbi nlm nih gov/pubmed/7470379
Hardin, B. D.; Bond, G. P.; Sikov, M. R.; Andrew, F. D.; Beliles, R. P.; Niemeier, R. W. (1981). Testing
of selected workplace chemicals for teratogenic potential. Scand J Work Environ Health, 7, 66-
75. http://www.ncbi.nlm.nih.gov/pubmed/7330632
Harrington-Brock, K.; Doerr, C. L.; Moore, M. M. (1998). Mutagenicity of three disinfection by-
products: Di- and trichloroacetic acid and chloral hydrate in L5178Y(+/-) --3.7.2C mouse
lymphoma cells. Mutat Res Genet Toxicol Environ Mutagen, 413, 265-276.
http: //www .ncbi .nlm .nih. gov/pubmed/9651541
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-26 DRAFT—DO NOT CITE OR QUOTE
-------
Harrington, J. M.; Whitby, H.; Gray, C. N.; Reid, F. J.; Aw, T. C.; Waterhouse, J. A. (1989). Renal
disease and occupational exposure to organic solvents: A case referent approach. Br J Ind Med,
46, 643-650. http://dx.doi.Org/10.l 136/oem.46.9.643
Hartmann, A. and Speit, G. (1995). Genotoxic effects of chemicals in the single cell gel (SCG) test with
human blood cells in relation to the induction of sister-chromatid exchanges (SCE). Mutat Res
Lett, 346, 49-56. http://dx.doi.org/10.1016/0165-7992(95)90068-3
Hasmall, S.; West, D.; Olsen, K.; Roberts, R. (2000). Role of hepatic non-parenchymal cells in the
response of rat hepatocytes to the peroxisome proliferator nafenopin in vitro. Carcinogenesis, 21,
2159-2165. http://www.ncbi.nlm.nih.g0v/pubmed/l 1133804
Hassoun, E. A. and Dey, S. (2008). Dichloroacetate- and trichloroacetate-induced phagocytic activation
and production of oxidative stress in the hepatic tissues of mice after acute exposure. J Biochem
Mol Toxicol, 22, 27-34. http://dx.doi.org/10.1002/ibt.2021Q
Hattis, D.; White, P.; Marmorstein, L.; Koch, P. (1990). Uncertainties in pharmacokinetic modeling for
perchloroethylene. I. Comparison of model structure, parameters, and predictions for low-dose
metabolism rates for models derived by different authors. Risk Anal, 10, 449-458.
http://dx.doi.org/10.1111/i. 1539-6924.1990.tb00528.x
Haworth, S.; Lawlor, T.; Mortelmans, K.; Speck, W.; Zeiger, E. (1983). Salmonella mutagenicity test
results for 250 chemicals. Environ Mutagen, 5, 3-142. http://dx.doi.org/10.1002/em.28600507Q3
Hayes, J. R.; Condie, L. W. J.; Borzelleca, J. F. (1986). The subchronic toxicity of tetrachloroethylene
(perchloroethylene) administered in the drinking water of rats. Fundam Appl Toxicol, 7, 119-125.
http://www.ncbi.nlm.nih.gov/pubmed/3732662
Hegi, M. E.; Fox, T. R.; Belinsky, S. A.; Devereux, T. R.; Anderson, M. W. (1993). Analysis of activated
protooncogenes in B6C3F1 mouse liver tumors induced by ciprofibrate, a potent peroxisome
proliferator. Carcinogenesis, 14, 145-149. http://www.ncbi.nlm.nih.gov/pubmed/8425263
Heineman, E. F.; Cocco, P.; Gomez, M. R.; Dosemeci, M.; Stewart, P. A.; Hayes, R. B., . . . Blair, A.
(1994). Occupational exposure to chlorinated aliphatic hydrocarbons and risk of astrocytic brain
cancer. Am J Ind Med, 26, 155-169. http://dx.doi.org/10.1002/aiim.47002602Q3
Henschler, D. (1977). Metabolism and mutagenicity of halogenated olefins - A comparison of structure
and activity. Environ Health Perspect, 21, 61-64. http://www.ncbi.nlm.nih.gov/pubmed/348459
Henschler, D. and Bonse, G. (1977). Metabolic activation of chlorinated ethylenes: Dependence of
mutagenic effect on electrophilic reactivity of the metabolically formed epoxides. Arch Toxicol,
39, 7-12. http://dx.doi.org/10.1007/BF0034327Q
Hernberg, S.; Kauppinen, T.; Riala, R.; Korkala, M. L.; Asikainen, U. (1988). Increased risk for primary
liver cancer among women exposed to solvents. Scand J Work Environ Health, 14, 356-365.
http: //www .ncbi .nlm .nih. gov/pubmed/3212412
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. http://dx.doi.org/10.1007/BF0Q378517
Herren-Freund, S. L.; Pereira, M. A.; Khoury, M. D.; Olson, G. (1987). The carcinogenicity of
trichloroethylene and its metabolites, trichloroacetic acid and dichloroacetic acid, in mouse liver.
Toxicol Appl Pharmacol, 90, 183-189. http://dx.doi.org/10.1016/0041-008X(87)90325-5
Hill, A. B. (1965). The environment and disease: Association or causation? Proc R Soc Med, 58, 295-300.
http: //www .ncbi .nlm .nih. gov/pubmed/14283 879
Hinchman, C. A. and Ballatori, N. (1990). Glutathione-degrading capacities of liver and kidney in
different species. Biochem Pharmacol, 40, 1131-1135. http://dx.doi.org/10.1016/00Q6-
2952(90)90503-D
Hinnen, U.; Schmid-Grendelmeier, P.; Muller, E.; Eisner, P. (1995). Exposure to solvents in scleroderma:
Disseminated circumscribed scleroderma (morphea) in a painter exposed to perchloroethylene.
Schweizerische medizinische Wochenschrift, 125, 2433-2437.
http: //www .ncbi .nlm .nih. gov/pubmed/8553031
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-27 DRAFT—DO NOT CITE OR QUOTE
-------
Hobara, T.; Kobayashi, H.; Kawamoto, T.; Sato, T.; Iwamoto, S.; Hirota, S.; Sakai, T. (1986). Biliary
excretion of trichloroethylene and its metabolites in dogs. Toxicol Lett, 32, 119-122.
http: //www .ncbi .nlm .nih. gov/pubmed/3 73 8923
Holoshitz, N.; Alsheikh-Ali, A. A.; Karas, R. H. (2008). Relative safety of gemfibrozil and fenofibrate in
the absence of concomitant cerivastatin use. Am J Cardiol, 101, 95-97.
http://dx.doi.Org/10.1016/i.amicard.2007.07.057
Holson, R. R. and Pearce, B. (1992). Principles and pitfalls in the analysis of prenatal treatment effects in
multiparous species. Neurotoxicol Teratol, 14, 221-228. http://dx.doi.org/10.1016/Q892-
0362(92)90020-B
Honma, T.; Hasegawa, H.; Sato, M.; Sudo, A. (1980a). Changes of free amino acid content in rat brain
after exposure to trichloroethylene and tetrachloroethylene. Ind Health, 18, 1-7.
http://dx.doi.org/10.2486/indhealth. 18.1
Honma, T.; Sudo, A.; Miyagawa, M.; Sato, M.; Hasegawa, H. (1980b). Effects of exposure to
trichloroethylene and tetrachloroethylene on the contents of acetylcholine, dopamine,
norepinephrine and serotonin in rat brain. Ind Health, 18, 171-178.
http: //dx .doi.org/10.2486/indhe alth .18.171
Horvath, M. and Frantik, E. (1973). To the relative sensitivity of nervous functions and behaviour to
nonspecific effects of foreign substances. Homeost Health Dis, 15, 25-27.
http://www.ncbi.nlm.nih.gov/pubmed/469809Q
Hoshida, Y.; Toffanin, S.; Lachenmayer, A.; Villanueva, A.; Minguez, B.; Llovet, J. M. (2010).
Molecular classification and novel targets in hepatocellular carcinoma: recent advancements.
Semin Liver Dis, 30, 35-51. http://dx.doi.org/10.1055/s-003Q-1247131
Houten, L. and Sonnesso, G. (1980). Occupational exposure and cancer of the liver. Arch Environ Health,
35, 51-53. http://www.ncbi.nlm.nih.gov/pubmed/7362270
Hu, Y.; Hakkola, J.; Oscarson, M.; Ingelman-Sundberg, M. (1999). Structural and functional
characterization of the 5'-flanking region of the rat and human cytochrome P450 2E1 genes:
Identification of a polymorphic repeat in the human gene. Biochem Biophys Res Commun, 263,
286-293. http://dx.doi.org/10.1006/bbrc.1999.1362
Hudnell, H. K.; Boyes, W. K.; Otto, D. A.; House, D. E.; Creason, J. P.; Geller, A. M., . . . Broadwell, D.
K. (1996a). Battery of neurobehavioral tests recommended to ATSDR: Solvent-induced deficits
in microelectronic workers. Toxicol Ind Health, 12, 235-243.
http: //www .ncbi .nlm .nih. gov/pubmed/8794536
Hudnell, H. K.; House, D.; Schmid, J.; Koltai, D.; Stopford, W.; Wilkins, J., . . . Music, S. (2001). Human
visual function in the North Carolina clinical study on possible estuary-associated syndrome. J
Toxicol Environ Health A, 62, 575-594. http://dx.doi.org/10.1080/15287390151Q79633
Hudnell, H. K.; Otto, D. A.; House, D. E. (1996b). The influence of vision on computerized
neurobehavioral test scores: A proposal for improving test protocols. Neurotoxicol Teratol, 18,
391-400. http://dx.doi.org/10.1016/0892-0362(96)00040-2
Hudnell, H. K. and Schreiber, J. S. (2004). Residential tetrachloroethylene exposure: Response. Environ
Health Perspect, 112, A864-A865.
Hurst, C. H. and Waxman, D. J. (2003). Activation of PPARalpha and PPARgamma by environmental
phthalate monoesters. Toxicol Sci, 74, 297-308. http://dx.doi.org/10.1093/toxsci/kfgl45
Huttunen, J.; Heinonen, O.; Manninen, V.; Koskinen, P.; Hakulinen, T.; Teppo, L., . . . Frick, M. (1994).
The Helsinki Heart Study: An 8.5-year safety and mortality follow-up. J Intern Med, 235, 31-39.
http: //www .ncbi.nlm.nih.gov/pubmed/8283157
IARC. (International Agency for Research on Cancer). (1995). Tetrachloroethylene. Lyon, France:
Author. Retrieved from http://monographs.iarc.fr/ENG/Monographs/vol63/index.php.
IARC. (International Agency for Research on Cancer). (2004). Chloral and Chloral Hydrate. Lyon,
France: Author. Retrieved from http://monographs.iarc.fr/ENG/Monographs/vol84/index.php.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-28 DRAFT—DO NOT CITE OR QUOTE
-------
Ikeda, M. and Imamura, T. (1973). Biological half-life of trichloroethylene and tetrachloroethylene in
human subjects. Int Arch Occup Environ Health, 31, 209-224.
http://www.ncbi.nlm.nih.gov/pubmed/4593976
Ikeda, M.; Koizumi, A.; Watanabe, T.; Endo, A.; Sato, K. (1980). Cytogenetic and cytokinetic
investigations on lymphocytes from workers occupationally exposed to tetrachloroethylene.
Toxicol Lett, 5, 251-256. http://dx.doi.org/10.1016/0378-4274(80)90068-5
Ikeda, M. and Ohtsuji, H. (1972). A comparative study of the excretion of Fujiwara reaction-positive
substances in urine of humans and rodents given trichloro- or tetrachloro-derivatives of ethane
and ethylene. Occup Environ Med, 29, 99-104. http://www.ncbi.nlm.nih.gov/pubmed/5060252
Ikeda, M.; Ohtsuji, 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, 328-333. http://dx.doi.Org/10.l 136/oem.29.3.328
ILSI. (International Life Sciences Institute). (1992). Similarities and differences between children and
adults: implications for risk assessment. ILSI Press.
Imbriani, M.; Ghittori, S.; Pezzagno, G.; Capodaglio, E. (1988). Urinary excretion of tetrachloroethylene
(perchloroethylene) in experimental and occupational exposure. Arch Environ Occup Health, 43,
292-298.
. Immune effects of trichloroethylene on autoimmune disease in mice. (2004). In L. C. Mohr, D. G. Hoel
& D. Jollow (Eds.), Trichloroethylene : The scientific basis of risk assessment (pp. 87-98).
Charleston, SC: The Medical University of South Carolina Press.
Infante-Rivard, C.; Siemiatycki, J.; Lakhani, R.; Nadon, L. (2005). Maternal exposure to occupational
solvents and childhood leukemia. Environ Health Perspect, 113, 787-792.
http://dx.doi.org/10.1289/ehp.7707
IOM. (Institute of Medicine). (2002). Gulf War and health. Vol. 2: Insecticides and solvents.
Washington, DC: National Academies Press.
Ishmael, J. and Dugard, P. H. (2006). A review of perchloroethylene and rat mononuclear cell leukemia.
Regul Toxicol Pharmacol, 45, 178-184. http://dx.doi.Org/10.1016/i.vrtph.2006.02.009
Ishmael, J. and Lock, E. A. (1986). Nephrotoxicity of hexachlorobutadiene and its glutathione-derived
conjugates. Toxicol Pathol, 14, 258-262. http://www.ncbi.nlm.nih.gov/pubmed/3764322
Issemann, I. and Green, S. (1990). Activation of a member of the steroid hormone receptor superfamily
by peroxisome proliferators. Nature, 347, 645-650.
Ito, Y.; Yamanoshita, O.; Asaeda, N.; Tagawa, Y.; Lee, C. H.; Aoyama, T., . . . Nakajima, T. (2007).
Di(2-ethylhexyl)phthalate induces hepatic tumorigenesis through a peroxisome proliferator-
activated receptor alpha-independent pathway. J Occup Health, 49, 172-182.
http: //www .ncbi .nlm .nih. gov/pubmed/17575397
Jakobson, I.; Wahlberg, J. E.; Holmberg, B.; Johansson, G. (1982). Uptake via the blood and elimination
of 10 organic solvents following epicutaneous exposure of anesthetized guinea pigs. Toxicol
Appl Pharmacol, 63, 181-187. http://dx.doi.org/10.1016/0041-008X(82)90039-4
Jang, J. Y. and Droz, P. O. (1997). Ethnic differences in biological monitoring of several organic
solvents: II. A simulation study with a physiologically based pharmacokinetic model. Int Arch
Occup Environ Health, 70, 41-50. http://dx.doi.org/10.1007/s00420005Q184
Janulewicz, P. A.; White, R. F.; Winter, M. R.; Weinberg, J. M.; Gallagher, L. E.; Vieira, V., . . .
Aschengrau, A. (2008). Risk of learning and behavioral disorders following prenatal and early
postnatal exposure to tetrachloroethylene (PCE)-contaminated drinking water. Neurotoxicol
Teratol, 30, 175-185. http://dx.doi.Org/10.1016/i.ntt.2008.01.007
Ji, J.; Granstrom, C.; Hemminki, K. (2005a). Occupation and bladder cancer: a cohort study in Sweden.
Br J Cancer, 92, 1276-1278. http://dx.doi.org/10.1038/si.bic.6602473
Ji, J.; Granstrom, C.; Hemminki, K. (2005b). Occupational risk factors for kidney cancer: a cohort study
in Sweden. World Journal of Urology, 23, 271-278. http://dx.doi.org/10.1007/s00345-005-00Q7-5
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-29 DRAFT—DO NOT CITE OR QUOTE
-------
Ji, J. and Hemminki, K. (2005a). Occupation and upper aerodigestive tract cancers: A follow-up study in
Sweden. J Occup Environ Med, 47, 785-795.
http://dx.doi.org/10.1097/01.iom.000Q165798.28569.b5
Ji, J. and Hemminki, K. (2005b). Occurrences of leukemia subtypes by socioeconomic and occupational
groups in Sweden. J Occup Environ Med, 47, 1131-1140.
http://dx.doi.org/10.1097/01.iom.00001743Q2.63621.e8
Ji, J. and Hemminki, K. (2005c). 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. http://dx.doi.org/10.1007/s00420-005-0Q15-l
Ji, J. and Hemminki, K. (2006). Socioeconomic/occupational risk factors for lymphoproliferative diseases
in Sweden. Ann Epidemiol, 16, 370-376. http://dx.doi.Org/10.1016/i.annepidem.2005.09.002
JISA. (Japan Industrial Safety Association). (1993). Carcinogenicity study of tetrachloroethylene by
inhalation in rats and mice. Hadano, Japan: Author.
Johansen, K.; Tinnerberg, H.; Lynge, E. (2005). Use of history science methods in exposure assessment
for occupational health studies. Occup Environ Med, 62, 434-441.
http://dx.doi.Org/10.l 136/oem.2004.016493
Johnson, P. D.; Dawson, B. V.; Goldberg, S. J. (1998). Cardiac teratogenicity of trichloroethylene
metabolites. J Am Coll Cardiol, 32, 540-545. http://dx.doi.org/10.1016/S0735-1097(98)00232-0
Jonker, D.; Woutersen, R. A.; Feron, V. J. (1996). Toxicity of mixtures of nephrotoxicants with similar or
dissimilar mode of action. Food Chem Toxicol, 34, 1075-1082. http://dx.doi.org/10.1016/SQ278-
6915(97)00077-X
Juntunen, J. (1986). Occupational toxicology of trichloroethylene with special reference to neurotoxicity.
In P. L. Chambers, P. Gehring & F. Sakai (Eds.), New concepts and developments in toxicology
(pp. 189-200). Amsterdam, Netherlands: Elsevier Science Publishers.
Kacew, S. and Lambert, G. H. (1997). Environmental toxicology and pharmacology of human
development. Washington, DC: Taylor and Francis.
Kaemmerer, K.; Fink, J.; Keitzmann, M. (1982). Studies on the pharmacodynamics of perchloroethylene.
Prakt Tierarzt, 63, 171-182.
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.
http://dx.doi.org/10.1016/0165-1161(86)90006-3
Kargalioglu, Y.; McMillan, B. J.; Minear, R. A.; Plewa, M. J. (2002). Analysis of the cytotoxicity and
mutagenicity of drinking water disinfection by-products in Salmonella typhimurium. Teratog
Carcinog Mutagen, 22, 113-128. http://dx.doi.org/10.1002/tcm. 10010
Karlsson, J. E.; Rosengren, L. E.; Kjellstrand, P.; Haglid, K. G. (1987). Effects of low-dose inhalation of
three chlorinated aliphatic organic solvents on deoxyribonucleic acid in gerbil brain. Scand J
Work Environ Health, 13, 453-458. http://www nchi nlm nih gov/pubmed/3433047
Kato, I.; Koenig, K. L.; Watanabe-Meserve, H.; Baptiste, M. S.; Lillquist, P. P.; Frizzera, G., . . . Shore,
R. E. (2005). Personal and occupational exposure to organic solvents and risk of non-Hodgkin's
lymphoma (NHL) in women (United States). Cancer Causes Control, 16, 1215-1224.
Kaufman, D. S.; Shipley, W. U.; Feldman, A. S. (2009). Bladder cancer. Lancet, 374, 239-249.
http://dx.doi.0rg/lO.lOl6/SO 140-6736(09)60491-8
Kauppinen, T.; Heikkila, P.; Plato, N.; Woldbaek, T.; Lenvik, K.; Hansen, J., . . . Pukkala, E. (2009).
Construction of job-exposure matrices for the Nordic Occupational Cancer Study (NOCCA).
Acta Oncol, 48, 791-800. http://dx.doi.org/10.1080/028418609Q2718747
Kawata, K.; Shimazaki, R.; Okabe, S. (2009). Comparison of gene expression profiles in HepG2 cells
exposed to arsenic, cadmium, nickel, and three model carcinogens for investigating the
mechanisms of metal carcinogenesis. Environ Mol Mutagen, 50, 46-59.
http://dx.doi.org/10.1002/em.20438
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-30 DRAFT—DO NOT CITE OR QUOTE
-------
Keech, A.; Simes, R. J.; Barter, P.; Best, J.; Scott, R.; Taskinen, M. R., . . . investigators, F. s. (2005).
Effects of long-term fenofibrate therapy on cardiovascular events in 9795 people with type 2
diabetes mellitus (the FIELD study): randomised controlled trial. Lancet, 366, 1849-1861.
http://dx.doi.org/10.1016/SQ 140-6736(05)67667-2
Keller, D. A. and HdA, H. (1988). Mechanistic studies on chloral toxicity: Relationship to
trichloroethylene carcinogenesis. Toxicol Lett, 42, 183-191. http://dx.doi.org/10.1016/Q378-
4274(88)90076-8
Kernan, G. J.; Ji, B. T.; Dosemeci, M.; Silverman, D. T.; Balbus, J.; Zahm, S. H. (1999). Occupational
risk factors for pancreatic cancer: A case-control study based on death certificates from 24 U.S.
states. Am J Ind Med, 36, 260-270. http://dx.doi.org/10.1002/(SICI) 1097-
0274(199908)36:2<260::AID-AJIM5>3.0.CQ;2-P
Kessler, W.; Numtip, W.; Grote, K.; Csanady, G. A.; Chahoud, I.; Filser, J. G. (2004). Blood burden of
di(2-ethylhexyl) phthalate and its primary metabolite mono(2-ethylhexyl) phthalate in pregnant
and nonpregnant rats and marmosets. Toxicol Appl Pharmacol, 195, 142-153.
http://dx.doi.Org/10.1016/i.taap.2003.ll.014
Kezic, S.; Monster, A.; Kriise, J.; Verberk, M. (2000). Skin absorption of some vaporous solvents in
volunteers. Int Arch Occup Environ Health, 73, 415-422.
http://www.ncbi.nlm.nih.gov/pubmed/11007346
Khan, M.; Wu, X.; Ansari, G. (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. http://dx.doi.org/10.1006/taap.2000.9Q86
Khanov, M.; Kurmukov, A.; Sultanov, M.; Akhmedkhodzhaeva, K. (1968). [Effect of vincarin on the
central nervous system], Farmakol Toksikol, 31, 538-541.
http: //www .ncbi .nlm .nih. gov/pubmed/43 87717
Kim, D. and Ghanayem, B. I. (2005). Role Of Cytochrome P450 2E1 (CYP2E1) In Trichloroethylene
(TCE) Metabolism And Disposition: Comparative Studies Using CYP2E1-/- And Wild-Type
Mice. Toxicol Sci, 84, 148.
Kim, K. S.; Min, J. Y.; Dickman, M. B. (2008). Oxalic acid is an elicitor of plant programmed cell death
during Sclerotinia sclerotiorum disease development. Mol Plant Microbe Interact, 21, 605-612.
http://dx.doi.org/10.1094/MPMI-21-5-0605
Kim, M. Y.; Oskarsson, T.; Acharyya, S.; Nguyen, D. X.; Zhang, X. H.; Norton, L.; Massague, J. (2009).
Tumor self-seeding by circulating cancer cells. Cell, 139, 1315-1326.
http://dx.doi.Org/10.1016/i.cell.2009.ll.025
Kjellstrand, P.; Holmquist, B.; Jonsson, I.; Romare, S.; Mansson, L. (1985). Effects of organic solvents
on motor activity in mice. Toxicology, 35, 35-46. http://dx.doi.org/10.1016/030Q-
483X(85)90130-1
Kjellstrand, P.; Holmquist, B.; Kanje, M.; Aim, P.; Romare, S.; Jonsson, I., . . . Bjerkemo, M. (1984).
Perchloroethylene: Effects on body and organ weights and plasma butyrylcholinesterase activity
in mice. Acta Pharmacol Toxicol, 54, 414-424. http://dx.doi.Org/10.l 11 l/i.1600-
0773.1984.tb01951.x
Klaunig, J. E.; Babich, M. A.; Baetcke, K. P.; Cook, J. C.; Corton, J. C.; David, R. M., . . . Fenner-Crisp,
P. A. (2003). PPARalpha agonist-induced rodent tumors: Modes of action and human relevance.
Crit Rev Toxicol, 33, 655-780. http://dx.doi.org/10.1080/713608372
Kline, S. A.; McCoy, E. C.; Rosenkranz, H. S.; Van Duuren, B. L. (1982). Mutagenicity of chloroalkene
epoxides in bacterial systems. Mutat Res, 101, 115-125. http://dx.doi.org/10.1016/Q165-
1218(82)90002-7
Kline, S. A.; Solomon, J. J.; Van Duuren, B. L. (1978). Synthesis and reactions of chloroalkene epoxides.
J Org Chem, 43, 3596-3600.
Koch, R.; Schlegelmilch, R.; Wolf, H. U. (1988). Genetic effects of chlorinated ethylenes in the yeast
Saccharomyces cerevisiae. Mutat Res, 206, 209-216. http://dx.doi.org/10.1016/0165-
1218(88)90162-0
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-31 DRAFT—DO NOT CITE OR QUOTE
-------
Kogevinas, M.; 't Mannetje, A.; Cordier, S.; Ranfit, U.; Gonzalez, C.; Vineis, P., . . . Boffetta, P. (2003).
Occupation and bladder cancer among men in Western Europe. Cancer Causes Control, 14, 907-
914. http://dx.doi.Org/10.1023/B:CACO.0000007962.19066.9c
Kok, R. J.; Haas, M.; Moolenaar, F.; de Zeeuw, D.; Meijer, D. K. (1998). Drug delivery to the kidneys
and the bladder with the low molecular weight protein lysozyme. Ren Fail, 20, 211-217.
Koppel, C.; Arndt, I.; Arendt, U.; Koeppe, P. (1985). Acute tetrachloroethylene poisoning - Blood
elimination kinetics during hyperventilation therapy. Clin Toxicol, 23, 103-115.
htto://dx.doi.org/10.3109/15563658508990621
Kostrzewski, P.; Jakubowski, M.; Kolacinski, Z. (1993). Kinetics of trichloroethylene elimination from
venous blood after acute inhalation poisoning. Clin Toxicol, 31, 353-363.
http://www.ncbi.nlm.nih.gov/pubmed/8492349
Kringstad, K. P.; Ljungquist, P. O.; de Sousa, F.; Stromberg, L. M. (1981). Identification and mutagenic
properties of some chlorinated aliphatic compounds in the spent liquor from kraft pulp
chlorination. Environ Sci Technol, 15, 562-566. http://dx.doi.org/10.1021/esQ0087a006
Kung, J. W.; Currie, I. S.; Forbes, S. J.; Ross, J. A. (2010). Liver development, regeneration, and
carcinogenesis. J Biomed Biotechnol, 2010, 984248. http://dx.doi.Org/10.l 155/2010/984248
Kurata, Y.; Kidachi, F.; Yokoyama, M.; Toyota, N.; Tsuchitani, M.; Katoh, M. (1998). Subchronic
toxicity of Di(2-ethylhexyl)phthalate in common marmosets: lack of hepatic peroxisome
proliferation, testicular atrophy, or pancreatic acinar cell hyperplasia. Toxicol Sci, 42, 49-56.
Kylin, B.; Reichard, H.; Sumegi, I.; Yllner, S. (1963). Hepatotoxicity of inhaled trichloroethylene,
tetrachloroethylene and chloroform Single exposure. Basic Clin Pharmacol Toxicol, 20, 16-26.
Kylin, B.; Sumegi, I.; Yllner, S. (1965). Hepatotoxicity of inhaled trichloroethylene and
tetrachloroethylene Long-term exposure. Acta Pharmacol Toxicol, 22, 379-385.
http: //dx .doi.org/10.1111/i.l 600-0773.1965 ,tb01833 ,x
Kyrklund, T.; Ailing, C.; Kjellstrand, P.; Haglid, K. G. (1984). Chronic effects of perchloroethylene on
the composition of lipid and acyl groups in cerebral cortex and hippocampus of the gerbil.
Toxicol Lett, 22, 343-349. http://dx.doi.org/10.1016/0378-4274(84)90112-7
Kyrklund, T. and Haglid, K. (1991). Brain lipid composition in guinea pigs after intrauterine exposure to
perchloroethylene. Pharmacol Toxicol, 68, 146-148. http://dx.doi.Org/10.l 11 l/i.1600-
0773.199 l.tb02054.x
Kyrklund, T.; Kjellstrand, P.; Haglid, K. G. (1987). Lipid composition and fatty acid pattern of the gerbil
brain after exposure to perchloroethylene. Arch Toxicol, 60, 397-400.
http://dx.doi.org/10.1007/BF0Q295762
Kyrklund, T.; Kjellstrand, P.; Haglid, K. G. (1988). Effects of exposure to Freon 11, 1,1,1-trichloroethane
or perchloroethylene on the lipid and fatty-acid composition of rat cerebral cortex. Scand J Work
Environ Health, 14, 91-94.
Kyrklund, T.; Kjellstrand, P.; Haglid, K. G. (1990). Long-term exposure of rats to perchloroethylene, with
and without a post-exposure solvent-free recovery period: Effects on brain lipids. Toxicol Lett,
52, 279-285. http://dx.doi.org/10.1016/0378-4274(90)90037-M
Kyyronen, P.; Taskinen, H.; Lindbohm, M. L.; Hemminki, K.; Heinonen, O. P. (1989). Spontaneous
abortions and congenital malformations among women exposed to tetrachloroethylene in dry
cleaning. J Epidemiol Community Health, 43, 346-351. http://dx.doi.Org/10.l 136/iech.43.4.346
Lacey, J. V. J.; Garabrant, D. H.; Laing, T. J.; Gillespie, B. W.; Mayes, M. D.; Cooper, B. C.;
Schottenfeld, D. (1999). Petroleum distillate solvents as risk factors for undifferentiated
connective tissue disease (UCTD). Am J Epidemiol, 149, 761-770.
http://www.ncbi.nlm.nih.gov/pubmed/10206626
Lagakos, S. W.; Wessen, B. J.; Zelen, M. (1986). An analysis of comtaminated well water and health
effects in Woburn, Massachusetts. J Am Stat Assoc, 81, 583-596.
Lake, B. G.; Evans, J. G.; Cunninghame, M. E.; Price, R. J. (1993). Comparison of the hepatic effects of
nafenopin and WY-14,643 on peroxisome proliferation and cell replication in the rat and Syrian
hamster. Environ Health Perspect, 101, 241-247. http://www nchi nlm nih gov/pubmed/8013414
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-32 DRAFT—DO NOT CITE OR QUOTE
-------
Lane, S. E.; Watts, R. A.; Bentham, G.; Innes, N. J.; Scott, D. G. I. (2003). Are environmental factors
important in primary systemic vasculitis? A case-control study. Arthritis Rheum, 48, 814-823.
http://dx.doi.org/10.1002/art.1083Q
Lapis, K.; Zalatnai, A.; Timar, F.; Thorgeirsson, U. P. (1995). Quantitative evaluation of lysozyme- and
CD68-positive Kupffer cells in diethylnitrosamine-induced hepatocellular carcinomas in
monkeys. Carcinogenesis, 16, 3083-3085. htto://www ncbi nlm nih gov/pubmed/8603489
Larsen, G. L. (1985). Distribution of cysteine conjugate beta-lyase in gastrointestinal bacteria and in the
environment. Xenobiotica, 15, 199-209. http://dx.doi.org/10.3109/0049825850904535Q
Larsen, G. L. and Stevens, J. L. (1986). Cysteine conjugate beta-lyase in the gastrointestinal bacterium
Eubacterium limosum. Mol Pharmacol, 29, 97-103.
http: //www .ncbi .nlm .nih. gov/pubmed/3 945231
Larsen, N. L.; Nielsen, B.; Rayn-Nielsen, A. (1977). Perchloroethylene intoxication. A hazard in the use
of coin laundries. Ugeskr Laeger, 139, 270-275.
Lash, L. H.; Fisher, J. W.; Lipscomb, J. C.; Parker, J. C. (2000). Metabolism oftrichloroethylene. Environ
Health Perspect, 108, 177-200.
Lash, L. H.; Lipscomb, J. C.; Putt, D. A.; Parker, J. C. (1999). Glutathione conjugation of
trichloroethylene in human liver and kidney: Kinetics and individual variation. Drug Metab
Dispos, 27, 351-359. http ://www.ncbi .nlm.nih. gov/pubmed/100645 65
Lash, L. H.; Nelson, R. M.; Van Dyke, R. A.; Anders, M. W. (1990). Purification and characterization of
human kidney cytosolic cysteine conjugate beta-lyase activity. Drug Metab Dispos, 18, 50-54.
http: //www .ncbi .nlm .nih. gov/pubmed/2139845
Lash, L. H. and Parker, J. C. (2001). Hepatic and renal toxicities associated with perchloroethylene.
Pharmacol Rev, 53, 177-208. http://www ncbi nlm nih gov/pubmed/11356983
Lash, L. H.; Putt, D. A.; Huang, P.; Hueni, S. E.; Parker, J. C. (2007). Modulation of hepatic and renal
metabolism and toxicity of trichloroethylene and perchloroethylene by alterations in status of
cytochrome P450 and glutathione. Toxicology, 235, 11-26.
http://dx.doi.Org/10.1016/i.tox.2007.03.001
Lash, L. H.; Qian, W.; Putt, D. A.; Desai, K.; Elfarra, A. A.; Sicuri, A. R.; Parker, J. C. (1998).
Glutathione conjugation of perchloroethylene in rats and mice in vitro: Sex-, species-, and tissue-
dependent differences. Toxicol Appl Pharmacol, 150, 49-57.
http://dx.doi.org/10.1006/taap.1998.8402
Lash, L. H.; Qian, W.; Putt, D. A.; Hueni, S. E.; Elfarra, A. A.; Krause, R. J.; Parker, J. C. (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, 155-164.
http://www.ncbi.nlm.nih.gov/pubmed/11259540
Lash, L. H.; Qian, W.; Putt, D. A.; Hueni, S. E.; Elfarra, A. A.; Sicuri, A. R.; Parker, J. C. (2002). Renal
toxicity of perchloroethylene and S-(l,2,2-trichlorovinyl)glutathione in rats and mice: sex- and
species-dependent differences. Toxicol Appl Pharmacol, 179, 163-171.
http://dx.doi.org/10.1006/taap.2Q01.9358
Laslo-Baker, D.; Barrera, M.; Knittel-Keren, D.; Kozer, E.; Wolpin, J.; Khattak, S., . . . Koren, G. (2004).
Child neurodevelopmental outcome and maternal occupational exposure to solvents. Arch Pediatr
Adolesc Med, 158, 956-961. http://dx.doi.org/10.1001/archpedi.158.10.956
Laughter, A. R.; Dunn, C. S.; Swanson, C. L.; Howroyd, P.; Cattley, R. C.; Corton, J. C. (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. http://dx.doi.Org/10.1016/i.tox.2004.06.014
Lauwerys, R.; Herbrand, J.; Buchet, J. P.; Bernard, A.; Gaussin, J. (1983). Health surveillance of workers
exposed to tetrachloroethylene in dry-cleaning shops. Int Arch Occup Environ Health, 52, 69-77.
http://dx.doi.org/10.1007/BF003806Q9
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-33 DRAFT—DO NOT CITE OR QUOTE
-------
Leavitt, S. A.; DeAngelo, A. B.; George, M. H.; Ross, J. A. (1997). Assessment of the mutagenicity of
dichloroacetic acid in lacl transgenic B6C3F1 mouse liver. Carcinogenesis, 18, 2101-2106.
http ://www .ncbi .nlm .nih. gov/pubmed/93 95208
Lee, L.; Chung, C.; Ma, Y.; Wang, G.; Chen, P.; Hwang, Y.; Wang, J. (2003). Increased mortality odds
ratio of male liver cancer in a community contaminated by chlorinated hydrocarbons in
groundwater. Occup Environ Med, 60, 364-369. http://dx.doi.Org/10.l 136/oem.60.5.364
Lee, S. S.; Pineau, T.; Drago, J.; Lee, E. J.; Owens, J. W.; Kroetz, D. L., . . . Gonzalez, F. J. (1995).
Targeted disruption of the alpha isoform of the peroxisome proliferator-activated receptor gene in
mice results in abolishment of the pleiotropic effects of peroxisome proliferators. Mol Cell Biol,
15, 3012-3022. http://www nchi nlm nih gov/pubmed/7539101
Lehmann, I.; Rehwagen, M.; Diez, U.; Seiffart, A.; Rolle-Kampczyk, U.; Richter, M., . . . Herbarth, O.
(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. http://dx.doi.org/10.1078/1438-4639-0010Q
Lehmann, I.; Thoelke, A.; Rehwagen, M.; Rolle-Kampczyk, U.; Schlink, U.; Schulz, R., . . . Herbarth, O.
(2002). The influence of maternal exposure to volatile organic compounds on the cytokine
secretion profile of neonatal T cells. Environ Toxicol, 17, 203-210.
http://dx.doi.org/10.1002/tox.10Q55
Leibman, K. C. and Ortiz, E. (1970). Epoxide intermediates in microsomal oxidation of olefins to glycols.
J Pharmacol Exp Ther, 173, 242-246.
Leibman, K. C. and Ortiz, E. (1977). Metabolism of halogenated ethylenes. Environ Health Perspect, 21,
91-97. http://www.ncbi.nlm.nih.gov/pubmed/612463
Lemen, R. (2003). RE: Scrutinizing ACGIH risk assessments: the trichloroethylene case-Ruden C., 2003.
Am J Ind Med 44:207-213. Am J Ind Med, 44, 560. http://dx.doi.org/10.1002/aiim.103Q2
Letkiewicz, F.; Johnston, P.; Macaluso, C.; Elder, R.; Yu, W. (1982). Occurrence in tetrachloroethylene
(perchloroethylene) in drinking water, food and air. (JRB28130385229). McLean, VA: JRB
Associates, Inc.
[Letter from Scot Eustis, National Toxicology Program, to William Farland, Director, Office of Health
and Environmental Assessment, U.S. EPA] (1988).
Levine, B.; Fierro, M. F.; Goza, S. W.; Valentour, J. C. (1981). A tetrachloroethylene fatality. J Forensic
Sci, 26, 206-209. http://dx.doi.org/10.1520/JFS1135QJ
Liang, J. C. and Brinkley, B. R. (1985). Chemical probes and possible targets for the induction of
aneuploidy. Basic Life Sci, 36, 491-505. http://www.ncbi.nlm.nih.gov/pubmed/4096703
Liang, J. C. and Pacchierotti, F. (1988). Cytogenetic investigation of chemically-induced aneuploidy in
mouse spermatocytes. Mutat Res-Fundam Mol Mech Mutagen, 201, 325-335.
http://dx.doi.org/10.1016/0027-5107(88)90021-8
Lide, D. R. (Ed.). (1990). CRC Handbook of Chemistry and Physics (Vol. 71st). Boca Raton, FL: CRC
Press.
Lieber, C. S. (1997). Cytochrome P-4502E1: Its physiological and pathological role. Physiol Rev, 77,
517-544.
Lillford, L.; Beevers, C.; Bowen, D.; Kirkland, D. (2010). RE: DNA damage detected by the alkaline
comet assay in the liver of mice after oral administration of tetrachloroethylene. (Mutagenesis,
25, 133-138, 2010). Mutagenesis, 25, 427-428. http://dx.doi.org/10.1093/mutage/gea020
Lin, R. S. and Kessler, 1.1. (1981). A multifactorial model for pancreatic cancer in man: epidemiologic
evidence. JAMA, 245, 147-152. http://www.ncbi.nlm.nih.gov/pubmed/7452829
Lindbohm, M.-L.; Hemminki, K.; Bonhomme, M. G.; Anttila, A.; Rantala, K.; Heikkila, P.; Rosenberg,
M. J. (1991). Effects of paternal occupational exposure on spontaneous abortions. Am J Public
Health, 81, 1029-1033.
Lindbohm, M. L. (1995). Effects of parental exposure to solvents on pregnancy outcome. J Occup
Environ Med, 37, 908-914. http://www ncbi nlm nih gov/pubmed/8520952
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-34 DRAFT—DO NOT CITE OR QUOTE
-------
Lindbohm, M. L. and Hemminki, K. (1988). Nationwide database on medically diagnosed spontaneous
abortions in Finland. Int J Epidemiol, 17, 568-573. http://dx.doi.Org/10.1093/iie/17.3.568
Lindbohm, M. L.; Taskinen, H.; Sallmen, M.; Hemminki, K. (1990). Spontaneous abortions among
women exposed to organic solvents. Am J Ind Med, 17, 449-463.
http://dx.doi.org/10.1002/aiim.47001704Q4
Lipscomb, J. C.; Fisher, J. W.; Confer, P. D.; Byczkowski, J. Z. (1998). In vitro to in vivo extrapolation
for trichloroethylene metabolism in humans. Toxicol Appl Pharmacol, 152, 376-387.
http://dx.doi.org/10.1006/taap.1998.8485
Little, J.; Satlin, A.; Sunderland, T.; Volicer, L. (1995). Sundown syndrome in severely demented patients
with probable Alzheimer's disease. J Geriatr Psychiatry Neurol, 8, 103-106.
http: //www .ncbi .nlm .nih. gov/pubmed/7794472
Liu, Y.; Guyton, K. Z.; Gorospe, M.; Xu, Q.; Lee, J. C.; Holbrook, N. J. (1996). Differential activation of
ERK, JNK/SAPK and P38/CSBP/RK map kinase family members during the cellular response to
arsenite. Free Radic Biol Med, 21, 771-781. http://www.ncbi.nlm.nih.gov/pubmed/8902523
Lock, E. A.; Sani, Y.; Moore, R. B.; Finkelstein, M. B.; Anders, M. W.; Seawright, A. A. (1996). Bone
marrow and renal injury associated with haloalkene cysteine conjugates in calves. Arch Toxicol,
70, 607-619. http://dx.doi.org/10.1007/s00204005Q319
Loizou, G. D. (2001). The application of physiologically based pharmacokinetic modelling in the analysis
of occupational exposure to perchloroethylene. Toxicol Lett, 124, 59-69.
http://dx.doi.org/10.1016/S0378-4274(W)00283-6
Lovell, D. (2010). Is tetrachloroethylene genotoxic or not. Mutagenesis, 25, 443-446.
http://dx.doi.org/10.1093/mutage/geq036
Lowengart, R. A.; Peters, J. M.; 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, 39-46. http://www.ncbi.nlm.nih.gov/pubmed/3474448
Lukaszewski, T. (1979). Acute tetrachloroethylene fatality. Clin Toxicol, 15, 411-415.
http://dx.doi.org/10.3109/15563657908989895
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. http://dx.doi.org/10.3109/03009749409Q99278
Lundgren, B.; Meijer, J.; DePierre, J. W. (1987). Induction of cytosolic and microsomal epoxide
hydrolases and proliferation of peroxisomes and mitochondria in mouse liver after dietary
exposure to p-chlorophenoxyacetic acid, 2,4-dichlorophenoxyacetic acid and 2,4,5-
trichlorophenoxyacetic acid. Biochem Pharmacol, 36, 815-821.
http: //www .ncbi .nlm .nih. gov/pubmed/3 032197
Luster, M. I.; Simeonova, P. P.; Gallucci, R. M.; Bruccoleri, A.; Blazka, M. E.; Yucesoy, B. (2001). Role
of inflammation in chemical-induced hepatotoxicity. Toxicol Lett, 120, 317-321.
http://www.ncbi.nlm.nih.gOv/pubmed/l 1323190
Lybarger, J. A.; Lichtveld, M. Y.; Amler, R. W. (1999). Biomedical testing of the kidney for persons
exposed to hazardous substances in the environment. Ren Fail, 21, 263-274.
Lynge, E. (2008). [Comments on toxicological review of tetrachloroethylene CAS no. 127-18-4
EPA/635/R-08/011A external draft review].
Lynge, E.; Andersen, A.; Rylander, L.; Tinnerberg, H.; Lindbohm, M. L.; Pukkala, E., . . . Johansen, K.
(2006). Cancer in persons working in dry cleaning in the Nordic countries. Environ Health
Perspect, 114, 213-219. http://dx.doi.org/10.1289/ehp.8425
Lynge, E.; Carstensen, B.; Andersen, O. (1995). Primary liver cancer and renal cell carcinoma in laundry
and dry-cleaning workers in Denmark. Scand J Work Environ Health, 21, 293-295.
http: //www .ncbi .nlm .nih. gov/pubmed/8553005
Lynge, E. and Thygesen, L. (1990). Primary liver cancer among women in laundry and dry-cleaning work
in Denmark. Scand J Work Environ Health, 16, 108-112.
http: //www .ncbi .nlm .nih. gov/pubmed/235 3193
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-35 DRAFT—DO NOT CITE OR QUOTE
-------
Lynge, E.; Tinnerberg, H.; Rylander, L.; Romundstad, P.; Johansen, K.; Lindbohm, M. L., . . . Thorsted,
B. L. (2011). Exposure to tetrachloroethylene in dry cleaning shops in the nordic countries. Ann
Occup Hyg, 55, 387-396. http://dx.doi.org/10.1093/annhyg/meql01
Ma, J.; Lessner, L.; Schreiber, J.; Carpenter, D. O. (2009). Association between residential proximity to
PERC dry cleaning establishments and kidney cancer in New York City. J Environ Public Health,
2009, 183920. http://dx.doi.org/10.1155/2009/183920
MacArthur, A.; Le, N.; Fang, R.; Band, P. (2009). Identification of occupational cancer risk in British
Columbia: A population-based case-control study of 2,998 lung cancers by histopathological
subtype. Am J Ind Med, 52, 221-232. http://dx.doi.org/10.1002/aiim.2Q663
Macchiarini, P. (2006). Primary tracheal tumours. Lancet Oncol, 7, 83-91.
http://dx.doi.Org/10.1016/S 1470-2045(05)70541-6
Mackay, J. M.; Fox, V.; Griffiths, K.; Fox, D. A.; Howard, C. A.; Coutts, C., . . . Styles, J. A. (1995).
Trichloroacetic acid: Investigation into the mechanism of chromosomal damage in the in virto
human lymphocyte cytogenetic assay and the mouse bone marrow micronucleus test.
Carcinogenesis, 16, 1127-1133. http://dx.doi.org/10.1093/carcin/16.5.1127
Maddalena, R. L.; McKone, T. E.; Layton, D. W.; Hsieh, D. P. (1995). Comparison of multi-media
transport and transformation models: regional fugacity model vs. CalTOX. Chemosphere, 30,
869-889. http://dx.doi.org/10.1016/0045-6535(94)00447-3
Mahle, D. A.; Gearhart, J. M.; Grigsby, C. C.; Mattie, D. R.; Barton, H. A.; Lipscomb, J. C.; Cook, R. S.
(2007). Age-dependent partition coefficients for a mixture of volatile organic solvents in
Sprague-Dawley rats and humans. J Toxicol Environ Health A, 70, 1745-1751.
http://dx.doi.org/10.1080/152873907Q1458991
Mailhes, J.; Aardema, M.; Marchetti, F. (1993). Investigation of aneuploidy induction in mouse oocytes
following exposure to vinblastine-sulfate, pyrimethamine, diethylstilbestrol diphosphate, or
chloral hydrate. Environ Mol Mutagen, 22, 107-114.
http: //www .ncbi .nlm .nih. gov/pubmed/8359152
Maitre, A.; Hours, M.; Bonneterre, V.; Arnaud, J.; Arslan, M. T.; Carpentier, P., . . . de Gaudemaris, R.
(2004). Systemic sclerosis and occupational risk factors: role of solvents and cleaning products. J
Rheumatol, 31, 2395-2401. http://www.ncbi.nlm.nih.gov/pubmed/15570640
Malherbe, P.; Alberati-Giani, D.; Kohler, C.; Cesura, A. M. (1995). Identification of a mitochondrial form
of kynurenine aminotransferase/glutamine transaminase K from rat brain. FEBS Lett, 367, 141-
144. http://www.ncbi.nlm.nih.gov/pubmed/7796908
Mallin, K. (1990). Investigation of a bladder cancer cluster in northwestern Illinois. Am J Epidemiol, 132,
S96-106. http://www.ncbi.nlm.nih.gov/pubmed/2356842
Malone, K. E.; Koepsell, T. D.; Daling, J. R.; Weiss, N. S.; Morris, P. D.; Taylor, J. W., . . . Lyon, J. L.
(1989). Chronic lymphocytic leukemia in relation to chemical exposures. Am J Epidemiol, 130,
1152-1158. http://www ncbi nlm nih gov/pubmed/2589308
Maloney, E. K. and Waxman, D. J. (1999). trans-Activation of PPARalpha and PPARgamma by
structurally diverse environmental chemicals. Toxicol Appl Pharmacol, 161, 209-218.
http://dx.doi.org/10.1006/taap.1999.8809
Mandel, J. S.; McLaughlin, J. K.; Schlehofer, B.; Mellemgaard, A.; Helmert, U.; Lindblad, P., . . . Adami,
H.-O. (1995). International renal-cell cancer study. IV. Occupation. Int J Cancer, 61, 601-605.
http://dx.doi.org/10.1002/iic.29106105Q3
Mannervik, B. (1985). The isoenzymes of glutathione transferase. Adv Enzymol Relat Areas Mol Biol,
57, 357-417. http://www.ncbi.nlm.nih.gov/pubmed/3898742
Manzo, L.; Artigas, F.; Martinez, E.; Mutti, A.; Bergamaschi, E.; Nicotera, P., . . . Costa, L. G. (1996).
Biochemical markers of neurotoxicity A review of mechanistic studies and applications. Hum
Exp Toxicol, 1], S20-S35.
Marano, D.; Boice, J.; Fryzek, J.; Morrison, J.; Sadler, C.; McLaughlin, J. (2000). Exposure assessment
for a large epidemiological study of aircraft manufacturing workers. Appl Occup Environ Hyg,
15, 644-656. http://dx.doi.org/10.1080/10473220050Q75653
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-36 DRAFT—DO NOT CITE OR QUOTE
-------
Marquardt, J. U. and Thorgeirsson, S. S. (2010). Stem cells in hepatocarcinogenesis: Evidence from
genomic data. Semin Liver Dis, 30, 26-34. http://dx.doi.org/10.1055/s-0030-1247130
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.
htto: //www .ncbi .nlm .nih. gov/pubmed/7854141
Marsh, G.; Youk, A.; Stone, R.; Buchanich, J.; Gula, M.; Smith, T.; Quinn, M. (2001). Historical cohort
study of US man-made vitreous fiber production workers: I. 1992 fiberglass cohort follow-up:
initial findings. J Occup Environ Med, 43, 741-756.
http: //www .ncbi .nlm .nih. gov/pubmed/11561358
Marsman, D. S.; Cattley, R. C.; Conway, J. G.; Popp, J. A. (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-pyrimidinylthio]acetic
acid (Wy-14,643) in rats. Cancer Res, 48, 6739-6744.
http: //www .ncbi .nlm .nih. gov/pubmed/3180084
Marsman, D. S.; Goldsworthy, T. L.; Popp, J. A. (1992). Contrasting hepatocytic peroxisome
proliferation, lipofuscin accumulation and cell turnover for the hepatocarcinogens Wy-14,643 and
clofibric acid. Carcinogenesis, 13, 1011-1017. http://dx.doi.Org/10.1093/carcin/13.6.1011
Marth, E. (1987). Metabolic changes following oral exposure to tetrachloroethylene in subtoxic
concentrations. Arch Toxicol, 60, 293-299.
Marth, E.; Stuenzner, D.; Binder, H.; Moese, J. R. (1985a). Tetrachlorethylen Eine Studie uber die
Wirkung niedriger Konzentrationen von 1,1,2,2-Tetrachlorethylen (Perchlorethylen) am
Organismus der Maus I Laborchemische Untersuchungen [Tetrachloroethylene A study of the
effect of low concentrations of 1,1,2,2-tetrachloroethylene on the organism of the mouse. I.
Clinical-chemical investigation], Int J Hyg Environ Health, 181, 525-540.
Marth, E.; Stuenzner, D.; Binder, H.; Moese, J. R. (1985b). Tetrachlorethylen: eine Studie uber die
Wirkung niedriger Konzentrationen von 1,1,2,2,-Tetrachlorethylen (Perchlorethylen) am
Organismus der Maus II Ruckstandsuntersuchungen von Tetrachlorethylen in verschiedenen
Organen und Nachweis von histologischen Veranderungen der untersuchten organe
[Tetrachloroethylene. a study of the effect of low concentrations of 1,1,2,2-tetrachloroethylene on
the organism of the mouse. II. Examinations of tetrachloroethylene-residues in various organs and
establishment of the examined organs]. Int J Hyg Environ Health, 181, 541-547.
Marth, E.; Stunzner, D.; Binder, H.; Mose, J. R. (1985c). [Tetrachloroethylene: effect of low
concentrations of 1,1,2,2-tetrachloroethylene (perchloroethylene) on organisms in the mouse. I.
Laboratory chemical research], Zentralbl Bakteriol Mikrobiol Hyg, 181, 525-540.
Marth, E.; Stunzner, D.; Kock, M.; Mose, J. R. (1989). Toxicokinetics of chlorinated hydrocarbons. J Hyg
Epidemiol Microbiol Immunol, 33, 514-520. http://www.ncbi.nlm.nih.gov/pubmed/2634072
Matsushima, T.; Hayashi, M.; Matsuoka, A.; Ishidate, M.; Miura, K. F.; Shimizu, H., . . . Sofuni, T.
(1999). Validation study of the in vitro micronucleus test in a Chinese hamster lung cell line
(CHL/IU). Mutagenesis, 14, 569-580. http://dx.doi.Org/10.1093/mutage/14.6.569
Mattsson, J.; Albee, R. R.; Yano, B. L.; Bradley, G. J.; PJ, S. (1998). Neurotoxicologic examination of
rats exposed to 1,1,2,2-tetrachloroethylene (perchloroethylene) vapor for 13 weeks. Neurotoxicol
Teratol, 20, 83-98.
May, R. (1976). Aufnahme Metabolisierung und Ausscheidung von Tetrachlorathylen und Beeinflussung
durch gleichzeitige Aufnahme von Athanol beim Menschen. Universitat Wurzburg, Wurzburg,
Germany.
Mazzullo, M.; Grilli, S.; Lattanzi, G.; Prodi, G.; Turina, M. P.; Colacci, A. (1987). Evidence of DNA
binding activity of perchloroethylene. Res Comm Chem Pathol Pharmacol, 58, 215-235.
http: //www .ncbi .nlm .nih. gov/pubmed/2447621
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-37 DRAFT—DO NOT CITE OR QUOTE
-------
McCarver, D. G.; Byun, R.; Hines, R. N.; 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, 276-
281. http://dx.doi.org/10.10Q6/taap.1998.8532
McConnell, G.; Ferguson, D. M.; Pearson, C. R. (1975). Chlorinated hydrocarbons and the environment.
Endeavour, 34, 13-18. http://www.ncbi.nlm.nih.gov/pubmed/54249
McCredie, M. and Stewart, J. H. (1993). Risk factors for kidney cancer in New South Wales. IV.
Occupation. Br J Ind Med, 50, 349-354. http://www.ncbi.nlm.nih.gov/pubmed/8494775
McDermott, M. J.; Mazor, K. A.; Shost, S. J.; Narang, R. S.; Aldous, K. M.; Storm, J. E. (2005).
Tetrachloroethylene (PCE, Perc) levels in residential dry cleaner buildings in diverse
communities in New York City. Environ Health Perspect, 113, 1336-1343.
McDonald, A. D.; Armstrong, B.; Cherry, N. M.; Delorme, C.; Diodati-Nolin, A.; McDonald, J. C.;
Robert, D. (1986). Spontaneous abortion and occupation. J Occup Med, 28, 1232-1238.
McDonald, A. D.; McDonald, J. C.; Armstrong, B.; Cherry, N.; Delorme, C.; Nolin, A.; Robert, D.
(1987). Occupation and pregnancy outcome. Br J Ind Med, 44, 521-526.
McDougal, J. N.; Jepson, G. W.; 3rd, C. H.; Gargas, M. L.; Andersen, M. E. (1990). Dermal absorption of
organic chemical vapors in rats and humans. Toxicol Sci, 14, 299-308.
McKone, T. E. and Bogen, K. T. (1992). Uncertainties in health-risk assessment: an integrated case study
based on tetrachloroethylene in California groundwater. Regul Toxicol Pharmacol, 15, 86-103.
McLean, D.; Mannetje, A.; Dryson, E.; Walls, C.; McKenzie, F.; Maule, M., . . . Pearce, N. (2009).
Leukaemia and occupation: A New Zealand Cancer Registry-based case-control Study. Int J
Epidemiol, 38, 594-606. http://dx.doi.org/10.1093/iie/dvn220
MDPH. (Massachusetts Department of Public Health), (1997a). Woburn childhood leukemia follow-up
study: Information booklet. Boston, MA: Massachusetts Department of Public Health, Bureau of
Environmental Health Assessment.
MDPH. (Massachusetts Department of Public Health). (1997b). Woburn childhood leukemia follow-up
study: Volume I: Analyses. Boston, MA: Massachusetts Department of Public Health, Bureau of
Environmental Health Assessment. Retrieved from
http://www.mass.gov/Eeohhs2/docs/dph/environmental/investigations/woburn cancer leukemia
follow up study 1997.pdf.
Meade, T. W.; Framework, F. t. B. M. R. C. G. P. R.; clinics, p. v. (2001). Design and intermediate results
of the Lower Extremity Arterial Disease Event Reduction (LEADER)* trial of bezafibrate in men
with lower extremity arterial disease [ISRCTN4119421], Current Controlled Trials in
Cardiovascular Medicine, 2, 195-204. http://dx.doi.org/10.1186/cvm-2-4-195
Meckler, L. C. and Phelps, D. K. (1966). Liver disease secondary to tetrachloroethylene exposure: A case
report. JAMA, 197, 662-663.
Mellemgaard, A.; Engholm, G.; McLaughlin, J. K.; Olsen, J. H. (1994). Occupational risk factors for
renal-cell carcinoma in Denmark. Scand J Work Environ Health, 20, 160-165.
http: //www .ncbi .nlm .nih. gov/pubmed/7973487
Melnick, R. L. (2001). Is peroxisome proliferation an obligatory precursor step in the carcinogenicity of
di(2-ethylhexyl)phthalate (DEHP)? Environ Health Perspect, 109, 437-442.
http: //www .ncbi .nlm .nih. gov/pubmed/11401753
Melnick, R. L.; Kohn, M. C.; Portier, C. J. (1996). Implications for risk assessment of suggested
nongenotoxic mechanisms of chemical carcinogenesis. Environ Health Perspect, 104 Suppl 1,
123-134. http://www.ncbi.nlm.nih.gov/pubmed/8722116
Mera, N.; Ohmori, S.; Itahashi, K.; Kiuchi, M.; Igarashi, T.; Rikihisa, T.; Kitada, M. (1994).
Immunochemical evidence for the occurrence of Mu class glutathione S-transferase in human
fetal livers. J Biochem, 116, 315-320. http://www.ncbi.nlm.nih.gov/pubmed/7822249
Mergler, D. (1987). Worker participation in occupational health research: Theory and practice. Int J
Health Serv, 17, 151-167.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-38 DRAFT—DO NOT CITE OR QUOTE
-------
Mergler, D.; Belanger, S.; De Grosbois, S.; Vachon, N. (1988a). Chromal focus of acquired chromatic
discrimination loss and solvent exposure among printshop workers. Toxicology, 49, 341-348.
http://dx.doi.org/10.1016/0300-483X(88)90017-0
Mergler, D. and Blain, L. (1987). Assessing color vision loss among solvent-exposed workers. Am J Ind
Med, 12, 195-203. http://dx.doi.org/10.1002/aiim.47001202Q8
Mergler, D.; Blain, L.; Lemaire, J.; Lalande, F. (1988b). Colour vision impairment and alcohol
consumption. Neurotoxicol Teratol, 10, 255-260. http://dx.doi.org/10.1016/0892-0362(88)90025-
6
Mergler, D.; Huel, G.; Bowler, R.; Frenette, B.; Cone, J. (1991). Visual dysfunction among former
microelectronics assembly workers. Arch Environ Health, 46, 326-334.
http: //www .ncbi .nlm .nih. gov/pubmed/1772256
Meskar, A.; Plee-Gautier, E.; Amet, Y.; Berthou, F.; Lucas, D. (2001). Interactions alcool-xenobiotiques.
Role du cytochrome P450 2E1 [Alcohol-xenobiotic interactions. Role of cytochrome P450 2E1],
Pathol Biol, 49, 696-702. http://dx.doi.org/10.1016/S0369-8114(01)00235-8
Mester, B.; Nieters, 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.
http://dx.doi.Org/10.l 136/oem.2005.020453
Metabolism of trichloroethylene and covalent binding of reaction products. (2004). Symposium on new
scientific research related to the health effects of trichloroethylene. Washington, DC.
Miettinen, O. S. (1976). Stratification by a multivariate confounder score. Am J Epidemiol, 104, 609-620.
http: //www .ncbi .nlm .nih. gov/pubmed/998608
Miligi, L.; Costantini, A. S.; Benvenuti, A.; Kriebel, D.; Bolejack, V.; Tumino, R., . . . Vineis, P. (2006).
Occupational exposure to solvents and the risk of lymphomas. Epidemiology, 17, 552-561.
http://dx.doi.org/10.1097/01.ede.0000231279.3Q988.4d
Miligi, L.; Seniori, C. A.; Crosignani, P.; Fontana, A.; Masala, G.; Nanni, O., . . . Vineis, P. (1999).
Occupational, environmental, and life-style factors associated with the risk of
hematolymphopoietic malignancies in women. Am J Ind Med, 36, 60-69.
http://dx.doi.org/10.1002/(SICP 1097-0274(199907)36:1<60::AID-AJIM9>3.0.CQ;2-Z
Millar, R.; Warden, J.; Cooperman, L.; Price, H. (1969). Central sympathetic discharge and mean arterial
pressure during halothane anaesthesia. Br J Anaesth, 41, 918-928.
http://www.ncbi.nlm.nih.gov/pubmed/4982187
Miller, B. M. and Adler, I.-D. (1992). Aneuploidy induction in mouse spermatocytes. Mutagenesis, 7, 69-
76.
Miller, R. E. and Guengerich, F. P. (1982). Oxidation of trichloroethylene by liver microsomal
cytochrome P-450: evidence for chlorine migration in a transition state not involving
trichloroethylene oxide. Biochemistry, 21, 1090-1097. http://dx.doi.org/10.102l/bi00534a041
Miller, R. E. and Guengerich, F. P. (1983). Metabolism of trichloroethylene in isolated hepatocytes,
microsomes, and reconstituted enzyme systems containing cytochrome P-450. Cancer Res, 43,
1145-1152. http://www.ncbi.nlm.nih.gov/pubmed/6825087
Mitchell, A. E.; Morin, D.; Lakritz, J.; Jones, A. D. (1997). Quantitative profiling of tissue- and gender-
related expression of glutathione S-transferase isoenzymes in the mouse. Biochem J, 325, 207-
216.
Mitchell, A. M.; Lhuguenot, J. C.; Bridges, J. W.; Elcombe, C. R. (1985). Identification of the proximate
peroxisome proliferator(s) derived from di(2-ethylhexyl) phthalate. Toxicol Appl Pharmacol, 80,
23-32.
Moen, B. E.; Kyvik, K. R.; Engelsen, B. A.; Riise, T. (1990). Cerebrospinal fluid proteins and free amino
acids in patients with solvent induced chronic toxic encephalopathy and healthy controls. Br J Ind
Med, 47, 277-280. http://www.ncbi.nlm.nih.gov/pubmed/2337535
Moghaddam, A. P.; Abbas, R.; Fisher, J. W.; Lipscomb, J. C. (1997). The role of mouse intestinal
microflora in the metabolism of trichloroethylene, an in vivo study. Hum Exp Toxicol, 16, 629-
635. http://dx.doi.org/10.1177/0960327197016Q1101
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-39 DRAFT—DO NOT CITE OR QUOTE
-------
Moghaddam, A. P.; Abbas, R.; Fisher, J. W.; Stavrou, S.; Lipscomb, J. C. (1996). Formation of
dichloroacetic acid by rat and mouse gut microflora, an in vitro study. Biochem Biophys Res
Commun, 228, 639-645. http://dx.doi.org/10.1006/bbrc.1996.17Q9
Monster, A.; Regouin-Peeters, W.; Van Schijndel, A.; Van der Tuin, J. (1983). Biological monitoring of
occupational exposure to tetrachloroethene. Scand J Work Environ Health, 9, 273-281.
Monster, A. C. (1979). Difference in uptake, elimination, and metabolism in exposure to
trichloroethylene, 1,1,1-trichloroethane, and tetrachloroethylene. Int Arch Occup Environ Health,
42, 311-317. http://www.ncbi.nlm.nih.gov/pubmed/422272
Monster, A. C.; Boersma, G.; Steenweg, H. (1979). Kinetics of tetrachloroethylene in volunteers;
influence of exposure concentration and work load. Int Arch Occup Environ Health, 42, 303-309.
Monster, A. C. and Houtkooper, J. M. (1979). Estimation of individual uptake of trichloroethylene, 1,1,1-
trichloroethane and tetrachloroethylene from biological parameters. Int Arch Occup Environ
Health, 42, 319-323. http://www.ncbi.nlm.nih.gov/pubmed/422273
Monster, A. C. and Smolders, J. F. (1984a). Tetrachloroethene in exhaled air of persons living near
pollution sources. Int Arch Occup Environ Health, 53, 331-336.
Monster, A. C. and Smolders, J. F. J. (1984b). Tetrachloroethene in exhaled air of persons living near
polution sources. Int Arch Occup Environ Health, 53, 331-336.
Moore, M. M. and Harrington-Brock, K. (2000). Mutagenicity of trichloroethylene and its metabolites:
Implications forthe risk assessment of trichloroethylene. Environ Health Perspect, 108, 215-223.
http: //www .ncbi .nlm .nih. gov/pubmed/10807553
Moos, F. and Richard, P. (1975). [Adrenergic and cholinergic control of oxytocin release evoked by
vaginal, vagal and mammary stimulation in lactating rats (author's transl)]. 70, 315-332.
http: //www .ncbi .nlm .nih. gov/pubmed/1524
Morrow, L. A.; Ryan, C. M.; Hodgson, M. J.; Robin, N. (1990). Alterations in cognitive and
psychological functioning after organic solvent exposure. J Occup Med, 32, 444-450.
Morse, H. C., Ill; Anver, M. R.; Fredrickson, T. N.; Haines, D. C.; Harris, A. W.; Harris, N. L., . . .
Ward, J. M. (2002). Bethesda proposals for classification of lymphoid neoplasms in mice.
Neoplasia, 100, 246-258. http://www.ncbi.nlm.nih.gov/pubmed/12070034
Morton, W. and Marjanovic, D. (1984). Leukemia incidence by occupation in the Portland-Vancouver
metropolitan area. Am J Ind Med, 6, 185-205. http://www.ncbi.nlm.nih.gov/pubmed/6475965
Moryama, M. T.; Domiki, C.; Miyazawa, K.; Tanaka, T.; Suzuki, K. (2005). Effects of oxalate exposure
on Madin-Darby canine kidney cells in culture: Renal prothrombin fragment-1 mRNA
expression. Urol Res, 33, 470-475. http://dx.doi.org/10.1007/s00240-005-051Q-6
Moser, V. C.; Cheek, B. M.; MacPhail, R. C. (1995). A multidisciplinary approach to toxicological
screening III Neurobehavioral toxicity. J Toxicol Environ Health, 45, 173-210.
http: //www .ncbi. nlm. nih. go v/pubmed/7783252
Moslen, M. T.; Reynolds, E. S.; Boor, P. J.; Bailey, K.; Szabo, S. (1977a). Trichloroethylene-induced
deactivation of cytochrome P-450 and loss of liver glutathione in vivo. Res Commun Mol Pathol
Pharmacol, 16, 109-120. http://www.ncbi.nlm.nih.gov/pubmed/841173
Moslen, M. T.; Reynolds, E. S.; Szabo, S. (1977b). Enhancement of the metabolism and hepatotoxicity of
trichloroethylene and perchloroethylene. Biochem Pharmacol, 26, 369-375.
http: //www .ncbi .nlm .nih. gov/pubmed/19223 9
Motohashi, Y.; Miyazaki, Y.; Takano, T. (1993). Assessment of behavioral effects of tetrachloroethylene
using a set of time-series analyses. Neurotoxicol Teratol, 15, 3-10.
Mukheijee, R.; Jow, L.; Noonan, D.; McDonnell, D. P. (1994). Human and rat peroxisome proliferator
activated receptors (PPARs) demonstrate similar tissue distribution but different responsiveness
to PPAR activators. J Steroid Biochem Mol Biol, 51, 157-166.
http://www.ncbi.nlm.nih.gov/pubmed/7981125
Murakami, K. and Horikawa, K. (1995). The induction of micronuclei in mice hepatocytes and
reticulocytes by tetrachloroethylene. Chemosphere, 31, 3733-3739.
http://dx.doi.org/10.1016/0045-6535(95)00222-T
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-40 DRAFT—DO NOT CITE OR QUOTE
-------
Mutti, A.; Alinovi, R.; Bergamaschi, E.; Biagini, C.; Cavazzini, S.; Franchini, I., . . . Herbort, C. (1992).
Nephropathies and exposure to perchloroethylene in dry-cleaners. Lancet, 330, 189-193.
http://dx.doi.org/10.1016/0140-6736(92)90463-0
Muttray, A.; Wolff, U.; Jung, D.; Konietzko, J. (1997). Blue-yellow deficiency in workers exposed to low
concentrations of organic solvents. Int Arch Occup Environ Health, 70, 407-412.
Nagano, K.; Nishizawa, T.; Yamamoto, S.; Matsushima, T. (1998). Inhalation carcinogenesis studies of
six halogenated hydrocarbons in rats and mice. In K. Chiyotani, Y. Hosoda & Y. Aizawa (Eds.),
Advances in Prevention of Occupational Respiratory Diseases (pp. 741-746). Amsterdam:
Elsevier Science.
Nakai, J. S.; Stathopulos, P. B.; Campbell, G. L.; Chu, I.; Li-Muller, A.; Aucoin, R. (1999). Penetration of
chloroform, trichloroethylene, and tetrachloroethylene through human skin. J Toxicol Environ
Health A, 58, 157-170. http://dx.doi.org/10.1080/009841Q99157368
Nakasa, H.; Mera, N.; Ohmori, S.; Itahashi, K.; Igarashi, T.; Ishii, I.; Kitada, M. (1997). Nucleotide
sequence of pi class glutathione S-transferase in human fetal liver. Res Commun Mol Pathol
Pharmacol, 97, 67-78. http://www.ncbi.nlm.nih.gov/pubmed/9507570
Nakatsuka, H.; Watanabe, T.; Takeuchi, Y.; Hisanaga, N.; Shibata, E.; Suzuki, H., . . . Ikeda, M. (1992).
Absence of blue-yellow color vision loss among workers exposed to toluene or
tetrachloroethylene, mostly at levels below occupational exposure limits. Int Arch Occup Environ
Health, 64, 113-117. http://dx.doi.org/10.1007/BF0Q381478
Narotsky, M. G. and Kavlock, R. J. (1995). A multidisciplinary approach to toxicological screening: II.
Developmental toxicity. J Toxicol Environ Health, 45, 145-171.
http://dx.doi.org/10.1080/15287399509531987
National primary and secondary ambient air quality standards, 36 8186-8201 (1971).
NCI. (National Institutes of Health, National Cancer Institute). (1977). Bioassay of tetrachloroethylene for
possible carcinogenicity. (NCI-CGTR-13; DHEW Publication No. (NIH) 77-813). Bethesda, Md:
National Institutes of Health. Retrieved from
http ://ntp .niehs .nih. gov/ntp/htdocs/LT rpts/trO 13 .pdf.
NCI. (National Institutes of Health, National Cancer Institute). (2008). SEER cancer statistics review,
1975-2005. Bethesda, MD. Retrieved from http://seer.cancer.gov/csr/1975 2005.
NCI. (National Institutes of Health, National Cancer Institute). (2010). What you need to know about
cervical cancer. (NIH Publication No. 08-2407). Washington, D.C.: U.S. Department of Health
and Human Services. Retrieved from http://www.cancer.gov/cancertopics/wvntk/cervix.pdf.
Neafsey, P.; Ginsberg, G.; Hattis, D.; Johns, D. O.; Guyton, K. Z.; Sonawane, B. (2009). Genetic
polymorphism in CYP2E1: Population distribution of CYP2E1 activity. J Toxicol Environ Health
B Crit Rev, 12, 362-388. http://dx.doi.org/10.1080/109374009Q3158359
Nelson, A.; Mendoza, T.; Hoyle, G.; Brody, A.; Fermin, C.; Morris, G. (2001). Enhancement of
fibrogenesis by the p53 tumor suppressor protein in asbestos-exposed rodents. Chest, 120, 33S-
34S. http://www.ncbi.nlm.nih.gOv/pubmed/l 1451905
Nelson, B. K.; Taylor, B. J.; Setzer, J. V.; Hornung, R. W. (1980). Behavioral teratology of
perchloroethylene in rats. J Environ Pathol Toxicol Oncol, 3, 233-250.
Nelson, M. A. and Bull, R. J. (1988). Induction of strand breaks in DNA by trichloroethylene and
metabolites in rat and mouse liver in vivo. Toxicol Appl Pharmacol, 94, 45-54.
http://dx.doi.org/10.1016/004 l-008X(88')90335-3
Nelson, M. A.; Lansing, A. J.; Sanchez, I. M.; Bull, R. J.; Springer, D. L. (1989). Dichloroacetic acid and
trichloroacetic acid-induced DNA strand breaks are independent of peroxisome proliferation.
Toxicology, 58, 239-248. http://dx.doi.org/10.1016/0300-483X(89)90139-X
Nesslany, F. and Marzin, D. (1999). A micromethod for the in vitro micronucleus assay. Mutagenesis, 14,
403-410. http://www.ncbi.nlm.nih.gov/pubmed/103905Q8
Nestmann, E. R.; Chu, I.; Kowbel, D. J.; Matula, T. I. (1980). Short-lived mutagen in Salmonella
produced by reaction of trichloroacetic acid and dimethyl sulphoxide. Can J Genet Cytol, 22, 35-
40. http://dx.doi.Org/10.l 139/g80-006
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-41 DRAFT—DO NOT CITE OR QUOTE
-------
Ni, Y.; Wong, T.; Kadlubar, F.; Fu, P. (1994). Hepatic metabolism of chloral hydrate to free radical(s)
and induction of lipid peroxidation. Biochem Biophys Res Commun, 204, 937-943.
http://dx.doi.org/10.1006/bbrc. 1994.2550
Ni, Y. C.; Kadlubar, F. F.; Fu, P. P. (1995). Formation of malondialdehyde-modified 2'-deoxyguanosinyl
adduct from metabolism of chloral hydrate by mouse liver microsomes. Biochem Biophys Res
Commun, 216, 1110-1117. http://dx.doi.org/10.1006/bbrc. 1995.2735
NIOSH. (National Institute for Occupational Safety and Health). (1994). Fatality assessment and control
evaluation (FACE) for Colorado: 17-year-old worker at a plastic products manufacturing plant
died as a result of an overexposure to tetrachloroethylene (also known as perchloroethylene).
(94C0006A). Morgantown, West Virginia: Author. Retrieved from
http://www.cdc.gov/niosh/face/stateface/co/94co006.html.
Nomiyama, K.; Liu, S. J.; Nomiyama, H. (1992). Critical levels of blood and urinary cadmium, urinary
beta 2-microglobulin and retinol-binding protein for monitoring cadmium health effects. IARC
Sci Publ, 1992, 325-340. http://www.ncbi.nlm.nih.gov/pubmed/1303959
Norton, P. K.; Hoff, H. E.; Bellis, D. H.; Ham, D. (1968). VARIANTS OF THE RESPIRATORY
HEART RATE RESPONSE IN SHEEP MORPHINE CENT DEPRESS CHLORAL HYDRATE
CENT DEPRESS HUMAN. 7, 58-70.
Notice of availability; Documents entitled: Guidelines for carcinogen risk assessment and supplemental
guidance for assessing susceptibility from early-life exposure to carcinogens, 70 17765-17817
(2005).
NRC. (National Research Council). (1983). Risk assessment in the federal government: Managing the
process. Washington, DC: National Academy Press.
NRC. (National Research Council). (1993). Pesticides in the diets of infants and children. Washington,
DC: National Academy Press.
NRC. (National Research Council). (1995). Biologic markers in urinary toxicology. Washington, DC:
The National Academies Press.
NRC. (National Research Council). (2006). Assessing the human health risks of trichloroethylene: Key
scientific issues. Washington, DC: The National Academies Press. Retrieved from
http://nae.edu/nae/naepcms.nsf/weblinks/MKEZ-6SSHPD70penDocument.
NTP. (National Toxicology Program). (1983). Carcinogenesis studies of pentachloroethane (CAS No. 76-
01-7) in F344/N rats and B6C3F1 mice (gavage study). Research Triangle Park, NC: Public
Health Service, U.S. Department of Health and Human Services. Retrieved from
http ://ntp .niehs .nih. gov/ntp/htdocs/LT rpts/tr232 .pdf.
NTP. (National Toxicology Program). (1986a). Toxicology and carcinogenesis studies of isophorone
(CAS No. 78-59-1) in F344/N rats and B6C3F1 mice (gavage studies). Research Triangle Park,
NC: Public Health Service, U.S. Department of Health and Human Services. Retrieved from
http://ntp.niehs.nih.gov/ntp/htdocs/LT rpts/tr291 .pdf.
NTP. (National Toxicology Program). (1986b). Toxicology and carcinogenesis studies of
tetrachloroethylene (perchloroethylene) (CAS No. 127-18-4) in F344/N rats and B6C3F1 mice
(inhalation studies). (NIH Publication No. 86-2567). RTP, NC: Public Health Service, U.S.
Department of Health and Human Services. Retrieved from
http://ntp.niehs.nih.gov/ntp/htdocs/LT rpts/tr311 .pdf.
NTP. (National Toxicology Program). (1987). Toxicology and carcinogenesis studies of dimethyl
methylphosponate (CAS No. 756-79-6) in F344/N rats and B6C3F 1 mice. Research Triangle
Park, NC: Public Health Service, U.S. Department of Health and Human Services. Retrieved
from http://ntp.niehs.nih.gov/ntp/htdocs/LT rpts/tr323.pdf.
NTP. (National Toxicology Program). (1989). Toxicology and carcinogenesis studies of
hexachloroethane (CAS No. 67-72-1) in F344/N rats (gavage studies). Research Triangle Park,
NC: Public Health Service, U.S. Department of Health and Human Services. Retrieved from
http://ntp.niehs.nih.gov/ntp/htdocs/LT rpts/tr361 .pdf.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-42 DRAFT—DO NOT CITE OR QUOTE
-------
NTP. (National Toxicology Program). (1990). Toxicology and carcinogenesis studies of d-limonene
(CAS NO. 5989-27-5) in F344/N rats and B6C3F1 mice (gavage studies). (PB90231416).
Research Triangle Park, NC: U.S. Department of Health and Human Services, Public Health
Service, National Institutes of Health. Retrieved from
htto://nto.niehs.nih.gov/ntD/htdocs/LT rpts/tr347 .pdf.
NTP. (National Toxicology Program). (2005). Report on carcinogens (11th edition). Research Triangle
Park, NC: Public Health Service, U.S. Department of Health and Human Services. Retrieved
from http://ntp .niehs .nih.gov/index.cfm?obiectid=32BA9724-FlF6-975E-7FCE50709CB4C932.
NTP. (National Toxicology Program). (2007). NTP report on the toxicology studies of dichloroacetic acid
(CAS No. 79-43-6) in genetically modified (FVB Tg.AC hemizygous) mice (dermal and drinking
water studies) and carcinogenicity studies of dichloroacetic acid in genetically modified [B6.129-
Trp53tmlBrd(N5) haploinsufficient] mice (drinking water studies). (PB2008-109740; NTP
GMM 11; NIH PUB 07-4428). Research Triangle Park: Author. Retrieved from
http://www.ntis.gov/search/product.aspx?ABBR=PB2008109740.
Nutley, E.; Tcheong, A.; Allen, J.; Collins, B.; Ma, M.; Lowe, X., . . . Wyrobek, A. (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, 80-89.
http://dx.doi.org/10.1002/(SICD1098-2280(1996)28:2<80::AID-EM3>3.0.CQ;2-I
NY State Department of Health. (New York State Department of Health). (1997). Tetrachloroethene
ambient air criteria document. Final report. Albany, NY: Author.
NY State Department of Health. (New York State Department of Health). (2004). Comment focusing on
"Neurotoxicity of Tetrachloroethylene (Perchloroethylene) Discussion Paper" submitted by Jan E.
Storm and Kimberly A. Mazor, New York State Department of Health for EPA Docket ORD-
2003-0014. March 3, 2004. New York: Center for Environmental Health, Bureau of Toxic
Substance Assessment. Retrieved from
http://www.regulations.gov/search/Regs/home .html#documentDetail?R=09000064800c44c0.
NY State Department of Health. (New York State Department of Health). (2005a). Improving human risk
assessment for tetrachloroethylene by using biomakers and neurobehavioral testing. Washington,
DC: U.S. Environmental Protection Agency.
NY State Department of Health. (New York State Department of Health). (2005b). Pumpkin patch day
care center follow-up evaluation. Final Report. Troy, NY: Center for Environmental Health,
Bureau of Toxic Substance Assessment.
NY State Department of Health. (New York State Department of Health). (2005c). Pumpkin patch day
care center investigation. Final Report. Troy, NY: Center for Environmental Health, Bureau of
Toxic Substance Assessment.
NYS OAG. (New York State Office of the Attorney General). (2004). Comment focusing on
"Neurotoxicity of Tetrachloroethylene (Perchloroethylene) Discussion Paper" submitted by Judith
S. Schreiber, Ph.D., Environmental Protection Bureau for EPA Docket ORD-2003-0014. March
1, 2004. Albany, NY: State of New York Office of the Attorney General. Retrieved from
http://www.regulations.gov/search/Regs/home .html#documentDetail?R=09000064800c44ad.
NYSDOH. (New York State Department of Health). (2000). Evaluation of residential exposure to
tetrachloroethene using biomarkers of dose and neurological tests. Albany, NY: Author.
O'Brien, M.; Spear, B.; Glauert, H. (2005). Role of oxidative stress in peroxisome proliferator-mediated
carcinogenesis. Crit Rev Toxicol, 35, 61-88. http://www.ncbi.nlm.nih.gov/pubmed/15742903
O'Flaherty, E. J.; Scott, W.; Schreiner, C.; Beliles, R. P. (1992). A physiologically based kinetic model of
rat and mouse gestation: disposition of a weak acid. Toxicol Appl Pharmacol, 112, 245-256.
http://dx.doi.org/10.1016/0041-008X(92')90194-W
Odum, J.; Green, T.; Foster, J. R.; Hext, P. M. (1988a). The role of trichloracetic acid and peroxisome
proliferation in the differences in carcinogenicity of perchloroethylene in the mouse and rat.
Toxicol Appl Pharmacol, 92, 103-112. http://dx.doi.org/10.1016/0041-008X(88)90232-3
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-43 DRAFT—DO NOT CITE OR QUOTE
-------
Odum, J.; Green, T.; Foster, J. R.; Hext, P. M. (1988b). The role of trichloroacetic acid and peroxisome
proliferation in the differences in carcinogenicity of perchloroethylene in the mouse and rat.
Toxicol Appl Pharmacol, 92, 103-112.
Ogata, M.; Sugiyama, K.; Kuroda, Y. I. (1962). Investigation of a dry-cleaning shop using
tetrachloroethylene, with special reference to Fujiwara's substance in the urine of the employees.
Okayama Igakkai Zasshi, 74, 253.
Ogata, M.; Takatsuka, Y.; Tomokuni, K. (1971). Excretion of organic chlorine compounds in the urine of
persons exposed to vapours of trichloroethylene and tetrachloroethylene. Occup Environ Med, 28,
386-391. htto://dx.doi.org/10.1136/oem.28.4.386
Ohtsuki, T.; Sato, K.; Koizumi, A.; Kumai, M.; Ikeda, M. (1983). Limited capacity of humans to
metabolize tetrachloroethylene. IntArch Occup Environ Health, 51, 381-390.
http://dx.doi.org/10.1007/BF0Q378352
Olsen, G. D. and Weil, J. A. (1991). Noninvasive analysis of breathing heart rate and brain electrical
activity in neonatal guinea-pigs effect of chloral hydrate. FASEB J, 5, A1237.
Olsen, J.; Hemminki, K.; Ahlborg, G.; Bjerkedal, T.; Kyyronen, P.; Taskinen, H., . . . Egenaes, J. (1990).
Low birthweight, congenital malformations, and spontaneous abortions among dry-cleaning
workers in Scandinavia. Scand J Work Environ Health, 16, 163-168.
http: //www .ncbi .nlm .nih. gov/pubmed/2143312
Olshan, A. F.; Anderson, L.; Roman, E.; Fear, N.; Wolff, M.; Whyatt, R., . . . Potischman, N. (2000).
Workshop to identify critical windows of exposure for children's health: cancer work group
summary. Environ Health Perspect, 108, 595-597.
Olson, D. M.; Sheehan, M. G.; Thompson, W.; Hall, P. T.; Hahn, J. (2001). Sedation of children for
electroencephalograms. Pediatrics, 108, 163-165.
http: //www .ncbi .nlm .nih. gov/pubmed/11433070
Ono, Y.; Somiya, I.; Kawamura, M. (1991). Genotoxicity of the by-products produced in chlorination and
ozonation processes. Mutat Res Environ Mutagen Relat Subj, 252, 102.
http://dx.doi.org/10.1016/0165-116K9n90301-N
Opdam, J. J. (1989a). Intra and interindividual variability in the kinetics of a poorly and highly
metabolising solvent. Br J Ind Med, 46, 831-845. http://www.ncbi.nlm.nih.gov/pubmed/2611156
Opdam, J. J. (1989b). Respiratory input in inhalation experiments. Occup Environ Med, 46, 145-156.
http://dx.doi.Org/10.l 136/oem.46.3.145
Opdam, J. J. and Smolders, J. F. (1986). Alveolar sampling and fast kinetics of tetrachloroethene in man.
I. Alveolar sampling. Br J Ind Med, 43, 814-824. http://dx.doi.Org/10.l 136/oem.43.12.814
Oshiro, W. M.; Krantz, Q. T.; Bushnell, P. J. (2008). Characterization of the effects of inhaled
perchloroethylene on sustained attention in rats performing a visual signal detection task.
Neurotoxicol Teratol, 30, 167-174. http://dx.doi.Org/10.1016/i.ntt.2008.01.002
Otieno, M. A.; Baggs, R. B.; Hayes, J. D.; Anders, M. W. (1997). Immunolocalization of microsomal
glutathione S-transferase in rat tissues. Drug Metab Dispos, 25, 12-20.
http: //www .ncbi .nlm .nih. gov/pubmed/9010624
Overby, L. H.; Gardlik, S.; Philpot, R. M.; Serabjit-Singh, C. J. (1994). Unique distribution profiles of
glutathione S-transferases in regions of kidney, ureter, and bladder of rabbit. Lab Invest, 70, 468-
478.
Pahler, A.; Parker, J.; Dekant, W. (1999a). Dose-dependent protein adduct formation in kidney, liver, and
blood of rats and in human blood after perchloroethene inhalation. Toxicol Sci, 48, 5-13.
http: //www .ncbi .nlm .nih. gov/pubmed/10330678
Pahler, A.; Volkel, W.; Dekant, W. (1999b). Quantitation ofN epsilon-(dichloroacetyl)-L-lysine in
proteins after perchloroethene exposure by gas chromatography-mass spectrometry using
chemical ionization and negative ion detection following immunoaffinity chromatography. J
Chromatogr A, 847, 25-34.
Palmer, C. N.; Hsu, M. H.; Griffin, K. J.; Raucy, J. L.; Johnson, E. F. (1998). Peroxisome proliferator
activated receptor-alpha expression in human liver. Mol Pharmacol, 53, 14-22.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-44 DRAFT—DO NOT CITE OR QUOTE
-------
Parent, M. E.; Hua, Y.; Siemiatycki, J. (2000a). Occupational risk factors for renal cell carcinoma in
Montreal. Am J Ind Med, 38, 609-618. http://www ncbi nlm nib gov/pubmed/11071683
Parent, M. E.; Siemiatycki, J.; Fritschi, L. (2000b). Workplace exposures and oesophageal cancer. Occup
Environ Med, 57, 325-334. http://www.ncbi.nlm.nih.gov/pubmed/10769298
Park, B.; Vogelstein, B.; Kinzler, K. (2001). Genetic disruption of PPARdelta decreases the
tumorigenicity of human colon cancer cells. PNAS, 98, 2598-2603.
http://dx.doi.org/10.1073/pnas.051630998
Park, E. J.; Lee, J. H.; Yu, G. Y.; He, G.; Ali, S. R.; Holzer, R. G., . . . Karin, M. (2010). Dietary and
genetic obesity promote liver inflammation and tumorigenesis by enhancing IL-6 and TNF
expression. Cell, 140, 197-208. http://dx.doi.Org/10.1016/i.cell.2009.12.052
Park, J.-H.; Spengler, J. D.; Yoon, D.-W.; Dumyahn, T.; Lee, K.; Ozkaynak, H. (1998). Measurement of
air exchange rate of stationary vehicles and estimation of in-vehicle exposure. J Expo Sci Environ
Epidemiol, 8, 65-78.
Parkinson, A.; Mudra, D. R.; Johnson, C.; Dwyer, A.; Carroll, K. M. (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, 193-209.
http://dx.doi.Org/10.1016/i.taap.2004.01.010
Partanen, T.; Heikkila, P.; Hernberg, S.; Kauppinen, T.; Moneta, G.; Ojajarvi, A. (1991). Renal cell
cancer and occupational exposure to chemical agents. Scand J Work Environ Health, 17, 231-
239. http://www.ncbi.nlm.nih.gov/pubmed/1925434
Parzefall, W.; Berger, W.; Kainzbauer, E.; Teufelhofer, O.; Schulte-Hermann, R.; Thurman, R. (2001).
Peroxisome proliferators do not increase DNA synthesis in purified rat hepatocytes.
Carcinogenesis, 22, 519-523. http://www.ncbi.nlm.nih.gOv/pubmed/l 1238195
Paulu, C.; Aschengrau, A.; Ozonoff, D. (1999). Tetrachloroethylene-contaminated drinking water in
Massachusetts and the risk of colon-rectum, lung, and other cancers. Environ Health Perspect,
107, 265-271. http://www nchi nlm nih gov/pubmed/10090704
Paulu, C.; Aschengrau, A.; Ozonoff, D. (2002). Exploring associations between residential location and
breast cancer incidence in a case-control study. Environ Health Perspect, 110, 471-478.
http: //www .ncbi .nlm .nih. gov/pubmed/12003750
Pearce, N.; Checkoway, H.; Kriebel, D. (2007). Bias in occupational epidemiology studies. Occup
Environ Med, 64, 562-568. http://dx.doi.Org/10.l 136/oem.2006.026690
Pegg, D. G.; Zempel, J. A.; Braun, W. H.; Watanabe, P. G. (1979). Disposition of
tetrachloro(14C)ethylene following oral and inhalation exposure in rats. Toxicol Appl Pharmacol,
51,465-474.
Pelekis, M.; Gephart, L. A.; Lerman, S. E. (2001). Physiological-model-based derivation of the adult and
child pharmacokinetic intraspecies uncertainty factors for volatile organic compounds. Regul
Toxicol Pharmacol, 33, 12-20. http://dx.doi.org/10.1006/rtph.2000.1436
Pellizzari, E. D.; Hartwell, T. D.; 3rd, H. B.; Waddell, R. D.; Whitaker, D. A.; Erickson, M. D. (1982).
Purgeable organic compounds in mother's milk. Bull Environ Contain Toxicol, 28, 322-328.
http: //www .ncbi .nlm .nih. gov/pubmed/7082873
Peplonska, B.; Stewart, P.; Szeszenia-Dabrowska, N.; Rusiecki, J.; Garcia-Closas, M.; Lissowska, J., . . .
Blair, A. (2007). Occupation and breast cancer risk in Polish women: a population-based case-
control study. Am J Ind Med, 50, 97-111. http://dx.doi.org/10.1002/aiim.2042Q
Pereira, M. A. (1996). Carcinogenic activity of dichloroacetic acid and trichloroacetic acid in the liver of
female B6C3F1 mice. Fundam Appl Toxicol, 31, 192-199.
http://dx.doi.org/10.1006/faat.1996.0Q91
Pereira, M. A.; Kramer, P. M.; Conran, P. B.; 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, 1511-1519.
http: //www .ncbi .nlm .nih. gov/pubmed/11532874
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-45 DRAFT—DO NOT CITE OR QUOTE
-------
Pereira, M. A.; Li, K.; Kramer, P. M. (1997). Promotion by mixtures of dichloroacetic acid and
trichloroacetic acid of N-methyl-N-nitrosourea-initiated cancer in the liver of female B6C3F1
mice. Cancer Lett, 115, 15-23. htto://dx.doi.org/10.1016/S0304-3835(97)04699-5
Pereira, M. A. and Phelps, J. B. (1996). Promotion by dichloroacetic acid and trichloroacetic acid of N-
methyl-N-nitrosourea-initiated cancer in the liver of female B6C3F1 mice. Cancer Lett, 102, 133-
141. http://dx.doi.org/10.1016/0304-3835(96)04156-0
Perkins, N. D. (2004). NF-kappaB: tumor promoter or suppressor. Trends Cell Biol, 14, 64-69.
http://dx.doi.Org/10.1016/i.tcb.2003.12.004
Perocco, P.; Bolognesi, S.; Alberghini, W. (1983). Toxic activity of seventeen industrial solvents and
halogenated compounds on human lymphocytes cultured in vitro. Toxicol Lett, 16, 69-75.
http://dx.doi.org/10.1016/0378-4274(83)90012-7
Perrin, M. C.; Opler, M. G.; Harlap, S.; Harkavy-Friedman, J.; Kleinhaus, K.; Nahon, D., . . . Malaspina,
D. (2007). Tetrachloroethylene exposure and risk of schizophrenia: offspring of dry cleaners in a
population birth cohort, preliminary findings. Schizophr Res, 90, 251-254.
Pesch, B.; Haerting, J.; Ranft, U.; Klimpel, A.; Oelschlagel, B.; Schill, W. (2000a). Occupational risk
factors for urothelial carcinoma: Agent-specific results from a case-control study in Germany. Int
J Epidemiol, 29, 238-274. http://www ncbi nlm nih gov/pubmed/10817119
Pesch, B.; Haerting, J.; Ranft, U.; Klimpel, A.; Oelschlagel, B.; Schill, W.; Group, M. S. (2000b).
Occupational risk factors for renal cell carcinoma: Agent-specific results from a case-control
study in Germany. Int J Epidemiol, 29, 1014-1024.
Peters, J. M.; Cattley, R. C.; Gonzalez, F. J. (1997). Role of PPAR alpha in the mechanism of action of
the nongenotoxic carcinogen and peroxisome proliferator Wy-14,643. Carcinogenesis, 18, 2029-
2033.
Peters, J. M.; Cheung, C.; Gonzalez, F. J. (2005). Peroxisome proliferator-activated receptor-alpha and
liver cancer: where do we stand. J Mol Med, 83, 774-785. http://dx.doi.org/10.1007/sQ0109-005-
0678-9
Pezzagno, G.; Imbriani, M.; Ghittori, S.; Capodaglio, E. (1988). Urinary concentration, environmental
concentration, and respiratory uptake of some solvents: effect of the work load. Am Ind Hyg
Assoc J, 49, 546-552.
Philip, B. K.; Mumtaz, M. M.; Latendresse, J. R.; Mehendale, H. M. (2007). Impact of repeated exposure
on toxicity of perchloroethylene in Swiss Webster mice. Toxicology, 232, 1-14.
http://dx.doi.Org/10.1016/i.tox.2006.12.018
Physician's desk reference. (1981).
Plewa, M. J.; Kargalioglu, Y.; Vankerk, D.; Minear, R. A.; Wagner, E. D. (2002). Mammalian cell
cytotoxicity and genotoxicity analysis of drinking water disinfection by-products. Environ Mol
Mutagen, 40, 134-142. http://dx.doi.org/10.1002/em.10Q92
Poet, T. S.; Weitz, K. K.; Gies, R. A.; Edwards, J. A.; Thrall, K. D.; Corley, R. A., . . . Wester, R. C.
(2002). PBPK modeling of the percutaneous absorption of perchloroethylene from a soil matrix in
rats and humans. Toxicol Sci, 67, 17-31.
Pohl, H. R.; Roney, N.; Wilbur, S.; Hansen, H.; De Rosa, C. T. (2003). Six interaction profiles for simple
mixtures. Chemosphere, 53, 183-197.
Pohlabeln, H.; Boffetta, P.; Ahrens, W.; Merletti, F.; Agudo, A.; Benhamou, E., . . . Jockel, K. H. (2000).
Occupational risks for lung cancer among nonsmokers. Epidemiology, 11, 532-538.
http://www.ncbi.nlm.nih.gov/pubmed/10955405
Popp, W.; Muller, G.; Baltes-Schmitz, B.; Wehner, B.; Vahrenholz, C.; Schmieding, W., . . . Norpoth, K.
(1992). Concentrations of tetrachloroethene in blood and trichloroacetic acid in urine in workers
and neighbours of dry-cleaning shops. Int Arch Occup Environ Health, 63, 393-395.
Potter, C. L.; Chang, L. W.; DeAngelo, A. B.; Daniel, F. B. (1996). Effects of four trihalomethanes on
DNA strand breaks, renal hyaline droplet formation and serum testosterone in male F-344 rats.
Cancer Lett, 106, 235-242. http://www.ncbi .nlm.nih. gov/pubmed/8 844978
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-46 DRAFT—DO NOT CITE OR QUOTE
-------
Price, P. J.; Hassett, C. M.; Mansfield, J. I. (1978). Transforming activities of trichloroethylene and
proposed industrial alternatives. In Vitro Cell Dev Biol Anim, 14, 290-293.
Price, R. G.; Taylor, S. A.; Chivers, I.; Arce-Tomas, M.; Crutcher, E.; Franchini, I., . . . Eisenberger, U.
(1996). Development and validation of new screening tests for nephrotoxic effects. Hum Exp
Toxicol, 1, S10-S19.
Pukkala, E.; Martinsen, J.; Lynge, E.; Gunnarsdottir, H.; Sparen, P.; Tryggvadottir, L., . . . Kjaerheim, K.
(2009). Occupation and cancer - follow-up of 15 million people in five Nordic countries. Acta
Oncol, 48, 646-790. http://dx.doi.org/10.1080/028418609Q2913546
Quondamatteo, F.; Schulz, T. G.; Bunzel, N.; Hallier, E.; Herken, R. (1998). Immunohistochemical
localization of glutathione S-transferase-Tl in murine kidney, liver, and lung. Histochem Cell
Biol, 110,417-423.
Rachootin, P. and Olsen, J. (1983). The risk of infertility and delayed conception associated with
exposures in the Danish workplace. J Occup Med, 25, 394-402.
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. http://dx.doi.org/10.1097/JOM.0b013e3181845f7f
Raijmakers, M. T. M.; Steegers, E. A. P.; Peters, W. H. M. (2001). Glutathione S-transferases and thiol
concentrations in embryonic and early fetal tissues. Hum Reprod, 16, 2445-2450.
http: //www .ncbi .nlm .nih. gov/pubmed/11679536
Raisanen, J.; Niemela, R.; Rosenberg, C. (2001). Tetrachloroethylene emissions and exposure in dry
cleaning. J Air Waste Manag Assoc, 51, 1671-1675.
http://www.ncbi.nlm.nih.gov/pubmed/15666472
Rampy, L. W.; Quast, J. F.; Balmer, M. F.; Leong, B. K. J.; Gehring, P. J. (1978). Results of a long-term
inhalation toxicity study on rats of a perchloroethylene in rats (tetrachloroethylene) formulation.
Midland, MI: Dow Chemical.
Rao, H. V. and Brown, D. R. (1993). A physiologically based pharmacokinetic assessment of
tetrachloroethylene in groundwater for a bathing and showering determination. Risk Anal, 13, 37-
49. http://dx.doi.org/10.1111/i. 1539-6924.1993,tb00727.x
Raucy, J. L. (1995). Risk assessment: Toxicity from chemical exposure resulting from enhanced
expression of CYP2E1. Toxicology, 105, 217-224. http://dx.doi.org/10.1016/030Q-
483X(95)03216-3
Raunio, H.; HusgafVel-Pursiainen, K.; Anttila, S.; Hietanen, E.; Hirvonen, A.; Pelkonen, O. (1995).
Diagnosis of polymorphisms in carcinogen-activating and inactivating enzymes and cancer
susceptibility—a review. Gene, 159, 113-121.
Rea, W. J.; Ross, G. H.; Johnson, A. R.; Smiley, R. E.; Fenyes, E. J. (1991). Chemical sensitivity in
physicians. Bol Asoc Med P R, 83, 383-388. http://www.ncbi.nlm.nih.gov/pubmed/1807271
Recommended policy on control of volatile organic compounds, 42 35314-35316 (1977).
Reddy, J. K.; Reddy, M. K.; Usman, M. I.; Lalwani, N. D.; Rao, M. S. (1986). Comparison of hepatic
peroxisome proliferative effect and its implication for hepatocarcinogenicity of phthalate esters,
di(2-ethylhexyl) phthalate, and di(2-ethylhexyl) adipate with a hypolipidemic drug. Environ
Health Perspect, 65, 317-327.
Regan, D. (1989). Human brain eletrophysiology. New York: Elsevier Science.
Reichert, D.; Neudecker, T.; Spengler, U.; Henschler, D. (1983). Mutagenicity of dichloroacetylene and
its degradation products trichloroacetyl chloride, trichloroacryloyl chloride and
hexachlorobutadiene. Mutat Res Genet Toxicol Environ Mutagen, 117, 21-29.
http://dx.doi.org/10.1016/0165-1218(83)90149-0
Reitz, R. H.; Gargas, M. L.; Mendrala, A. L.; Schumann, A. M. (1996). In vivo and in vitro studies of
perchloroethylene metabolism for physiologically based pharmacokinetic modeling in rats, mice,
and humans. Toxicol Appl Pharmacol, 136, 289-306. http://dx.doi.org/10.1006/taap.1996.0Q36
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-47 DRAFT—DO NOT CITE OR QUOTE
-------
Reitz, R. H.; Mendrala, A. L.; Guengerich, F. P. (1989). In vitro metabolism of methylene chloride in
human and animal tissues: use in physiologically based pharmacokinetic models. Toxicol Appl
Pharmacol, 97, 230-246.
Renwick, A. G. (1998). Toxicokinetics in infants and children in relation to the ADI and TDI. Food Addit
Contam, 15, 17-35. http://www.ncbi.nlm.nih.gov/pubmed/9602909
Reulen, R.; Kellen, E.; Buntinx, F.; Zeegers, M. (2007). Bladder cancer and occupation: a report from the
Belgian case-control study on bladder cancer risk. Am J Ind Med, 50, 449-454.
http://dx.doi.org/10.1002/aiim.2Q469
Reya, T.; Morrison, S. J.; Clarke, M. F.; Weissman, I. L. (2001). Stem cells, cancer, and cancer stem
cells. Nature, 414, 105-111. http://dx.doi.org/10.1038/35102167
Rhomberg, L. (1992). A cross-species scaling factor for carcinogen risk assessment based on equivalence
of mg/kg3/4/day. 57, 24152-24173.
Richiardi, L.; Boffetta, P.; Simonato, L.; Forastiere, F.; Zambon, P.; Fortes, C., . . . Merletti, F. (2004).
Occupational risk factors for lung cancer in men and women: a population-based case-control
study in Italy. Cancer Causes Control, 15, 285-294.
http://dx.doi.Org/10.1023/B:CACO.0000024223.91059.ed
Riihimaki, V. and Pfaffli, P. (1978). Percutaneous absorption of solvent vapors in man. Scand J Work
Environ Health, 4, 73-85.
Ripp, S. L.; Overby, L. H.; Philpot, R. M.; Elfarra, A. A. (1997). Oxidation of cysteine S-conjugates by
rabbit liver microsomes and cDNA-expressed flavin-containing mono-oxygenases: studies with
S-(l,2-dichlorovinyl)-L-cysteine, S-(l,2,2-trichlorovinyl)-L-cysteine, S-allyl-L-cysteine, and S-
benzyl-L-cysteine. Mol Pharmacol, 51, 507-515. http://www.ncbi.nlm.nih.gov/pubmed/9058607
Rodilla, V.; Benzie, A. A.; Veitch, J. M.; Murray, G. I.; Rowe, J. D.; Hawksworth, G. M. (1998).
Glutathione S-transferases in human renal cortex and neoplastic tissue: enzymatic activity,
isoenzyme profile and immunohistochemical localization. Xenobiotica, 28, 443-456.
Rodriguez, C. E.; Mahle, D. A.; Gearhart, J. M.; Mattie, D. R.; Lipscomb, J. C.; Cook, R. S.; Barton, H.
A. (2007). Predicting age-appropriate pharmacokinetics of six volatile organic compounds in the
rat utilizing physiologically based pharmacokinetic modeling. Toxicol Sci, 98, 43-56.
http: //dx. doi. org/10.1093/toxsci/kfmO 82
Roldan-Aijona, T.; Garcia-Pedrajas, M. D.; Luque-Romero, F. L.; Hera, C.; Pueyo, C. (1991). An
association between mutagenicity of the Ara test of Salmonella typhimurium and carcinogenicity
in rodents for 16 halogenated aliphatic hydrocarbons. Mutagenesis, 6, 199-205.
http://dx.doi.Org/10.1093/mutage/6.3.199
Rose, M. G. and Berliner, N. (2004). T-cell large granular lymphocyte leukemia and related disorders.
Oncologist, 9, 247-258. http://www.ncbi.nlm.nih.gov/pubmed/15169980
Rosengren, L. E.; Kjellstrand, P.; Haglid, K. G. (1986). Tetrachloroethylene: Levels of DNA and S-100 in
the gerbil CNS after chronic exposure. Neurobehav Toxicol Teratol, 8, 201-206.
http: //www .ncbi .nlm .nih. gov/pubmed/3 713968
Rubins, H. B.; Robins, S. J.; Collins, D.; Fye, C. L.; Anderson, J. W.; Elam, M. B., . . . Wittes, J. (1999).
Gemfibrozil for the secondary prevention of coronary heart disease in men with low levels of
high-density lipoprotein cholesterol. Veterans Affairs High-Density Lipoprotein Cholesterol
Intervention Trial Study Group. N Engl J Med, 341, 410-418.
http://dx.doi.org/10.1056/NEJM1999080534106Q4
Rubins, H. B.; Robins, S. J.; Iwane, M. K.; Boden, W. E.; Elam, M. B.; Fye, C. L., . . . Wittes, J. T.
(1993). Rationale and design of the Department of Veterans Affairs High-Density Lipoprotein
Cholesterol Intervention Trial (HIT) for secondary prevention of coronary artery disease in men
with low high-density lipoprotein cholesterol and desirable low-density lipoprotein cholesterol.
Am J Cardiol, 71, 45-52. http://www.ncbi.nlm.nih.gov/pubmed/8420235
Ruder, A. M.; Ward, E. M.; Brown, D. P. (1994). Cancer mortality in female and male dry-cleaning
workers. J Occup Med, 36, 867-874. http://www.ncbi.nlm.nih.gov/pubmed/7807267
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-48 DRAFT—DO NOT CITE OR QUOTE
-------
Ruder, A. M.; Ward, E. M.; Brown, D. P. (2001). Mortality in dry-cleaning workers: An update. Am J Ind
Med, 39, 121-132. http://dx.doi.org/10.1002/1097-0274(200102)39:2<121::AID-
AJIM1000>3.0.CQ:2-H
Russo, A. and Levis, A. (1992a). Detection of aneuploidy in male germ cells of mice by means of a
meiotic micronucleus assay. Mutat Res, 281, 187-191.
http://www.ncbi.nlm.nih.gov/pubmed/1371841
Russo, A. and Levis, A. G. (1992b). Further evidence for the aneuploidogenic properties of chelating
agents: induction of micronuclei in mouse male germ cells by EDTA. Environ Mol Mutagen, 19,
125-131. http://www.ncbi.nlm.nih.gov/pubmed/1541253
Russo, A.; Pacchierotti, F.; Metalli, P. (1984). Nondisjunction induced in mouse spermatogenesis by
chloral hydrate, a metabolite of trichloroethylene. Environ Mutagen, 6, 695-703.
http: //www .ncbi .nlm .nih. gov/pubmed/6479114
Rusyn, I.; Peters, J. M.; Cunningham, M. L. (2006). Modes of action and species-specific effects of di-(2-
ethylhexyl)phthalate in the liver. Crit Rev Toxicol, 36, 459-479.
http://dx.doi.org/10.1080/10408440600779Q65
Saillenfait, A. M.; Langonne, I.; Sabate, J. P. (1995). Developmental toxicity of trichloroethylene,
tetrachloroethylene and four of their metabolites in rat whole embryo culture. Arch Toxicol, 70,
71-82. http://dx.doi.org/10.1007/BF02733666
Sakamoto, N. (1976). [Metabolism of tetrachloroethylene in guinea pigs (author's transl)]. Sangyo Igaku,
18, 11-16.
Sallmen, M.; Lindbohm, M. L.; 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, 24-30. http://www.ncbi.nlm.nih.gov/pubmed/9536159
Sallmen, M.; Lindbohm, M. L.; Kyyronen, P.; Nykyri, E.; Anttila, A.; Taskinen, H.; Hemminki, K.
(1995). Reduced fertility among women exposed to organic solvents. Am J Ind Med, 27, 699-
713. http://dx.doi.org/10.1002/aiim.47002705Q6
Sarangapani, R.; Gentry, P. R.; Covington, T. R.; Teeguarden, J. G.; 3rd, C. H. (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. http://dx.doi.org/10.1080/713857276
Sata, F.; Sapone, A.; Elizondo, G.; Stacker, P.; Miller, V. P.; Zheng, W., . . . Gonzalez, F. J. (2000).
CYP3A4 allelic variants with amino acid substitutions in exons 7 and 12: evidence for an allelic
variant with altered catalytic activity. Clin Pharmacol Ther, 67, 48-56.
Sato, T.; Mukaida, M.; Ose, Y.; Nagase, H.; Ishikawa, T. (1985). Mutagenicity of chlorinated products
from soil humic substances. Sci Total Environ, 46, 229-241. http://dx.doi.org/10.1016/0Q48-
9697(85)90296-7
Savolainen, H.; Pfaffli, P.; Tengen, M.; Vainio, H. (1977a). Trichloroethylene and 1,1,1-trichloroethane:
effects on brain and liver after five days intermittent inhalation. Arch Toxicol, 38, 229-237.
http: //www .ncbi .nlm .nih. gov/pubmed/5 78726
Savolainen, H.; Pfaffli, P.; Tengen, M.; Vainio, H. (1977b). Biochemical and behavioural effects of
inhalation exposure to tetrachlorethylene and dichlormethane. J Neuropathol Exp Neurol, 36,
941-949. http://www ncbi nlm nih gov/pubmed/925719
Sax, N. I. and Lewis, R. J. S. (1987). Hazardous chemicals desk reference. New York, NY: Van Nostrand
Reinhold.
Scharre, J. E.; Cotter, S. A.; Block, S. S.; Kelly, S. A. (1990). Normative contrast sensitivity data for
young children. Optometry and Vision Science, 67, 826-832.
http ://www .ncbi .nlm .nih. gov/pubmed/225 0891
Schatten, H. and 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, 9-
20.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-49 DRAFT—DO NOT CITE OR QUOTE
-------
Scheff, P. A. and Wadden, R. A. (1993). Receptor modeling of volatile organic compounds 1 Emission
inventory and validation. Environ Sci Technol, 27, 617-625.
http://dx.doi.org/10.1021/es00041a0Q5
Schenk, M.; Purdue, M.; Colt, J.; Hartge, P.; Blair, A.; Stewart, P., . . . Severson, R. (2009).
Occupation/industry and risk of non-Hodgkin's lymphoma in the United States. Occup Environ
Med, 66, 23-31. http://dx.doi.org/10.1136/oem.2007.036723
Scherr, P. A.; Hutchison, G. B.; Neiman, R. S. (1992). Non-Hodgkin's lymphoma and occupational
exposure. Cancer Res, 52, 5503s-5509s. http ://www.ncbi.nlm.nih.gov/pubmed/1394164
Schlehofer, B.; Heuer, C.; Blettner, M.; Niehoff, D.; Wahrendorf, J. (1995). Occupation, smoking and
demographic factors, and renal cell carcinoma in Germany. Int J Epidemiol, 24, 51-57.
http ://www .ncbi .nlm .nih. gov/pubmed/7797356
Schlesselman, J. J. and Stolley, P. D. (1982). Case control studies: design, conduct, analysis.
Schlichting, L. M.; Wright, P. F.; Stacey, N. H. (1992). Effects of tetrachloroethylene on hepatic and
splenic lymphocytotoxic activities in rodents. Toxicol Ind Health, 8, 255-266.
http: //www .ncbi .nlm .nih. gov/pubmed/1455436
Schoenberg, J.; Stemhagen, A.; Mogielnicki, A.; Altman, R.; Abe, T.; Mason, T. (1984a). Case-control
study of bladder cancer in New Jersey. I. Occupational exposures in white males. J Natl Cancer
Inst, 72, 973-981. http ://www.ncbi .nlm .nih.gov/pubmed/65 85596
Schoenberg, J. B.; Stemhagen, A.; Mogielnicki, A. P.; Altman, R.; Abe, T.; Mason, T. J. (1984b). Case-
control study of bladder cancer in New Jersey: I. Occupational exposures in white males. J Natl
Cancer Inst, 72, 973-981. http://www.ncbi.nlm.nih.gov/pubmed/6585596
Schreiber, J. S. (1993). Predicted infant exposure to tetrachloroethene in human breastmilk. Risk Anal,
13, 515-524. http://dx.doi.Org/10.llll/i.1539-6924.1993.tb00010.x
Schreiber, J. S. (1997). Transport of organic chemicals to breast milk: Tetrachloroethene case study. In S.
Kacew & G. Lambert (Eds.), Environmental toxicology and pharmacology of human
development (pp. 95-143). Washington, DC: Taylor and Francis.
Schreiber, J. S.; House, S.; Prohonic, E.; Smead, G.; Hudson, C.; Styk, M.; Lauber, J. (1993). An
investigation of indoor air contamination in residences above dry cleaners. Risk Anal, 13, 335-
344. http://dx.doi.Org/10.llll/i.1539-6924.1993.tb01085.x
Schreiber, J. S.; Hudnell, H. K.; Geller, A. M.; House, D. E.; Aldous, K. M.; Force, M. S., . . . Parker, J.
C. (2002). Apartment residents' and day care workers' exposures to tetrachloroethylene and
deficits in visual contrast sensitivity. Environ Health Perspect, 110, 655-664.
http: //www .ncbi .nlm .nih. gov/pubmed/12117642
Schumann, A. M.; Quast, J. F.; Watanabe, P. G. (1980). The pharmacokinetics and macromolecular
interactions of perchloroethylene in mice and rats as related to oncogenicity. Toxicol Appl
Pharmacol, 55, 207-219. http://dx.doi.org/10.1016/0041-008X(80)90082-4
Schwetz, B. A.; Leong, B. K. J.; Gehring, P. J. (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.
http://dx.doi.org/10.1016/0041-008X(75')90197-0
Scott, C. S.; Richards, S. J.; Sivakumaran, M. (1994). Disorders of large granular lymphocytes and
natural killer-associated cells. Blood, 83, 301-303.
http://www.ncbi.nlm.nih.gov/pubmed/8274746
Seeber, A. (1989). Neurobehavioral toxicity of long-term exposure to tetrachloroethylene. Neurotoxicol
Teratol, 11, 579-583. http://dx.doi.org/10.1016/0892-0362(89)90041-X
Seidel, H. J.; Weber, L.; Barthel, E. (1992). Hematological toxicity of tetrachloroethylene in mice. Arch
Toxicol, 66, 228-230. http://dx.doi.org/10.1007/BF01974Q21
Seidler, A.; Mohner, M.; Berger, J.; Mester, B.; Deeg, E.; Eisner, G., . . . Becker, N. (2007). Solvent
exposure and malignant lymphoma: A population-based case-control study in Germany. J Occup
Med Toxicol, 2, 2. http://dx.doi.org/10.1186/1745-6673-2-2
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-50 DRAFT—DO NOT CITE OR QUOTE
-------
Seiji, K.; Jin, C.; Watanabe, T.; Nakatsuka, H.; Ikeda, M. (1990). Sister chromatid exchanges in
peripheral lymphocytes of workers exposed to benzene, trichloroethylene, or tetrachloroethylene,
with reference to smoking habits. Int Arch Occup Environ Health, 62, 171-176.
http://dx.doi.org/10.1007/BF0Q383594
Selden, A. and Ahlborg, G. (2011). Cancer morbidity in Swedish dry-cleaners and laundry workers:
historically prospective cohort study. Int Arch Occup Environ Health, 84, 435-443.
http://dx.doi.org/10.1007/s00420-010-Q582-7
Seo, M.; Ikeda, K.; Okamura, T.; Kida, K.; Satoh, M.; Inagaki, N., . . . Nagase, H. (2008). Enhancing
effect of chlorinated organic solvents on histamine release and inflammatory mediator
production. Toxicology, 243, 75-83. http://dx.doi.Org/10.1016/i.tox.2007.09.024
Sertznig, P.; Seifert, M.; Tilgen, W.; Reichrath, J. (2007). Present concepts and future outlook: Function
of peroxisome proliferator-activated receptors (PPARs) for pathogenesis, progression, and
therapy of cancer. J Cell Physiol, 212, 1-12. http://dx.doi.org/10.1002/icp.20998
Sexton, K.; Adgate, J. L.; Church, T. R.; Ashley, D. L.; Needham, L. L.; Ramachandran, G., . . . Ryan, A.
D. (2005). Children's exposure to volatile organic compounds as determined by longitudinal
measurements in blood. Environ Health Perspect, 113, 342-349.
http://dx.doi.org/10.1289/ehp.7412
Shafer, T. J.; Bushnell, P. J.; Benignus, V. A.; Woodward, J. J. (2005). Perturbation of voltage-sensitive
Ca2+ channel function by volatile organic solvents. J Pharmacol Exp Ther, 315, 1109-1118.
http://dx.doi.org/10.1124/ipet. 105.090027
Shao, J.; Stapleton, P. L.; Lin, Y. S.; Gallagher, E. P. (2007). Cytochrome p450 and glutathione s-
transferase mRNA expression in human fetal liver hematopoietic stem cells. Drug Metab Dispos,
35, 168-175. http://dx.doi.org/10.1124/dmd.106.012757
Sharanjeet-Kaur; Mursyid, A.; Kamaruddin, A.; Ariffin, A. (2004). Effect of petroleum derivatives and
solvents on colour perception. Clin Exp Optom, 87, 339-343. http://dx.doi.Org/10.l 11 l/i.1444-
0938.2004.tb05064.x
Sheldon, L. S.; Handy, R. W.; Hartwell, T. D.; Leininger, C.; Zelon, H. (1985). Human exposure
assessment to environmental chemicals: nursing mothers study. Final report. U.S. Environmental
Protection Agency.
Sher, T.; Yi, H. F.; McBride, O. W.; Gonzalez, F. J. (1993). cDNA cloning, chromosomal mapping, and
functional characterization of the human peroxisome proliferator activated receptor.
Biochemistry, 32, 5598-5604.
Shevchenko, V. V.; Grinikh, L. I.; Alekseenok, A. Y. (1985). Clastogenic effect of oxalic acid and its
specificity. Sov Genet, 21, 606-610.
Shimada, T.; Swanson, A. F.; Leber, P.; Williams, G. M. (1985). Activities of chlorinated ethane and
ethylene compounds in the Salmonella/rat microsome mutagenesis and rat hepatocyte/DNA
repair assays under vapor phase exposure conditions. Cell Biol Toxicol, 1, 159-179.
http://dx.doi.org/10.1007/BF0012Q162
Shirasu, Y.; Moriya, M.; Kato, K.; Furuhashi, A.; Kada, T. (1976). Mutagenicity screening of pesticides
in the microbial system. Mutat Res, 40, 19-30. http://dx.doi.org/10.1016/0165-1218(76)90018-5
Shu, X. O.; Stewart, P.; Wen, W. Q.; Han, D.; Potter, J. D.; Buckley, J. D., . . . Robison, L. L. (1999).
Parental occupational exposure to hydrocarbons and risk of acute lymphocytic leukemia in
offspring. Cancer Epidemiol Biomarkers Prev, 8, 783-791.
http: //www .ncbi .nlm .nih. gov/pubmed/104983 97
Siemiatycki, J.; Wacholder, S.; Richardson, L.; Dewar, R.; Gerin, M. (1987). Discovering carcinogens in
the occupational environment. Methods of data collection and analysis of a large case-referent
monitoring system. Scand J Work Environ Health, 13, 486-492.
http: //www .ncbi .nlm .nih. gov/pubmed/3433050
Silverman, A. (1988). An ethologist's approach to behavioural toxicology. Neurotoxicol Teratol, 10, 85-
92. http://www.ncbi.nlm.nih.gov/pubmed/3398827
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-51 DRAFT—DO NOT CITE OR QUOTE
-------
Silverman, D.; Levin, L.; Hoover, R. (1990). Occupational risks of bladder cancer among white women in
the United States. Am J Epidemiol, 132, 453-461. http://www.ncbi.nlm.nih.gov/pubmed/2389750
Silverman, D. T.; Levin, L. I.; Hoover, R. N. (1989a). Occupational risks of bladder cancer in the United
States: II. Nonwhite men. J Natl Cancer Inst, 81, 1480-1483.
http://dx.doi.org/10.1093/inci/81.19.1480
Silverman, D. T.; Levin, L. I.; Hoover, R. N.; Hartge, P. (1989b). Occupational risks of bladder cancer in
the United States: I. White men. J Natl Cancer Inst, 81, 1472-1480.
http://dx.doi.org/10.1093/jnci/81.19.1472
Singh, M. and Sinha, U. (1976). Chloral hydrate induced haploidization in Aspergillus nidulans.
Experientia, 32, 1144-1145.
Singh, M. and Sinha, U. (1979). Mitotic haploidization and growth of Aspergillus nidulans on media
containing chloral hydrate. 14, 1-4.
Skender, L. J.; Karacic, V.; Prpic-Majic, D. (1991). A comparative study of human levels of
trichloroethylene and tetrachloroethylene after occupational exposure. Arch Environ Occup
Health, 46, 174-178. http://www.ncbi.nlm.nih.gov/pubmed/2039273
Skowron, J.; Miranowicz-Dzierzawska, K.; Zapor, L.; Golofit-Szymczak, M.; Starek, A. (2001).
Interactions of some organic solvents: hydrocarbons and chloroalkene. Int J Occup Saf Ergon, 7,
35-47.
Slikker, J. W.; Andersen, M. E.; Bogdanffy, M. S.; Bus, J. S.; Cohen, S. D.; Conolly, R. B., . . . Wallace,
K. (2004). Dose-dependent transitions in mechanisms of toxicity: Case studies. Toxicol Appl
Pharmacol, 201, 226-294. http://dx.doi.Org/10.1016/i.taap.2004.06.027
Smith, E. M.; Miller, E. R.; Woolson, R. F.; Brown, C. K. (1985). Bladder cancer risk among laundry
workers, dry cleaners, and others in chemically-related occupations. J Occup Med, 27, 295-297.
http: //www .ncbi .nlm .nih. gov/pubmed/3 998883
Smith, M. K.; Randall, J. L.; Read, E. J.; Stober, J. A. (1989). Teratogenic activity of trichloroacetic acid
in the rat. Teratology, 40, 445-451. http://dx.doi.org/10.1002/tera.14204005Q6
Sofuni, T.; Hayashi, M.; Matsuoka, A.; Sawada, M.; Hatanaka, M.; M Jr, I. (1985). [Mutagenicity tests on
organic chemical contaminants in city water and related compounds. II. Chromosome aberration
tests in cultured mammalian cells]. Kokuritsu Iyakuhin Shokuhin Eisei Kenkyusho Hokoku, 103,
64-75. http://www.ncbi .nlm.nih. gov/pubmed/3 830315
Sokol, L. and Loughran, T. P. J. (2006). Large granular lymphocyte leukemia. Oncologist, 11, 263-273.
http://dx.doi.org/10.1634/theoncologist.ll-3-263
Solet, D. and Robins, T. G. (1991). Renal function in dry cleaning workers exposed to perchloroethylene.
AmJInd Med, 20, 601-614.
Solet, D.; Robins, T. G.; Sampaio, C. (1990). Perchloroethylene exposure assessment among dry cleaning
workers. AIHA J, 51, 566-574.
Solleveld, H. A.; Haseman, J. K.; McConnell, E. E. (1984). Natural history of body weight gain, survival,
and neoplasia in the F344 rat. J Natl Cancer Inst, 72, 929-940.
http://www.ncbi.nlm.nih.gov/pubmed/6584668
Solvent exposure: a wolf in sheep's clothing? Recognition and assessment from a clinical perspective.
(2007). J Occup Environ Med, 49, 813-815.
http://dx.doi.org/10.1097/01.iom.000025162Q.62826.8d
Sonnenfeld, N.; Hertz-Picciotto, I.; Kaye, W. E. (2001). Tetrachloroethylene in drinking water and birth
outcomes at the US Marine Corps Base at Camp Lejeune, North Carolina. Am J Epidemiol, 154,
902-908. http://www.ncbi.nlm.nih.gov/pubmed/11700244
Sora, S. and Agostini Carbone, M. L. (1987). Chloral hydrate, methylmercury hydroxide and ethidium
bromide affect chromosomal segregation during meiosis of Saccharomyces cerevisiae. Mutat Res
Lett, 190, 13-17. http://dx.doi.org/10.1016/0165-7992(87)90075-3
Sparrow, G. P. (1977). A connective tissue disorder similar to vinyl chloride disease in a patient exposed
to perchlorethylene. Clin Exp Dermatol, 2, 17-22.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-52 DRAFT—DO NOT CITE OR QUOTE
-------
Speerschneider, P. and Dekant, W. (1995). Renal tumorigenicity of 1,1-dichloroethene in mice: the role of
male-specific expression of cytochrome P450 2E1 in the renal bioactivation of 1,1-
dichloroethene. Toxicol Appl Pharmacol, 130, 48-56.
Spencer, H. B.; Hussein, W. R.; Tchounwou, P. B. (2002). Effects of tetrachloroethylene on the viability
and development of embryos of the Japanese medaka, Oryzias latipes. Arch Environ Contam
Toxicol, 42, 463-469. http://dx.doi.org/10.1007/s00244-001-005Q-l
Spinatonda, G.; Colombo, R.; Capodaglio, E. M.; Imbriani, M.; Pasetti, C.; Minuco, G.; Pinelli, P. (1997).
Processes of speech production: Application in a group of subjects chronically exposed to organic
solvents (II). G Ital Med Lav Ergon, 19, 85-88. http://www.ncbi.nlm.nih.gov/pubmed/9463050
Spirtas, R.; Stewart, P. A.; Lee, J. S.; Marano, D. E.; Forbes, C. D.; Grauman, D. J., . . . Cohen, J. L.
(1991). Retrospective cohort mortality study of workers at an aircraft maintenance facility. I.
Epidemiological results. Br J Ind Med, 48, 515-530. http://dx.doi.Org/10.1136/oem.48.8.515
Stanley, L. A.; Blackburn, D. R.; Devereaux, S.; Foley, J.; Lord, P. G.; Maronpot, R. R., . . . Anderson,
M. W. (1994). Ras mutations in methylclofenapate-induced B6C3F1 and C57BL/10J mouse liver
tumours. Carcinogenesis, 15, 1125-1131. http://www.ncbi.nlm.nih.gov/pubmed/8020144
Stauber, A. J. and Bull, R. J. (1997). Differences in phenotype and cell replicative behavior of hepatic
tumors induced by dichloroacetate (DCA) and trichloroacetate (TCA). Toxicol Appl Pharmacol,
144, 235-246. http://dx.doi.org/10.1006/taap.1997.8159
Staunton, D.; Young, S.; Groves, P. (1980). The effect of long-term amphetamine administration on the
activity of dopaminergic neurons of the substantia nigra. Brain Res, 188, 107-117.
http://www.ncbi.nlm.nih.gov/pubmed/6989446
Steineck, G.; Plato, N.; Gerhardsson, M.; Norell, S. E.; Hogstedt, C. (1990). Increased risk of urothelial
cancer in Stockholm during 1985-87 after exposure to benzene and exhausts. Int J Cancer, 45,
1012-1017. http://dx.doi.org/10.1002/iic.29104506Q5
Stemhagen, A.; Slade, J.; Altman, R.; Bill, J. (1983). Occupational risk factors and liver cancer: A
retrospective case-control study of primary liver cancer in New Jersey. Am J Epidemiol, 117,
443-454. http://www ncbi nlm nih gov/pubmed/6837558
Stephens, E. A.; Taylor, J. A.; Kaplan, N.; Yang, C. H.; Hsieh, L. L.; Lucier, G. W.; Bell, D. A. (1994).
Ethnic variation in the CYP2E1 gene: polymorphism analysis of 695 African-Americans,
European-Americans and Taiwanese. Pharmacogenetics, 4, 185-192.
Stern, F.; Lehman, E.; Ruder, A. (2001). Mortality among unionized construction plasterers and cement
masons. Am J Ind Med, 39, 373-388. http://www.ncbi.nlm.nih.gov/pubmed/11323787
Stevens, J. (1985a). Cysteine conjugate beta-lyase activities in rat kidney cortex: subcellular localization
and relationship to the hepatic enzyme. Biochem Biophys Res Commun, 129, 499-504.
http: //www .ncbi .nlm .nih. gov/pubmed/4015643
Stevens, J.; Ayoubi, N.; Robbins, J. (1988). The role of mitochondrial matrix enzymes in the metabolism
and toxicity of cysteine conjugates. J Biol Chem, 263, 3395-3401.
http://www.ncbi.nlm.nih.gov/pubmed/3343250
Stevens, J.; Hatzinger, P.; Hayden, P. (1989). Quantitation of multiple pathways for the metabolism of
nephrotoxic cysteine conjugates using selective inhibitors of L-alpha-hydroxy acid oxidase (L-
amino acid oxidase) and cysteine conjugate beta-lyase. Drug Metab Dispos, 17, 297-303.
http: //www .ncbi .nlm .nih. gov/pubmed/2568912
Stevens, J. and Jakoby, W. B. (1983). Cysteine conjugate beta-lyase. Mol Pharmacol, 23, 761-765.
Stevens, J.; Robbins, J.; Byrd, R. (1986). A purified cysteine conjugate beta-lyase from rat kidney
cytosol. Requirement for an alpha-keto acid or an amino acid oxidase for activity and identity
with soluble glutamine transaminase K. J Biol Chem, 261, 15529-15537.
http://www.ncbi.nlm.nih.gov/pubmed/3782077
Stevens, J. L. (1985b). Isolation and characterization of a rat liver enzyme with both cysteine conjugate
beta-lyase and kynureninase activity. J Biol Chem, 260, 7945-7950.
http://www.ncbi.nlm.nih.gov/pubmed/4008484
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-53 DRAFT—DO NOT CITE OR QUOTE
-------
Stewart, P. A.; Lee, J. S.; Marano, D. E.; Spirtas, R.; Forbes, C. D.; Blair, A. (1991). Retrospective cohort
mortality study of workers at an aircraft maintenance facility II Exposures and their assessment.
Occup Environ Med, 48, 531-537. http://www ncbi nlm nih gov/pubmed/1878309
Stewart, R. D.; Baretta, E. D.; Dodd, H. C.; Torkelson, T. R. (1970). Experimental human exposure to
tetrachloroethylene. Arch Environ Health, 20, 224-229.
Stewart, R. D. and Dodd, H. C. (1964). Absorption of carbon tetrachloride, trichloroethylene,
tetrachloroethylene, methylene chloride, and 1,1, 1 - trichloroethane through the human skin. Am
Ind Hyg Assoc J, 25, 439-446.
Stewart, R. D.; Erley, D. S.; Schaffer, A. W.; Gay, H. H. (1961). Accidental vapor exposure to anesthetic
concentrations of a solvent containing tetrachloroethylene. Ind Med Surg, 30, 327-330.
Stewart, R. D.; Hake, C. L.; Forster, A. J.; Lubrun, A. J.; Peterson, J. E.; Wu, A. (1974).
Tetrachloroethylene: development of a biologic standard for the industrial worker by breath
analysis. National Institute of Occupational Safety and Health.
Stewart, R. D.; Hake, C. L.; Wu, A.; Kalbfleisch, J.; Newton, P. E.; Marlow, S. K.; Vucicevic-Salama, M.
(1977). Effects of perchloroethylene/drug interaction on behavior and neurological function.
Storm, J. E. and Mazor, K. A. (2004). Update of residential tetrachloroethylene exposure and decreases in
visual contrast sensitivity. Environ Health Perspect, 112, A862-A864.
http://www.ncbi.nlm.nih.gov/pubmed/15531411
Storm, J. E.; Mazor, K. A.; Aldous, K. M.; Blount, B. C.; Brodie, S. E.; Serle, J. B. (In Press). Visual
Contrast Sensitivity (VCS) in children Exposed to tetrachloroethylene. Arch Environ Occup
Health.
Stott, W. T.; Quast, J. F.; Watanabe, P. G. (1982). The pharmacokinetics and macromolecular interactions
of trichloroethylene in mice and rats. Toxicol Appl Pharmacol, 62, 137-151.
http: //www .ncbi .nlm .nih. gov/pubmed/7064149
Stromberg, P. C. (1985). Large granular lymphocyte leukemia in F344 rats: Model for human T gamma
lymphoma, malignant histiocytosis, and T-cell chronic lymphocytic leukemia. Am J Pathol, 119,
517-519. http://www nchi nlm nih gov/pubmed/3874555
Styles, J. A.; Wyatt, I.; Coutts, C. (1991). Trichloroacetic acid: studies on uptake and effects on hepatic
DNA and liver growth in mouse. Carcinogenesis, 12, 1715-1719.
http: //www .ncbi .nlm .nih. gov/pubmed/1893532
Suarez, L.; Weiss, N. S.; Martin, J. (1989). Primary liver cancer death and occupation in Texas. Am J Ind
Med, 15, 167-175. http://www.ncbi.nlm.nih.gov/pubmed/2729281
Subramoniam, A.; Goel, S. K.; Pandya, K. P.; Seth, P. K. (1989). Influence of trichloroethylene treatment
on phosphoinositides in rat brain. Toxicol Lett, 49, 55-60.
http: //www .ncbi .nlm .nih. gov/pubmed/25 545 3 8
Sukhotina, K. I. (1969). Study of Trichloroethylene Metabolites in Workers Manufacturing
Trichloroethylene and Monochloroacetic Acid. 13, 35-38.
Sung, T.; Chen, P.; Jyuhn-Hsiarn Lee, L.; Lin, Y.; Hsieh, G.; Wang, J. (2007). Increased standardized
incidence ratio of breast cancer in female electronics workers. BMC Public Health, 7, 102.
http://dx.doi.org/10.1186/1471-2458-7-102
Sung, T.; Wang, J.; Chen, P. (2008). Increased risk of cancer in the offspring of female electronics
workers. Reprod Toxicol, 25, 115-119. http://dx.doi.Org/10.1016/i.reprotox.2007.08.004
Sung, T. I.; Wang, J. D.; Chen, P. C. (2009). Increased risks of infant mortality and of deaths due to
congenital malformation in the offspring of male electronics workers. Birth Defects Res A Clin
Mol Teratol, 85, 119-124. http://dx.doi.org/10.1002/bdra.20496
Supplementary data for TCE assessment: Cancer rodents input data. (2011).
Supplementary data for TCE assessment: Cancer rodents model selections. (2011).
Supplementary data for TCE assessment: Cancer rodents plots. (2011).
Supplementary data for TCE assessment: Cancer rodents results. (2011).
Supplementary data for TCE assessment: Cancer rodents uncertainty analysis. (2011).
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-54 DRAFT—DO NOT CITE OR QUOTE
-------
Supplementary data for TCE assessment: Cancer rodents uncertainty CSF-inhalation histograms,
inhalation bioassays. (2011).
Supplementary data for TCE assessment: Cancer rodents uncertainty CSF-inhalation historams, oral
bioassays. (2011).
Supplementary data for TCE assessment: Cancer rodents uncertainty CSF-oral histograms, oral bioassays.
(2011).
Supplementary data for TCE assessment: Hack human population calibration evaluation. (2011).
Supplementary data for TCE assessment: Hack human subject calibration. (2011).
Supplementary data for TCE assessment: Hack mouse population calibration evaluation. (2011).
Supplementary data for TCE assessment: Hack mouse subject calibration. (2011).
Supplementary data for TCE assessment: Hack rat populations calibration evaluation. (2011).
Supplementary data for TCE assessment: Hack rat subject calibration. (2011).
Supplementary data for TCE assessment: Human population example. (2011).
Supplementary data for TCE assessment: Human posteriors by subject. (2011).
Supplementary data for TCE assessment: Mouse population example. (2011).
Supplementary data for TCE assessment: Mouse posteriors by subject. (2011).
Supplementary data for TCE assessment: Non-cancer HECs altPOD plots from rodent inhalation studies.
(2011).
Supplementary data for TCE assessment: Non-cancer HECs altPOD plots from rodent oral studies.
(2011).
Supplementary data for TCE assessment: Non-cancer HECs plots from human inhalation studies. (2011).
Supplementary data for TCE assessment: Non-cancer HECs plots from rodent inhalation studies. (2011).
Supplementary data for TCE assessment: Non-cancer HECs plots from rodent oral studies. (2011).
Supplementary data for TCE assessment: Non-cancer HEDs altPOD plots from rodent inhalation studies.
(2011).
Supplementary data for TCE assessment: Non-cancer HEDs altPOD plots from rodent oral studies.
(2011).
Supplementary data for TCE assessment: Non-cancer HEDs plots from human inhalation studies. (2011).
Supplementary data for TCE assessment: Non-cancer HEDs plots from rodent inhalation studies. (2011).
Supplementary data for TCE assessment: Non-cancer HEDs plots from rodent oral studies. (2011).
Supplementary data for TCE assessment: Non-cancer input data contin. (2011).
Supplementary data for TCE assessment: Non-cancer input data dichot. (2011).
Supplementary data for TCE assessment: Non-cancer plots dichot. (2011).
Supplementary data for TCE assessment: Non-cancer results contin. (2011).
Supplementary data for TCE assessment: Non-cancer results dichot. (2011).
Supplementary data for TCE assessment: Rodents time to tumor results. (2011).
Sverdrup, B.; Kallberg, H.; Bengtsson, C.; Lundberg, I.; Padyukov, L.; Alfredsson, L., . . . Group, E. I. o.
R. A. S. (2005). Association between occupational exposure to mineral oil and rheumatoid
arthritis: results from the Swedish EIRA case-control study. Arthritis Res Ther, 7, R1296-R1303.
http://dx.doi.org/10.1186/arl824
Swanson, G. M. and Burns, P. B. (1995). Cancer incidence among women in the workplace: A study of
the association between occupation and industry and 11 cancer sites. J Occup Environ Med, 37,
282-287. http://www ncbi nlm nih gov/pubmed/7796194
Sweeney, L. M.; Kirman, C. R.; Gargas, M. L.; Dugard, P. H. (2009). Contribution of trichloroacetic acid
to liver tumors observed in perchloroethylene (perc)-exposed mice. Toxicology, 260, 77-83.
http://dx.doi.Org/10.1016/i.tox.2009.03.008
Swenberg, J. A. and Lehman-McKeeman, L. D. (1999). alpha 2-Urinary globulin-associated nephropathy
as a mechanism of renal tubule cell carcinogenesis in male rats. IARC Sci Publ95-118.
http: //www .ncbi .nlm .nih. gov/pubmed/1045 7913
Szakmary, E.; Ungvary, G.; Tatrai, E. (1997). The offspring-damaging effect of tetrachloroethylene in
rats, mice, and rabbits. Central Eur J Occup Env Med, 3, 31-39.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-55 DRAFT—DO NOT CITE OR QUOTE
-------
Tanaka, E. and Breimer, D. (1997). In vivo function tests of hepatic drug-oxidizing capacity in patients
with liver disease. J Clin Pharm Ther, 22, 237-249.
http://www.ncbi.nlm.nih.gov/pubmed/9548204
Tanaka, S. and Ikeda, M. (1968). A method for determination of trichloroethanol and trichloroacetic acid
in urine. Occup Environ Med, 25, 214-219. htto://www ncbi nlm nih gov/pubmed/5690921
Tanaka, T.; Ordovas, J. M.; Delgado-Lista, J.; Perez-Jimenez, F.; Marin, C.; Perez-Martinez, P., . . .
Lopez-Miranda, J. (2007). Peroxisome proliferator-activated receptor alpha polymorphisms and
postprandial lipemia in healthy men. J Lipid Res, 48, 1402-1408.
htto: //dx .doi.org/10.1194/ilr.M700066-JLR200
Tang, M. L.; Kemp, A. S.; Thorburn, J.; Hill, D. J. (1994). Reduced interferon-gamma secretion in
neonates and subsequent atopy. Lancet, 344, 983-985. http://dx.doi.org/10.1016/S014Q-
6736(94)91641-1
Tanios, M. A.; El Gamal, H.; Rosenberg, B. J.; Hassoun, P. M. (2004). Can we still miss
tetrachloroethylene-induced lung disease? The emperor returns in new clothes. Respiration, 71,
642-645. http://dx.doi.org/10.1159/000Q81768
Tao, L.; Kramer, P. M.; Ge, R.; Pereira, M. A. (1998). Effect of dichloroacetic acid and trichloroacetic
acid on DNA methylation in liver and tumors of female B6C3F1 mice. Toxicol Sci, 43, 139-144.
http://dx.doi.org/10.1006/toxs.1998.2449
Taskinen, H.; Anttila, A.; Lindbohm, M.-L.; 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, 345-352.
http://www.ncbi.nlm.nih.gov/pubmed/2799322
Tateishi, M.; Suzuki, S.; Shimizu, H. (1978). Cysteine conjugate beta-lyase in rat liver. A novel enzyme
catalyzing formation of thiol-containing metabolites of drugs. J Biol Chem, 253, 8854-8859.
http: //www .ncbi .nlm .nih. gov/pubmed/7 21818
Tateishi, T.; Nakura, H.; Asoh, M.; Watanabe, M.; Tanaka, M.; Kumai, T., . . . Kobayashi, S. (1997). A
comparison of hepatic cytochrome P450 protein expression between infancy and postinfancy.
Life Sci, 61, 2567-2574. http://dx.doi.org/10.1016/S0024-3205(97)01011-4
Tenkanen, L.; Manttari, M.; Kovanen, P. T.; Virkkunen, H.; Manninen, V. (2006). Gemfibrozil in the
treatment of dyslipidemia: an 18-year mortality follow-up of the Helsinki Heart Study. Arch
Intern Med, 166, 743-748. http://dx.doi.org/10.1001/archinte. 166.7.743
Terrier, P.; Townsend, A. J.; Coindre, J. M.; Triche, T. J.; Cowan, K. H. (1990). An
immunohistochemical study of pi class glutathione S-transferase expression in normal human
tissue. Am J Pathol, 137, 845-853.
Teschke, K.; Morgan, M. S.; Checkoway, H.; Franklin, G.; Spinelli, J. J.; van Belle, G.; Weiss, N. S.
(1997). Surveillance of nasal and bladder cancer to locate sources of exposure to occupational
carcinogens. Occup Environ Med, 54, 443-451. http ://www.ncbi .nlm .nih.gov/pubmed/9245 952
Testing drugs for physical dependence potential and abuse liability. The Committee on Problems of Drug
Dependence, Inc. (1984). NIDA Res Monogr, 52, 1-153.
http: //www .ncbi .nlm .nih. gov/pubmed/6148699
Thai, S.-F.; Allen, J. W.; DeAngelo, A. B.; George, M. H.; Fuscoe, J. C. (2003). Altered gene expression
in mouse livers after dichloroacetic acid exposure. Mutat Res Rev Mutat Res, 543, 167-180.
http://dx.doi.org/10.1016/S1383-5742(03)00014-0
Thomas, J.; Haseman, J. K.; Goodman, J. I.; Ward, J. M.; Loughran, T. P. J.; Spencer, P. J. (2007). A
review of large granular lymphocytic leukemia in Fischer 344 rats as an initial step toward
evaluating the Implication of the endpoint to human cancer risk assessment. Toxicol Sci, 99, 3-
19. http://dx.doi.org/10.1093/toxsci/kfm098
Thomas, K. W.; Pellizzari, E. D.; Perritt, R. L.; Nelson, W. C. (1991). Effect of dry-cleaned clothes on
tetrachloroethylene levels in indoor air, personal air, and breath for residents of several New
Jersey homes. J Expo Anal Environ Epidemiol, 1, 475-490.
http: //www .ncbi .nlm .nih. gov/pubmed/1824329
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-56 DRAFT—DO NOT CITE OR QUOTE
-------
Tichenor, B. A.; Sparks, L. E.; Jackson, M. D.; Guo, Z.; Mason, M. A.; Plunket, C. M.; Rasor, S. A.
(1990). Emissions of perchloroethylene from dry cleaned fabrics. Atmos Environ, 24, 1219-1229.
http://dx.doi.org/10.1016/0960-1686(90)90087-4
Till, C.; Koren, G.; Rovet, J. F. (2001a). Prenatal exposure to organic solvents and child neurobehavioral
performance. Neurotoxicol Teratol, 23, 235-245. http://dx.doi.org/10.1016/SQ892-
0362(01)00141-6
Till, C.; Rovet, J. F.; Koren, G.; Westall, C. A. (2003). Assessment of visual functions following prenatal
exposure to organic solvents. Neurotoxicology, 24, 725-731. http://dx.doi.org/10.1016/SQ161-
813X(02)00212-7
Till, C.; Westall, C. A.; Koren, G.; Nulman, I.; Rovet, J. F. (2005). Vision abnormalities in young
children exposed prenatally to organic solvents. Neurotoxicology, 26, 599-613.
http://dx.doi.Org/10.1016/i.neuro.2005.05.011
Till, C.; Westall, C. A.; Rovet, J. F.; Koren, G. (2001b). Effects of maternal occupational exposure to
organic solvents on offspring visual functioning: A prospective controlled study. Teratology, 64,
134-141. http://dx.doi.org/10.1002/tera.1056
Tinston, D. J. (1994). Perchloroethylene: A multigeneration inhalation study in the rat. (CTL/P/4097,
86950000190). Cheshire, UK: Zeneca Central Toxicology Laboratory.
't Mannetje, A.; Dryson, E.; Walls, C.; McLean, D.; McKenzie, F.; Maule, M., . . . Pearce, N. (2008).
High risk occupations for non-Hodgkin's lymphoma in New Zealand: case-control study. Occup
Environ Med, 65, 354-363. http://dx.doi.Org/10.l 136/oem.2007.035014
t Mannetje, A.; Kogevinas, M.; Chang-Claude, J.; Cordier, S.; Gonzalez, C.-A.; Hours, M., . . . Boffetta,
P. (1999). Occupation and bladder cancer in European women. Cancer Causes Control, 10, 209-
217. http://dx.doi.org/10.1023/A: 1008852127139
Tomisawa, H.; Ichihara, S.; Fukazawa, H.; Ichimoto, N.; Tateishi, M.; Yamamoto, I. (1986). Purification
and characterization of human hepatic cysteine-conjugate beta-lyase. Biochem J, 235, 569-575.
http://www.ncbi.nlm.nih.gov/pubmed/3741406
Tomisawa, H.; Suzuki, S.; Ichihara, S.; Fukazawa, H.; Tateishi, M. (1984). Purification and
characterization of C-S lyase from Fusobacterium varium . A C-S cleavage enzyme of cysteine
conjugates and some S-containing amino acids. J Biol Chem, 259, 2588-2593.
http: //www .ncbi .nlm .nih. gov/pubmed/6698982
Tong, Z.; Board, P. G.; Anders, M. W. (1998a). Glutathione transferase zeta-catalyzed biotransformation
of dichloroacetic acid and other alpha-haloacids. Chem Res Toxicol, 11, 1332-1338.
http: //dx. doi. org/10.1021 /tx9 80144f
Tong, Z.; Board, P. G.; Anders, M. W. (1998b). Glutathione transferase zeta catalyses the oxygenation of
the carcinogen dichloroacetic acid to glyoxylic acid. Biochem J, 331 ( Pt 2), 371-374.
http: //www .ncbi .nlm .nih. gov/pubmed/9531472
Toraason, M.; Butler, M. A.; Ruder, A.; Forrester, C.; Taylor, L.; Ashley, D. L., . . . Wey, H. (2003).
Effect of perchloroethylene, smoking, and race on oxidative DNA damage in female dry cleaners.
Mutat Res Genet Toxicol Environ Mutagen, 539, 9-18. http://dx.doi.org/10.1016/S1383-
5718(03)00130-X
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. http://dx.doi.org/10.1016/S0300-483X(99')00083-9
Travier, N.; Gridley, G.; De Roos, A. J.; Plato, N.; Moradi, T.; Boffetta, P. (2002). Cancer incidence of
dry cleaning, laundry and ironing workers in Sweden. Scand J Work Environ Health, 28, 341 -
348. http://www.ncbi.nlm.nih.gov/pubmed/12432988
Traylor, P. S.; Nastainczyk, W.; Ullrich, V. (1977). Conversion of trichloroethylene to carbon monoxide
by microsomal cytochrome P450.
Trevisan, A.; Cristofori, P.; Fanelli, G. (1999). Glutamine synthetase activity in rat urine as sensitive
marker to detect S3 segment-specific injury of proximal tubule induced by xenobiotics. Arch
Toxicol, 73, 255-262.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-57 DRAFT—DO NOT CITE OR QUOTE
-------
Trevisan, A.; I, M.; Rui, F.; Carrieri, M.; Bartolucci, G. B.; Manno, M. (2000). Kidney and liver
biomakers in female dry-cleaning workers exposed to perchloroethylene. Biomarkers, 5, 399-409.
Trichloroethylene, 54 2432-2434 (1989).
Triclofos sodium (Triclos) for insomnia. (1972). Med Lett Drugs Ther, 14, 78-80.
http://www.ncbi.nlm.nih.gov/pubmed/4567567
Tschopp, C.; Safiran, A. B.; Viviani, P.; Bullinger, A.; Reicherts, M.; Mermoud, C. (1998). Automated
visual field examination in children aged 5-8 years. Part I: Experimental validation of a testing
procedure. Vision Res, 38, 2203-2210. http://www.ncbi.nlm.nih.gov/pubmed/9797979
Tsuruta, H. (1989). Skin absorption of organic solvent vapors in nude mice in vivo. Ind Health, 27, 37-47.
Tu, A. S.; Murray, T. A.; Hatch, K. M.; Sivak, A.; Milman, H. A. (1985). In vitro transformation of
BALB/c-3T3 cells by chlorinated ethanes and ethylenes. Cancer Lett, 28, 85-92.
http://dx.doi.org/10.1016/0304-3835(85)90096-5
Tugwood, J. D.; Aldridge, T. C.; Lambe, K. G.; Macdonald, N.; Woodyatt, N. J. (1996). Peroxisome
proliferator-activated receptors: structures and function. Ann N Y Acad Sci, 804, 252-265.
Tzeng, H. F.; Blackburn, A. C.; Board, P. G.; Anders, M. W. (2000). Polymorphism- and species-
dependent inactivation of glutathione transferase zeta by dichloroacetate. Chem Res Toxicol, 13,
231-236. http://dx.doi.org/10.1021/tx990175a
U.S. EPA. (U.S. Environmental Protection Agency). (1982). An exposure and risk assessment for
tetrachloroethylene. (EPA/440/4-85/015). Office ofWater, Regulations, and Standards. Retrieved
from http://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=2000LLQH.txt.
U.S. EPA. (U.S. Environmental Protection Agency). (1985a). Chemical carcinogens: review of the
science and its associated principles. 50, 10442.
U.S. EPA. (U.S. Environmental Protection Agency). (1985b). Health assessment document for
tetrachloroethylene (perchloroethylene) Final report. (EPA/600/8-82/005F). Research Triangle
Park, NC: U.S. Environmental Protection Agency, Office of Health and Environmental
Assessment. Retrieved from http://cfbub.epa. gov/ncea/cfm/recordisplav.cfm?deid=3 8082.
U.S. EPA. (U.S. Environmental Protection Agency). (1986a). Addendum to the health assessment
document for tetrachloroethylene (perchloroethylene): updated carcinogenicity assessment for
tetrachloroethylene (perchloroethylene, PERC, PCE) [external review draft]. (EP A/600/8-
82/005FA). Washington, DC: U.S. Environmental Protection Agency, Office of Research and
Development. Retrieved from http://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=300016UC.txt.
U.S. EPA. (U.S. Environmental Protection Agency). (1986b). Guidelines for mutagenicity risk
assessment. (EPA/630/R-98/003). Washington, DC: U.S. Environmental Protection Agency, Risk
Assessment Forum. Retrieved from http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (U.S. Environmental Protection Agency). (1986c). Guidelines for the health risk assessment of
chemical mixtures. (EPA/630/R-98/002). Washington, DC: U.S. Environmental Protection
Agency, Risk Assessment Forum. Retrieved from
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=22567.
U.S. EPA. (U.S. Environmental Protection Agency). (1987). Evaluation of the carcinogenicity of
unleaded gasoline. (EPA/600/8-87/001). Office of Health and Environmental Assessment, Office
of Research and Development. Retrieved from http: //www ,p2pavs. org/ref/18/17565.pdf.
U.S. EPA. (U.S. Environmental Protection Agency). (1988a). Integrated Risk Information System (IRIS).
IRIS summary of tetrachloroethylene (CASRN 127-18-4). Office of Research and Development,
National Center for Environmental Assessment. Retrieved from
http://www.epa.gOv/iris/subst/0106.htm.
U.S. EPA. (U.S. Environmental Protection Agency). (1988b). Recommendations for and documentation
of biological values for use in risk assessment. (EPA/600/6-87/008). Cincinnati, OH: U.S.
Environmental Protection Agency, Office of Research and Development. Retrieved from
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=34855.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-58 DRAFT—DO NOT CITE OR QUOTE
-------
U.S. EPA. (U.S. Environmental Protection Agency). (1991a). Guidelines for developmental toxicity risk
assessment. (EPA/600/FR-91/001). Washington, DC: U.S. Environmental Protection Agency,
Risk Assessment Forum. Retrieved from http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (U.S. Environmental Protection Agency). (1991b). Response to issues and the data
submissions on the carcinogenicity of tetrachloroethylene (perchloroethylene). (EPA/600/6-
91/002F). Office of Health and Environmental Assessment. Retrieved from
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=35421.
U.S. EPA. (U.S. Environmental Protection Agency). (1994). Methods for derivation of inhalation
reference concentrations and application of inhalation dosimetry. (EPA/600/8-90/066F).
Washington, DC: U.S. Environmental Protection Agency, Office of Research and Development.
U.S. EPA. (U.S. Environmental Protection Agency). (1995). The use of the benchmark dose approach in
health risk assessment. (EPA/630/R-94/007). Washington, DC: U.S. Environmental Protection
Agency, Risk Assessment Forum.
U.S. EPA. (U.S. Environmental Protection Agency). (1996a). Guidelines for reproductive toxicity risk
assessment. (EPA/630/R-96/009). Washington, DC: U.S. Environmental Protection Agency, Risk
Assessment Forum.
U.S. EPA. (U.S. Environmental Protection Agency). (1996b). Modeled ambient concentrations:
perchloroethylene (CASRN 127-18-4). National Air Toxics Assessment, Technology Transfer
Network, Office of Air and Radiation. Retrieved from
http: //www .epa. gov/ttn/atw/nata/natsa2 .html.
U.S. EPA. (U.S. Environmental Protection Agency). (1998a). Dichloroacetic acid: carcinogenicity
identification characterization summary. (EPA 815-B-98-010). Washington, DC: U.S.
Environmental Protection Agency; National Center for Environmental Assessment. Retrieved
from http://www.ntis.gov/search/product.aspx?ABBR=PB99111387.
U.S. EPA. (U.S. Environmental Protection Agency). (1998b). Guidelines for neurotoxicity risk
assessment. (EPA/630/R-95/00IF). Washington, DC: U.S. Environmental Protection Agency,
Risk Assessment Forum.
U.S. EPA. (U.S. Environmental Protection Agency). (2000a). Benchmark dose technical guidance
document [external review draft]. (EPA/630/R-00/001). Washington, DC: U.S. Environmental
Protection Agency, Risk Assessment Forum. Retrieved from
http://www.epa.gov/raf/publications/benchmark-dose-doc-draft.htm.
U.S. EPA. (U.S. Environmental Protection Agency). (2000b). Science policy council handbook: Risk
characterization. (EPA 100-B-00-002). Washington, D.C.: U.S. Environmental Protection
Agency, Office of Research and Development, Office of Science Policy. Retrieved from
http://www.epa.gov/osa/spc/pdfs/rchandbk.pdf.
U.S. EPA. (U.S. Environmental Protection Agency). (2000c). Supplementary guidance for conducting
health risk assessment of chemical mixtures. (EPA/630/R-00/002). Washington, DC: U.S.
Environmental Protection Agency, Risk Assessment Forum. Retrieved from
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=20533.
U.S. EPA. (U.S. Environmental Protection Agency). (2001a). Sources, emission and exposure for
trichloroethylene (TCE) and related chemicals. (EPA/600/R-00/009). Washington, DC: U.S.
Environmental Protection Agency, Office of Research and Development.
U.S. EPA. (U.S. Environmental Protection Agency). (2001b). Trichloroethylene health risk assessment:
synthesis and characterization [external review draft]. (EPA/600/P-01/002A). National Center for
Environmental Assessment, Office of Research and Development. Retrieved from
http://oaspub.epa.gov/eims/eimscomm.getfile7p download id=4580.
U.S. EPA. (U.S. Environmental Protection Agency). (2002). A review of the reference dose and reference
concentration processes. (EPA/630/P-02/0002F). Washington, DC: U.S. Environmental
Protection Agency, Risk Assessment Forum.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-59 DRAFT—DO NOT CITE OR QUOTE
-------
U.S. EPA. (U.S. Environmental Protection Agency). (2003). Neurotoxicity of tetrachoroethylene
(perchloroethylene): discussion paper. (EPA/600/P-03/005A). National Center for Environmental
Assessment. Retrieved from http://cfpub.epa.gov/ncea/iris drafts/recordisplav.cfm?deid=75193.
U.S. EPA. (U.S. Environmental Protection Agency). (2004). Summary report of the peer review
workshop on the neurotoxicity of tetrachloroethylene (perchloroethylene) discussion paper.
(EPA/600/R-04/041). National Center for Environmental Assessment. Retrieved from
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=84351.
U.S. EPA. (U.S. Environmental Protection Agency). (2005a). Documentation of environmental indicator
determination: RCRA corrective action. IBM-Endicott. New York, NY: Author. Retrieved from
http://www.epa.gov/Region2/waste/ibm e725.pdf.
U.S. EPA. (U.S. Environmental Protection Agency). (2005b). Guidelines for carcinogen risk assessment
[Final report]. (EPA/630/P-03/001F). Washington, DC: U.S. Environmental Protection Agency,
Risk Assessment Forum.
U.S. EPA. (U.S. Environmental Protection Agency). (2005c). Supplemental guidance for assessing
susceptibility from early-life exposure to carcinogens. (EPA/630/R-03/003F). Washington, DC:
U.S. Environmental Protection Agency, Risk Assessment Forum. Retrieved from
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=160003.
U.S. EPA. (U.S. Environmental Protection Agency). (2005d). Trichloroethylene issue paper 3: Role of
peroxisome proliferator-activated receptor agonism and cell signaling in trichloroethylene
toxicity. (EPA/600/R-05/024). Washington, DC: U.S. Environmental Protection Agency, Office
of Research and Development. Retrieved from
http://oaspub.epa.gov/eims/eimscomm.getfile7p download id=438646.
U.S. EPA. (U.S. Environmental Protection Agency). (2006a). Aging and toxic response: issues relevant to
risk assessment. (EPA/600/P-03/004A). National Center for Environmental Assessment.
Retrieved from http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=156648.
U.S. EPA. (U.S. Environmental Protection Agency). (2006b). A framework for assessing health risk of
environmental exposures to children. (EPA/600/R-05/093F). Washington, DC: U.S.
Environmental Protection Agency, Office of Research and Development. Retrieved from
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=158363.
U.S. EPA. (U.S. Environmental Protection Agency). (2006c). Peer review handbook. (EPA/100/B-
06/002). Washington, DC: U.S. Environmental Protection Agency, Science Policy Council.
Retrieved from http://www.epa.gov/peerreview/pdfs/peer review handbook 2006.pdf.
U.S. EPA. (U.S. Environmental Protection Agency). (2008). Child-specific exposure factors handbook.
(EPA/600/R-06/096F). Washington, DC: U.S. Environmental Protection Agency, National Center
for Environmental Assessment. Retrieved from
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=199243.
U.S. EPA. (U.S. Environmental Protection Agency). (2009a). Benchmark dose software (BMDS), from
http: //www .epa. gov/N CEA/bmds
U.S. EPA. (U.S. Environmental Protection Agency). (2009b). Toxicological review of Trichloracetic
Acide (TCA) (External Review Draft). (EPA/635/R-09/003A). Washington, DC: Author.
Uhler, A. D. and Miller, L. J. (1988). Multiple headspace extraction gas chromatography for the
determination of volatile halocarbon compounds in butter. J Agric Food Chem, 36, 772-775.
http://dx.doi.org/10.1021/ifD0Q82a025
Umezu, T.; Yonemoto, J.; Soma, Y.; Miura, T. (1997). Behavioral effects of trichloroethylene and
tetrachloroethylene in mice. Pharmacol Biochem Behav, 58, 665-671.
http://dx.doi.org/10.1016/S0091-3057(97)00046-4
Uttamsingh, V.; Keller, D. A.; Anders, M. W. (1998). Acylase I-catalyzed deacetylation ofN-acetyl-L-
cysteine and S-alkyl-N-acetyl-L-cysteines. Chem Res Toxicol, 11, 800-809.
http://dx.doi.org/10.1021/tx980018b
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-60 DRAFT—DO NOT CITE OR QUOTE
-------
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.
http: //www .ncbi .nlm .nih. gov/pubmed/2263216
Vainio, H.; Parkki, M. G.; Marniemi, J. (1976). Effects of aliphatic chlorohydrocarbons on drug-
metabolizing enzymes in rat liver in vivo. Xenobiotica, 6, 599-604.
Valencia, R.; Mason, J. M.; Woodruff, R. C.; Zimmering, S. (1985). Chemical mutagenesis testing in
Drosophila. III. Results of 48 coded compounds tested for the National Toxicology Program.
Environ Mutagen, 7, 325-348. http://dx.doi.org/10.1002/em.28600703Q9
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, 403-406.
http: //www .ncbi .nlm .nih. gov/pubmed/943 9987
Vamvakas, S.; Dekant, W.; Berthold, K.; Schmidt, S.; Wild, D.; Henschler, D. (1987). Enzymatic
transformation of mercapturic acids derived from halogenated alkenes to reactive and mutagenic
intermediates. Biochem Pharmacol, 36, 2741-2748. http://dx.doi.org/10.1016/00Q6-
2952(87)90258-9
Vamvakas, S.; Dekant, W.; Henschler, D. (1989a). 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, 329-335. http://dx.doi.org/10.1016/0165-1218(89)90108-0
Vamvakas, S.; Dekant, W.; Henschler, D. (1989b). Genotoxicity of haloalkene and haloalkane glutathione
S-conjugates in porcine kidney cells. Toxicol In Vitro, 3, 151-156.
http://dx.doi.org/10.1016/0887-2333(89)90058-1
Vamvakas, S.; Herkenhoff, M.; Dekant, W.; Henschler, D. (1989c). Mutagenicity of tetrachloroethene in
the ames test: Metabolic activation by conjugation with glutathione. J Biochem Toxicol, 4, 21-27.
http://dx.doi.org/10.1002/ibt.25700401Q5
Vamvakas, S.; Kochling, A.; Berthold, K.; Dekant, W. (1989d). Cytotoxicity of cysteine S-conjugates:
structure-activity relationships. Chem Biol Interact, 71, 79-90.
van Lieshout, E. M. and Peters, W. H. (1998). Age and gender dependent levels of glutathione and
glutathione S-transferases in human lymphocytes. Carcinogenesis, 19, 1873-1875.
http: //www .ncbi .nlm .nih. gov/pubmed/9806172
van der Gulden, J. W. J. and Zielhuis, G. A. (1989). Reproductive hazards related to perchloroethylene: A
review. Int Arch Occup Environ Health, 61, 235-242. http://dx.doi.org/10.1007/BF0038142Q
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. http://dx.doi.org/10.1016/0045-6535(93')90165-2
Vaughan, T. L.; Stewart, P. A.; Davis, S.; Thomas, D. B. (1997). Work in dry cleaning and the incidence
of cancer of the oral cavity, larynx, and oesophagus. Occup Environ Med, 54, 692-695.
http: //www .ncbi .nlm .nih. gov/pubmed/9423 5 85
Veitch, J. M.; Murray, G. I.; Juronen, E.; Kilty, C. G.; Doyle, S.; Hawksworth, G. M.; Rodilla, V. (1997).
Theta-class glutathione S-transferases in human kidney and renal tumours. Biochem Soc Trans,
25, S605.
Verberk, M. M. and Scheffers, T. M. L. (1980). Tetrachloroethylene in exhaled air of residents near dry-
cleaning shops. Environ Res, 21, 432-437.
Verplanke, A. J.; Leummens, M. H.; Herber, R. F. (1999). Occupational exposure to tetrachloroethene
and its effects on the kidneys. J Occup Environ Med, 41, 11-16.
http: //www .ncbi .nlm .nih. gov/pubmed/9924715
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,
476-483. http://dx.doi.Org/10.llll/i.1432-1033.1996.0476z.x
Vieira, V.; Aschengrau, A.; Ozonoff, D. (2005a). Impact of tetrachloroethylene-contaminated drinking
water on the risk of breast cancer: using a dose model to assess exposure in a case-control study.
Environ Health, 4, 3. http://dx.doi.Org/10.l 186/1476-069X-4-3
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-61 DRAFT—DO NOT CITE OR QUOTE
-------
Vieira, V.; Webster, T.; Weinberg, J.; Aschengrau, A.; Ozonoff, D. (2005b). Spatial analysis of lung,
colorectal, and breast cancer on Cape Cod: An application of generalized additive models to case-
control data. Environ Health, 4,11. http://dx.doi.Org/10.l 186/1476-069X-4-11
Volkel, W.; Friedewald, M.; Lederer, E.; Pahler, A.; Parker, J.; Dekant, W. (1998). Biotransformation of
perchloroethene: Dose-dependent excretion of trichloroacetic acid, dichloroacetic acid, and N-
acetyl-S-(trichlorovinyl)-L-cysteine in rats and humans after inhalation. Toxicol Appl Pharmacol,
153, 20-27. http://dx.doi.org/10.10Q6/taap.1998.8548
Volkel, W.; Pahler, A.; Dekant, W. (1999). Gas chromatography-negative ion chemical ionization mass
spectrometry as a powerful tool for the detection of mercapturic acids and DNA and protein
adducts as biomarkers of exposure to halogenated olefins. J Chromatogr A, 847, 35-46.
von der Hude, W.; Behm, C.; Giirtler, R.; Basler, A. (1988). Evaluation of the SOS chromotest. Mutat
Res, 203, 81-94. http://dx.doi.org/10.1016/0165-l 161(88)90023-4
Von Tungeln, L. S.; Yi, P.; Bucci, T. J.; Samokyszyn, V. M.; Chou, M. W.; Kadlubar, F. F.; Fu, P. P.
(2002). Tumorigenicity of chloral hydrate, trichloroacetic acid, trichloroethanol,
malondialdehyde, 4-hydroxy-2-nonenal, crotonaldehyde, and acrolein in the B6C3F1 neonatal
mouse. Cancer Lett, 185, 13-19. http://dx.doi.org/10.1016/50304-3835(02)00231-8
Vyskocil, A.; Emminger, S.; Tejral, J.; Fiala, Z.; Ettlerova, E.; Cermanova, A. (1990). Study on kidney
function in female workers exposed to perchlorethylene. Hum Exp Toxicol, 9, 377-380.
http: //www .ncbi .nlm .nih. gov/pubmed/2271228
W.H.O. cooperative trial on primary prevention of ischaemic heart disease using clofibrate to lower
serum cholesterol: mortality follow-up. Report of the Committee of Principal Investigators.
(1980). Lancet, 2, 379-385. http://www.ncbi.nlm.nih.gov/pubmed/6105515
Walgren, J. E.; Kurtz, D. T.; McMillan, J. M. (2000). 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. http://dx.doi.Org/10.1023/A: 1007638227821
Walles, S. A. S. (1986). Induction of single-strand breaks in dna of mice by trichloroethylene and
tetrachloroethylene. Toxicol Lett, 31, 31-35. http://dx.doi.org/10.1016/0378-4274(86)90191-8
Wang, J.-L.; Chen, W.-L.; Tsai, S.-Y.; Sung, P.-Y.; Huang, R.-N. (2001). An in vitro model for
evaluation of vaporous toxicity of trichloroethylene and tetrachloroethylene to CHO-K1 cells.
Chem Biol Interact, 137, 139-154. http://dx.doi.org/10.1016/S0009-2797(0n00226-5
Wang, S.; Karlsson, J.-E.; Kyrklund, T.; Haglid, K. (1993). Perchloroethylene-induced reduction in glial
and neuronal cell marker proteins in rat brain. Basic Clin Pharmacol Toxicol, 72, 273-278.
http://dx.doi.Org/10.llll/i.1600-0773.1993.tb01649.x
Wanner, M.; Lehmann, E.; Morel, J.; Christen, R. (1982). [The transfer of perchloroethylene from animal
feed to milk], Mitteilungen aus dem Gebiete der Lebensmittel-untersuchung un Hygiene, 73, 82-
87.
Ward, J. M. and Reynolds, C. W. (1983). Large granular lymphocyte leukemia. A heterogeneous
lymphocytic leukemia in F344 rats. Am J Pathol, 111, 1-10.
http: //www .ncbi .nlm .nih. gov/pubmed/683 7719
Ward, R. C.; Travis, C. C.; Hetrick, D. M.; Andersen, M. E.; Gargas, M. L. (1988). Pharmacokinetics of
tetrachloroethylene. Toxicol Appl Pharmacol, 93, 108-117. http://dx.doi.org/10.1016/0Q41-
008X(88)90030-0
Warner, J. A.; Miles, E. A.; Jones, A. C.; Quint, D. J.; Colwell, B. M.; Warner, J. O. (1994). Is deficiency
of interferon gamma production by allergen triggered cord blood cells a predictor of atopic
eczema? Clin Exp Allergy, 24, 423-430. http://dx.doi.Org/10.l 111/i. 1365-2222.1994,tb00930.x
Warr, T.; Parry, E.; Parry, J. (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, 29-
46. http: //www. ncbi .nlm. nih. go v/pubme d/7683 3 82
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-62 DRAFT—DO NOT CITE OR QUOTE
-------
Warren, D. A.; Reigle, T. G.; Muralidhara, S.; Dallas, C. E. (1996). Schedule-controlled operant behavior
of rats following oral administration of perchloroethylene: Time course and relationship to blood
and brain solvent levels. J Toxicol Environ Health, 47, 345-362.
http://dx.doi.org/10.1080/00984109616169Q
Waskell, L. (1978). A study of the mutagenicity of anesthetics and their metabolites. DNA Repair, 57,
141-153. htto://www.ncbi.nlm.nih.go v/pubmed/351387
Watanabe, K.; Satamoto, K.; Sasaki, T. (1998). Comparisons on chemically-induced mutation among four
bacterial strains, Salmonella typhimurium TA102 and TA2638, and Escherichia coli
WP2/pKM101 and WP2 uvrA/pKMlOl: Collaborative study II. Mutat Res, 412, 17-31.
http://dx.doi.org/10.1016/S1383-5718(97)00155-1
Webler, T. and Brown, H. S. (1993). Exposure to tetrachloroethylene via contaminated drinking water
pipes in Massachusetts: a predictive model. Arch Environ Health, 48, 293-297.
Weichardt, H. and Lindner, J. (1975). Gesundheitsgefahren durch Perchlorathylene in Chemisch-
Reinigungsbetrieben arbeitsmedizinisch-toxikologischer Sicht [Health hazards caused by
perchloroethylene in dry cleaning plants, from the viewpoint of occupational medicine and
toxicology]. Gefahrst Reinhalt Luft, 35, 416-420.
Werner, M.; Birner, G.; Dekant, W. (1996). Sulfoxidation of mercapturic acids derived from tri- and
tetrachloroethene by cytochromes P450 3A: A bioactivation reaction in addition to deacetylation
and cysteine conjugate beta-lyase mediated cleavage. Chem Res Toxicol, 9, 41-49.
http://dx.doi.org/10.1021/tx950075u
Westlind, A.; Lofberg, L.; Tindberg, N.; Andersson, T. B.; Ingelman-Sundberg, M. (1999).
Interindividual differences in hepatic expression of CYP3A4: relationship to genetic
polymorphism in the 5'-upstream regulatory region. Biochem Biophys Res Commun, 259, 201-
205.
White, I. N.; Razvi, N.; Gibbs, A. H.; Davies, A. M.; Manno, M.; Zaccaro, C., . . . Dekant, W. (2001).
Neoantigen formation and clastogenic action of HCFC-123 and perchloroethylene in human
MCL-5 cells. Toxicol Lett, 124, 129-138. http://dx.doi.org/10.1016/S0378-4274(00)00281-2
WHO. (World Health Organization). (2006). Concise international chemical assessment document 68:
Tetrachloroethene. Geneva, Switzerland: World Health Organization, International Programme
on Chemical Safety. Retrieved from
http: //www .inchem ,org/documents/cicads/cicads/cicad6 8 .htm.
WHO cooperative trial on primary prevention of ischaemic heart disease with clofibrate to lower serum
cholesterol: final mortality follow-up. Report of the Committee of Principal Investigators. (1984).
Lancet, 2, 600-604. http://www ncbi nlm nih gov/pubmed/6147641
Wilcox, A. J. and Horney, L. F. (1984). Accuracy of spontaneous abortion recall. Am J Epidemiol, 120,
727-733.
Williams, R. L.; Creasy, R. K.; Cunningham, G. C.; Hawes, W. E.; Norris, F. D.; Tashiro, M. (1982).
Fetal growth and perinatal viability in California. Obstet Gynecol, 59, 624-632.
Wilson, R.; Donahue, M.; Gridley, G.; Adami, J.; El Ghormli, L.; Dosemeci, M. (2008). Shared
occupational risks for transitional cell cancer of the bladder and renal pelvis among men and
women in Sweden. Am J Ind Med, 51, 83-99. http://dx.doi.org/10.1002/aiim.20522
Windham, G. C.; Shusterman, D.; Swan, S. H.; Fenster, L.; Eskenazi, B. (1991). Exposure to organic
solvents and adverse pregnancy outcome. Am J Ind Med, 20, 241-259.
http://dx.doi.org/10.1002/aiim.470020021Q
Wiseman, H. and Hampel, G. (1978). Cardiac arrhythmias due to chloral hydrate poisoning. Br Med J, 2,
960. http ://www.ncbi .nlm.nih. gov/pubmed/709162
Wolf, C. J.; Takacs, M. L.; Schmid, J. E.; Lau, C.; Abbott, B. D. (2008). Activation of mouse and human
peroxisome proliferator-activated receptor alpha by perfluoroalkyl acids of different functional
groups and chain lengths. Toxicol Sci, 106, 162-171. http://dx.doi.org/10.1093/toxsci/kfhl66
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-63 DRAFT—DO NOT CITE OR QUOTE
-------
Wood, P. and Cheney, D. (1979). The effects of muscarinic receptor blockers on the turnover rate of
acetylcholine in various regions of the rat brain. Can J Physiol Pharmacol, 57, 404-411.
http://www.ncbi.nlm.nih.gov/pubmed/455141
Xu, Y.; Iyengar, S.; Roberts, R.; Shappell, S.; Peehl, D. (2003). Primary culture model of peroxisome
proliferator-activated receptor gamma activity in prostate cancer cells. J Cell Physiol, 196, 131-
143. http://dx.doi.org/10.1002/icp.10281
Yang, Q.; Ito, S.; Gonzalez, F. J. (2007). Hepatocyte-restricted constitutive activation of PPAR-A
induces hepatoproliferation but not hepatocarcinogenesis. Carcinogenesis, 28, 1171-1177.
Yang, R.; Elmasri, H.; Thomas, R.; Dobrev, I.; Dennisonjr, J.; Bae, D., . . . Andersen, M. (2004).
Chemical mixture toxicology: from descriptive to mechanistic, and going on to in silico
toxicology. Environ Toxicol Pharmacol, 18, 65-81. http://dx.doi.Org/10.1016/i.etap.2004.01.015
Yllner, S. (1961). Urinary metabolites of 14C-tetrachloroethylene in mice. Nature, 191, 820.
Yokley, K. A. and Evans, M. V. (2007). An example of model structure differences using sensitivity
analyses in physiologically based pharmacokinetic models of trichloroethylene in humans. Bull
Math Biol, 69, 2591-2625. http://dx.doi.org/10.1007/sll538-007-9233-x
Yoo, J. S.; Guengerich, F. P.; Yang, C. S. (1988). Metabolism of N-nitrosodialkylamines by human liver
microsomes. Cancer Res, 48, 1499-1504.
Yoon, J. S.; Mason, J. M.; Valencia, R.; Woodruff, R. C.; Zimmering, S. (1985). Chemical mutagenesis
testing in Drosophila. IV. Results of 45 coded compounds tested for the National Toxicology
Program. Environ Mutagen, 7, 349-367. http://dx.doi.org/10.1002/em.286007031Q
Youssef, J. and Badr, M. (1998). Extraperoxisomal targets of peroxisome proliferators: Mitochondrial,
microsomal, and cytosolic effects: Implications for health and disease. Crit Rev Toxicol, 28, 1-33.
http://dx.doi.org/10.1080/104Q8449891344182
Yu, S.; Cao, W. Q.; Kashireddy, P.; Meyer, K.; Jia, Y.; Hughes, D. E., . . . Reddy, J. K. (2001). Human
peroxisome proliferator-activated receptor alpha (PPARA ) supports the induction of peroxisome
proliferation in PPARA -deficient mouse liver. J Biol Chem, 276, 42485-42491.
Zeka, A.; Mannetje, A.; Zaridze, D.; Szeszenia-Dabrowska, N.; Rudnai, P.; Lissowska, J., . . . Boffetta, P.
(2006a). Lung cancer and occupation in nonsmokers: a multicenter case-control study in Europe.
Epidemiology, 17, 615-623. http://dx.doi.org/10.1097/01.ede.000Q239582.92495.b5
Zeka, A.; Zanobetti, A.; Schwartz, J. (2006b). Individual-level modifiers of the effects of particulate
matter on daily mortality. Am J Epidemiol, 163, 849-859.
Zendell, E. (1972). Chloral hydrate overdose. Case report. Anesth Prog, 19, 165-168.
http://www.ncbi.nlm.nih.gov/pubmed/4510320
Zhang, T.; Wang, L.; Hashmi, M.; Anders, M.; Thorpe, C.; Ridge, D. (1995). Fourier-transform ion
cyclotron resonance mass spectrometric evidence for the formation of alpha-chloroenethiolates
and thioketenes from chloroalkene-derived, cytotoxic 4-thiaalkanoates. Chem Res Toxicol, 8,
907-910. http://www ncbi nlm nih gov/pubmed/8555404
Zheng, T.; Cantor, K. P.; Zhang, Y.; Lynch, C. F. (2002). Occupation and bladder cancer: A population-
based, case-control study in Iowa. J Occup Environ Med, 44, 685-691.
http: //www .ncbi .nlm .nih. gov/pubmed/12134533
Zhou, Y. C. and Waxman, D. J. (1998). Activation of peroxisome proliferator-activated receptors by
chlorinated hydrocarbons and endogenous steroids. Environ Health Perspect, 106 Suppl 4, 983-
988. http://www nchi nlm nih gov/pubmed/9703482
Zielhuis, G. A.; Gijsen, R.; van der Gulden, J. W. J. (1989). Menstrual disorders among dry-cleaning
workers. Scand J Work Environ Health, 15, 238. http://www.ncbi.nlm.nih.gov/pubmed/2781255
Zingg, J. M. and Jones, P. A. (1997). Genetic and epigenetic aspects of DNA methylation on genome
expression, evolution, mutation and carcinogenesis. Carcinogenesis, 18, 869-882.
This document is a draft for review purposes only and does not constitute Agency policy.
6/24/11 7-64 DRAFT—DO NOT CITE OR QUOTE
------- |